{ "dataset": "LoomData caption-count sample", "model": "gpt-5.6-sol", "prompt_file": "videoloom_referring_expression_final_prompt.txt", "annotation_format": "Normalized to canonical field names; original model output retained per record as model_output_raw. Span offsets are exact caption code-point offsets; time intervals are clip-relative seconds with mapped source-video times.", "annotations": [ { "source_video_id": "v_0J55EGFLglQ", "caption_count_category": 1, "clip": { "caption_index": 1, "caption": "A person in a black long-sleeved jersey with the number 5 runs forward on the court toward the right. They start from the center circle and run into the penalty area. Their pace gradually slows as they approach the goal, and they eventually come to a complete stop, standing still in front of the goalpost.", "source_interval_sec": [ 8.11, 19.72 ], "loom_clip_id": "v_0J55EGFLglQ_243_591", "clip_frame_interval_4fps": [ 243, 591 ], "sample_count_1fps": 12, "sampled_frames": [ { "frame": "00000", "timestamp_sec": 8.11, "offset_sec": 0, "rgb_path": "assets/cat_01/v_0J55EGFLglQ_243_591/0000s_00000_rgb.jpg", "mask_path": "assets/cat_01/v_0J55EGFLglQ_243_591/0000s_00000_mask.png", "mask_status": "available" }, { "frame": "00028", "timestamp_sec": 9.11, "offset_sec": 1, "rgb_path": "assets/cat_01/v_0J55EGFLglQ_243_591/0001s_00028_rgb.jpg", "mask_path": "assets/cat_01/v_0J55EGFLglQ_243_591/0001s_00028_mask.png", "mask_status": "available" }, { "frame": "00060", "timestamp_sec": 10.11, "offset_sec": 2, "rgb_path": "assets/cat_01/v_0J55EGFLglQ_243_591/0002s_00060_rgb.jpg", "mask_path": "assets/cat_01/v_0J55EGFLglQ_243_591/0002s_00060_mask.png", "mask_status": "available" }, { "frame": "00088", "timestamp_sec": 11.11, "offset_sec": 3, "rgb_path": "assets/cat_01/v_0J55EGFLglQ_243_591/0003s_00088_rgb.jpg", "mask_path": "assets/cat_01/v_0J55EGFLglQ_243_591/0003s_00088_mask.png", "mask_status": "available" }, { "frame": "00120", "timestamp_sec": 12.11, "offset_sec": 4, "rgb_path": "assets/cat_01/v_0J55EGFLglQ_243_591/0004s_00120_rgb.jpg", "mask_path": "assets/cat_01/v_0J55EGFLglQ_243_591/0004s_00120_mask.png", "mask_status": "available" }, { "frame": "00148", "timestamp_sec": 13.11, "offset_sec": 5, "rgb_path": "assets/cat_01/v_0J55EGFLglQ_243_591/0005s_00148_rgb.jpg", "mask_path": "assets/cat_01/v_0J55EGFLglQ_243_591/0005s_00148_mask.png", "mask_status": "available" }, { "frame": "00180", "timestamp_sec": 14.11, "offset_sec": 6, "rgb_path": "assets/cat_01/v_0J55EGFLglQ_243_591/0006s_00180_rgb.jpg", "mask_path": "assets/cat_01/v_0J55EGFLglQ_243_591/0006s_00180_mask.png", "mask_status": "available" }, { "frame": "00208", "timestamp_sec": 15.11, "offset_sec": 7, "rgb_path": "assets/cat_01/v_0J55EGFLglQ_243_591/0007s_00208_rgb.jpg", "mask_path": "assets/cat_01/v_0J55EGFLglQ_243_591/0007s_00208_mask.png", "mask_status": "available" }, { "frame": "00240", "timestamp_sec": 16.11, "offset_sec": 8, "rgb_path": "assets/cat_01/v_0J55EGFLglQ_243_591/0008s_00240_rgb.jpg", "mask_path": "assets/cat_01/v_0J55EGFLglQ_243_591/0008s_00240_mask.png", "mask_status": "available" }, { "frame": "00268", "timestamp_sec": 17.11, "offset_sec": 9, "rgb_path": "assets/cat_01/v_0J55EGFLglQ_243_591/0009s_00268_rgb.jpg", "mask_path": "assets/cat_01/v_0J55EGFLglQ_243_591/0009s_00268_mask.png", "mask_status": "available" }, { "frame": "00300", "timestamp_sec": 18.11, "offset_sec": 10, "rgb_path": "assets/cat_01/v_0J55EGFLglQ_243_591/0010s_00300_rgb.jpg", "mask_path": "assets/cat_01/v_0J55EGFLglQ_243_591/0010s_00300_mask.png", "mask_status": "available" }, { "frame": "00328", "timestamp_sec": 19.11, "offset_sec": 11, "rgb_path": "assets/cat_01/v_0J55EGFLglQ_243_591/0011s_00328_rgb.jpg", "mask_path": "assets/cat_01/v_0J55EGFLglQ_243_591/0011s_00328_mask.png", "mask_status": "available" } ], "obj_id": [ 0 ], "anno_id": [ 225 ], "fps_inferred_from_clip_interval": 29.9742 }, "usage": { "input_tokens": 27691, "input_tokens_details": { "cache_write_tokens": 27688, "cached_tokens": 0 }, "output_tokens": 5212, "output_tokens_details": { "reasoning_tokens": 3259 }, "total_tokens": 32903 }, "model": "gpt-5.6-sol", "effort": "medium", "custom_id": "pilot", "annotation": { "caption": "A person in a black long-sleeved jersey with the number 5 runs forward on the court toward the right. They start from the center circle and run into the penalty area. Their pace gradually slows as they approach the goal, and they eventually come to a complete stop, standing still in front of the goalpost.", "entity_groups": [ { "id": "person_1", "label": "person", "spans": [ { "text": "A person", "range": [ 0, 8 ] }, { "text": "They", "range": [ 102, 106 ] }, { "text": "Their", "range": [ 167, 172 ] }, { "text": "they", "range": [ 197, 201 ] }, { "text": "they", "range": [ 225, 229 ] } ], "source_mask_link": "selected_person" }, { "id": "jersey_1", "label": "jersey", "spans": [ { "text": "a black long-sleeved jersey", "range": [ 12, 39 ] } ], "source_mask_link": null }, { "id": "court_1", "label": "court", "spans": [ { "text": "the court", "range": [ 74, 83 ] } ], "source_mask_link": null }, { "id": "center_circle_1", "label": "center circle", "spans": [ { "text": "the center circle", "range": [ 118, 135 ] } ], "source_mask_link": null }, { "id": "goal_1", "label": "goal", "spans": [ { "text": "the goal", "range": [ 211, 219 ] } ], "source_mask_link": null }, { "id": "goalpost_1", "label": "goalpost", "spans": [ { "text": "the goalpost", "range": [ 293, 305 ] } ], "source_mask_link": null } ], "actions": [ { "id": "action_1", "mentions": [ { "text": "runs", "ranges": [ [ 58, 62 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } }, { "text": "run", "ranges": [ [ 140, 143 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "type": "simple", "subject_refs": [ "person_1" ], "object_refs": [], "confidence": "high", "temporal_status": "localized", "intervals": [ { "start_s": 1.0, "end_s": 9.0, "source_start_s": 9.11, "source_end_s": 17.11 } ], "evidence": "The masked player wearing number 5 advances rapidly toward the right in running poses across the clean frames sampled from 1.000s through 8.000s.", "mask_relation": "The action subject matches the supplied selected-person mask.", "source_subject_refs": "person_1" }, { "id": "action_2", "mentions": [ { "text": "slows", "ranges": [ [ 188, 193 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "type": "simple", "subject_refs": [ "person_1" ], "object_refs": [], "confidence": "medium", "temporal_status": "localized", "intervals": [ { "start_s": 8.0, "end_s": 11.0, "source_start_s": 16.11, "source_end_s": 19.11 } ], "evidence": "From 8.000s onward, the masked player's displacement decreases and the running pose transitions toward an upright stationary posture by 10.000s.", "mask_relation": "The action subject matches the supplied selected-person mask.", "source_subject_refs": "person_1" }, { "id": "action_3", "mentions": [ { "text": "approach", "ranges": [ [ 202, 210 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "type": "simple", "subject_refs": [ "person_1" ], "object_refs": [ "goal_1" ], "confidence": "high", "temporal_status": "localized", "intervals": [ { "start_s": 4.0, "end_s": 10.0, "source_start_s": 12.11, "source_end_s": 18.11 } ], "evidence": "The masked player moves through the penalty area toward the visible goal between the clean frames at 4.000s and 9.000s.", "mask_relation": "The action subject matches the supplied selected-person mask.", "source_subject_refs": "person_1" }, { "id": "action_4", "mentions": [ { "text": "come to a complete stop", "ranges": [ [ 241, 264 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "type": "light_verb", "subject_refs": [ "person_1" ], "object_refs": [], "confidence": "high", "temporal_status": "localized", "intervals": [ { "start_s": 9.0, "end_s": 11.0, "source_start_s": 17.11, "source_end_s": 19.11 } ], "evidence": "The masked player transitions from a final step near the goal at 9.000s to an upright stopped position by 10.000s.", "mask_relation": "The action subject matches the supplied selected-person mask.", "source_subject_refs": "person_1" }, { "id": "action_5", "mentions": [ { "text": "standing", "ranges": [ [ 266, 274 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "type": "posture", "subject_refs": [ "person_1" ], "object_refs": [], "confidence": "high", "temporal_status": "localized", "intervals": [ { "start_s": 10.0, "end_s": 11.61, "source_start_s": 18.11, "source_end_s": 19.72 } ], "evidence": "The masked player is upright and stationary in front of the goal in the clean frames at 10.000s and 11.000s, with support continuing to the clip boundary.", "mask_relation": "The action subject matches the supplied selected-person mask.", "source_subject_refs": "person_1" } ], "notes": [ "The only supplied masklet identifies the black-uniformed player wearing number 5 and links to person_1.", "The mask is empty at 0.000s because the selected player is obscured by the opening graphic; action localization begins with the first visible masked sample at 1.000s.", "No separate masks were supplied for the jersey, court, center circle, goal, or goalpost, so their source-mask links are null.", "The phrases \"the right\" and \"the penalty area\" were not entity-grouped because they denote spatial regions; \"pace\" and \"a complete stop\" were not entity-grouped because they denote an abstract property and an event.", "\"start\" was not annotated as a separate action because it frames the beginning of the running event." ] }, "model_output_raw": { "caption": "A person in a black long-sleeved jersey with the number 5 runs forward on the court toward the right. They start from the center circle and run into the penalty area. Their pace gradually slows as they approach the goal, and they eventually come to a complete stop, standing still in front of the goalpost.", "entity_groups": [ { "id": "person_1", "label": "person", "spans": [ { "text": "A person", "start": 0, "end": 8 }, { "text": "They", "start": 102, "end": 106 }, { "text": "Their", "start": 167, "end": 172 }, { "text": "they", "start": 197, "end": 201 }, { "text": "they", "start": 225, "end": 229 } ], "source_mask": "selected_person" }, { "id": "jersey_1", "label": "jersey", "spans": [ { "text": "a black long-sleeved jersey", "start": 12, "end": 39 } ], "source_mask": null }, { "id": "court_1", "label": "court", "spans": [ { "text": "the court", "start": 74, "end": 83 } ], "source_mask": null }, { "id": "center_circle_1", "label": "center circle", "spans": [ { "text": "the center circle", "start": 118, "end": 135 } ], "source_mask": null }, { "id": "goal_1", "label": "goal", "spans": [ { "text": "the goal", "start": 211, "end": 219 } ], "source_mask": null }, { "id": "goalpost_1", "label": "goalpost", "spans": [ { "text": "the goalpost", "start": 293, "end": 305 } ], "source_mask": null } ], "actions": [ { "id": "action_1", "mentions": [ { "text": "runs", "ranges": [ [ 58, 62 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } }, { "text": "run", "ranges": [ [ 140, 143 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "type": "simple", "subject_ref": "person_1", "object_refs": [], "confidence": "high", "temporal_status": "localized", "intervals": [ { "start_s": 1.0, "end_s": 9.0, "source_start_s": 9.11, "source_end_s": 17.11 } ], "evidence": "The masked player wearing number 5 advances rapidly toward the right in running poses across the clean frames sampled from 1.000s through 8.000s.", "mask_relation": "The action subject matches the supplied selected-person mask." }, { "id": "action_2", "mentions": [ { "text": "slows", "ranges": [ [ 188, 193 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "type": "simple", "subject_ref": "person_1", "object_refs": [], "confidence": "medium", "temporal_status": "localized", "intervals": [ { "start_s": 8.0, "end_s": 11.0, "source_start_s": 16.11, "source_end_s": 19.11 } ], "evidence": "From 8.000s onward, the masked player's displacement decreases and the running pose transitions toward an upright stationary posture by 10.000s.", "mask_relation": "The action subject matches the supplied selected-person mask." }, { "id": "action_3", "mentions": [ { "text": "approach", "ranges": [ [ 202, 210 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "type": "simple", "subject_ref": "person_1", "object_refs": [ "goal_1" ], "confidence": "high", "temporal_status": "localized", "intervals": [ { "start_s": 4.0, "end_s": 10.0, "source_start_s": 12.11, "source_end_s": 18.11 } ], "evidence": "The masked player moves through the penalty area toward the visible goal between the clean frames at 4.000s and 9.000s.", "mask_relation": "The action subject matches the supplied selected-person mask." }, { "id": "action_4", "mentions": [ { "text": "come to a complete stop", "ranges": [ [ 241, 264 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "type": "light_verb", "subject_ref": "person_1", "object_refs": [], "confidence": "high", "temporal_status": "localized", "intervals": [ { "start_s": 9.0, "end_s": 11.0, "source_start_s": 17.11, "source_end_s": 19.11 } ], "evidence": "The masked player transitions from a final step near the goal at 9.000s to an upright stopped position by 10.000s.", "mask_relation": "The action subject matches the supplied selected-person mask." }, { "id": "action_5", "mentions": [ { "text": "standing", "ranges": [ [ 266, 274 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "type": "posture", "subject_ref": "person_1", "object_refs": [], "confidence": "high", "temporal_status": "localized", "intervals": [ { "start_s": 10.0, "end_s": 11.61, "source_start_s": 18.11, "source_end_s": 19.72 } ], "evidence": "The masked player is upright and stationary in front of the goal in the clean frames at 10.000s and 11.000s, with support continuing to the clip boundary.", "mask_relation": "The action subject matches the supplied selected-person mask." } ], "notes": [ "The only supplied masklet identifies the black-uniformed player wearing number 5 and links to person_1.", "The mask is empty at 0.000s because the selected player is obscured by the opening graphic; action localization begins with the first visible masked sample at 1.000s.", "No separate masks were supplied for the jersey, court, center circle, goal, or goalpost, so their source-mask links are null.", "The phrases \"the right\" and \"the penalty area\" were not entity-grouped because they denote spatial regions; \"pace\" and \"a complete stop\" were not entity-grouped because they denote an abstract property and an event.", "\"start\" was not annotated as a separate action because it frames the beginning of the running event." ] }, "normalization_repairs": 0 }, { "source_video_id": "v_-DaX_1bBNAo", "caption_count_category": 2, "clip": { "caption_index": 1, "caption": "A person with long, wavy blonde hair, wearing a white lace bra and large white angel wings, smiles while looking to their left. They raise a small blue party blower to their mouth and hold it there for a moment. After lowering it, they raise the blower again towards their open mouth before bringing their arm back down.", "source_interval_sec": [ 0.0, 11.5 ], "loom_clip_id": "v_-DaX_1bBNAo_1_345", "clip_frame_interval_4fps": [ 1, 345 ], "sample_count_1fps": 12, "sampled_frames": [ { "frame": "00000", "timestamp_sec": 0.0, "offset_sec": 0, "rgb_path": "assets/cat_02/v_-DaX_1bBNAo_1_345/0000s_00000_rgb.jpg", "mask_path": "assets/cat_02/v_-DaX_1bBNAo_1_345/0000s_00000_mask.png", "mask_status": "available" }, { "frame": "00028", "timestamp_sec": 1.0, "offset_sec": 1, "rgb_path": "assets/cat_02/v_-DaX_1bBNAo_1_345/0001s_00028_rgb.jpg", "mask_path": "assets/cat_02/v_-DaX_1bBNAo_1_345/0001s_00028_mask.png", "mask_status": "available" }, { "frame": "00060", "timestamp_sec": 2.0, "offset_sec": 2, "rgb_path": "assets/cat_02/v_-DaX_1bBNAo_1_345/0002s_00060_rgb.jpg", "mask_path": "assets/cat_02/v_-DaX_1bBNAo_1_345/0002s_00060_mask.png", "mask_status": "available" }, { "frame": "00088", "timestamp_sec": 3.0, "offset_sec": 3, "rgb_path": "assets/cat_02/v_-DaX_1bBNAo_1_345/0003s_00088_rgb.jpg", "mask_path": "assets/cat_02/v_-DaX_1bBNAo_1_345/0003s_00088_mask.png", "mask_status": "available" }, { "frame": "00120", "timestamp_sec": 4.0, "offset_sec": 4, "rgb_path": "assets/cat_02/v_-DaX_1bBNAo_1_345/0004s_00120_rgb.jpg", "mask_path": "assets/cat_02/v_-DaX_1bBNAo_1_345/0004s_00120_mask.png", "mask_status": "available" }, { "frame": "00148", "timestamp_sec": 5.0, "offset_sec": 5, "rgb_path": "assets/cat_02/v_-DaX_1bBNAo_1_345/0005s_00148_rgb.jpg", "mask_path": "assets/cat_02/v_-DaX_1bBNAo_1_345/0005s_00148_mask.png", "mask_status": "available" }, { "frame": "00180", "timestamp_sec": 6.0, "offset_sec": 6, "rgb_path": "assets/cat_02/v_-DaX_1bBNAo_1_345/0006s_00180_rgb.jpg", "mask_path": "assets/cat_02/v_-DaX_1bBNAo_1_345/0006s_00180_mask.png", "mask_status": "available" }, { "frame": "00208", "timestamp_sec": 7.0, "offset_sec": 7, "rgb_path": "assets/cat_02/v_-DaX_1bBNAo_1_345/0007s_00208_rgb.jpg", "mask_path": "assets/cat_02/v_-DaX_1bBNAo_1_345/0007s_00208_mask.png", "mask_status": "available" }, { "frame": "00240", "timestamp_sec": 8.0, "offset_sec": 8, "rgb_path": "assets/cat_02/v_-DaX_1bBNAo_1_345/0008s_00240_rgb.jpg", "mask_path": "assets/cat_02/v_-DaX_1bBNAo_1_345/0008s_00240_mask.png", "mask_status": "available" }, { "frame": "00268", "timestamp_sec": 9.0, "offset_sec": 9, "rgb_path": "assets/cat_02/v_-DaX_1bBNAo_1_345/0009s_00268_rgb.jpg", "mask_path": "assets/cat_02/v_-DaX_1bBNAo_1_345/0009s_00268_mask.png", "mask_status": "available" }, { "frame": "00300", "timestamp_sec": 10.0, "offset_sec": 10, "rgb_path": "assets/cat_02/v_-DaX_1bBNAo_1_345/0010s_00300_rgb.jpg", "mask_path": "assets/cat_02/v_-DaX_1bBNAo_1_345/0010s_00300_mask.png", "mask_status": "available" }, { "frame": "00328", "timestamp_sec": 11.0, "offset_sec": 11, "rgb_path": "assets/cat_02/v_-DaX_1bBNAo_1_345/0011s_00328_rgb.jpg", "mask_path": "assets/cat_02/v_-DaX_1bBNAo_1_345/0011s_00328_mask.png", "mask_status": "available" } ], "obj_id": [ 0 ], "anno_id": [ 43 ], "fps_inferred_from_clip_interval": 29.913 }, "usage": { "input_tokens": 24656, "input_tokens_details": { "cache_write_tokens": 24653, "cached_tokens": 0 }, "output_tokens": 5474, "output_tokens_details": { "reasoning_tokens": 3317 }, "total_tokens": 30130 }, "model": "gpt-5.6-sol", "effort": "medium", "custom_id": "cat02_cap01_v_-DaX_1bBNAo_1_345", "annotation": { "caption": "A person with long, wavy blonde hair, wearing a white lace bra and large white angel wings, smiles while looking to their left. They raise a small blue party blower to their mouth and hold it there for a moment. After lowering it, they raise the blower again towards their open mouth before bringing their arm back down.", "entity_groups": [ { "id": "person_1", "label": "person", "spans": [ { "text": "A person", "range": [ 0, 8 ] }, { "text": "their", "range": [ 116, 121 ] }, { "text": "They", "range": [ 128, 132 ] }, { "text": "their", "range": [ 168, 173 ] }, { "text": "they", "range": [ 231, 235 ] }, { "text": "their", "range": [ 267, 272 ] }, { "text": "their", "range": [ 300, 305 ] } ], "source_mask_link": "selected_person" }, { "id": "hair_1", "label": "hair", "spans": [ { "text": "long, wavy blonde hair", "range": [ 14, 36 ] } ], "source_mask_link": null }, { "id": "bra_1", "label": "bra", "spans": [ { "text": "a white lace bra", "range": [ 46, 62 ] } ], "source_mask_link": null }, { "id": "wings_1", "label": "angel wings", "spans": [ { "text": "large white angel wings", "range": [ 67, 90 ] } ], "source_mask_link": null }, { "id": "blower_1", "label": "party blower", "spans": [ { "text": "a small blue party blower", "range": [ 139, 164 ] }, { "text": "it", "range": [ 189, 191 ] }, { "text": "it", "range": [ 227, 229 ] }, { "text": "the blower", "range": [ 242, 252 ] } ], "source_mask_link": null }, { "id": "mouth_1", "label": "mouth", "spans": [ { "text": "their mouth", "range": [ 168, 179 ] }, { "text": "their open mouth", "range": [ 267, 283 ] } ], "source_mask_link": null }, { "id": "arm_1", "label": "arm", "spans": [ { "text": "their arm", "range": [ 300, 309 ] } ], "source_mask_link": null } ], "actions": [ { "id": "action_1", "mentions": [ { "text": "smiles", "ranges": [ [ 92, 98 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "type": "simple", "subject_refs": [ "person_1" ], "object_refs": [], "confidence": "high", "temporal_status": "supported", "intervals": [ { "start_s": 0.0, "end_s": 3.0, "source_start_s": 0.0, "source_end_s": 3.0 } ], "evidence": "The masked person visibly smiles in the opening frames, especially at 0 and 2 seconds.", "mask_relation": "The action subject matches the supplied selected-person mask.", "source_subject_refs": [ "person_1" ] }, { "id": "action_2", "mentions": [ { "text": "looking", "ranges": [ [ 105, 112 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "type": "prepositional", "subject_refs": [ "person_1" ], "object_refs": [], "confidence": "high", "temporal_status": "supported", "intervals": [ { "start_s": 0.0, "end_s": 2.0, "source_start_s": 0.0, "source_end_s": 2.0 } ], "evidence": "At 0 and 1 seconds, the masked person's head and gaze are turned toward her left.", "mask_relation": "The action subject matches the supplied selected-person mask.", "source_subject_refs": [ "person_1" ] }, { "id": "action_3", "mentions": [ { "text": "raise", "ranges": [ [ 133, 138 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "type": "simple", "subject_refs": [ "person_1" ], "object_refs": [ "blower_1" ], "confidence": "high", "temporal_status": "supported", "intervals": [ { "start_s": 2.0, "end_s": 4.0, "source_start_s": 2.0, "source_end_s": 4.0 } ], "evidence": "The masked person's hand moves from a lowered position at 2 seconds to a raised position beside her face at 3 seconds.", "mask_relation": "The action subject matches the supplied selected-person mask.", "source_subject_refs": [ "person_1" ] }, { "id": "action_4", "mentions": [ { "text": "hold", "ranges": [ [ 184, 188 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "type": "simple", "subject_refs": [ "person_1" ], "object_refs": [ "blower_1" ], "confidence": "high", "temporal_status": "supported", "intervals": [ { "start_s": 3.0, "end_s": 6.0, "source_start_s": 3.0, "source_end_s": 6.0 } ], "evidence": "From 3 through 5 seconds, the masked person keeps the small blower raised beside her mouth.", "mask_relation": "The action subject matches the supplied selected-person mask.", "source_subject_refs": [ "person_1" ] }, { "id": "action_5", "mentions": [ { "text": "lowering", "ranges": [ [ 218, 226 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "type": "simple", "subject_refs": [ "person_1" ], "object_refs": [ "blower_1" ], "confidence": "high", "temporal_status": "supported", "intervals": [ { "start_s": 5.0, "end_s": 7.0, "source_start_s": 5.0, "source_end_s": 7.0 } ], "evidence": "The masked person's hand is raised at 5 seconds and lowered by 6 seconds.", "mask_relation": "The action subject matches the supplied selected-person mask.", "source_subject_refs": [ "person_1" ] }, { "id": "action_6", "mentions": [ { "text": "raise", "ranges": [ [ 236, 241 ] ], "occurrence": 2, "particle": { "text": null, "occurrence": 1 } } ], "type": "simple", "subject_refs": [ "person_1" ], "object_refs": [ "blower_1" ], "confidence": "high", "temporal_status": "supported", "intervals": [ { "start_s": 6.0, "end_s": 8.0, "source_start_s": 6.0, "source_end_s": 8.0 } ], "evidence": "After having her hand down at 6 seconds, the masked person raises the blower beside her open mouth at 7 seconds.", "mask_relation": "The action subject matches the supplied selected-person mask.", "source_subject_refs": [ "person_1" ] }, { "id": "action_7", "mentions": [ { "text": "bringing", "ranges": [ [ 291, 299 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "type": "simple", "subject_refs": [ "person_1" ], "object_refs": [ "arm_1" ], "confidence": "high", "temporal_status": "supported", "intervals": [ { "start_s": 9.0, "end_s": 11.0, "source_start_s": 9.0, "source_end_s": 11.0 } ], "evidence": "The masked person's arm is still raised at 9 seconds and is down by 10 seconds; it remains down at 11 seconds.", "mask_relation": "The action subject matches the supplied selected-person mask.", "source_subject_refs": [ "person_1" ] } ], "notes": [ "The supplied mask consistently identifies the single person throughout the clip.", "No independent masks were supplied for the hair, bra, wings, party blower, mouth, or arm, so those entity groups have null source-mask links.", "The participle \"wearing\" describes a static clothing configuration rather than an eligible action event and was not annotated.", "In \"bringing their arm back down,\" \"back\" and \"down\" are directional modifiers and are excluded from the action span." ] }, "model_output_raw": { "caption": "A person with long, wavy blonde hair, wearing a white lace bra and large white angel wings, smiles while looking to their left. They raise a small blue party blower to their mouth and hold it there for a moment. After lowering it, they raise the blower again towards their open mouth before bringing their arm back down.", "entity_groups": [ { "id": "person_1", "label": "person", "spans": [ { "text": "A person", "range": [ 0, 8 ] }, { "text": "their", "range": [ 116, 121 ] }, { "text": "They", "range": [ 128, 132 ] }, { "text": "their", "range": [ 168, 173 ] }, { "text": "they", "range": [ 231, 235 ] }, { "text": "their", "range": [ 267, 272 ] }, { "text": "their", "range": [ 300, 305 ] } ], "source_mask_link": "selected_person" }, { "id": "hair_1", "label": "hair", "spans": [ { "text": "long, wavy blonde hair", "range": [ 14, 36 ] } ], "source_mask_link": null }, { "id": "bra_1", "label": "bra", "spans": [ { "text": "a white lace bra", "range": [ 46, 62 ] } ], "source_mask_link": null }, { "id": "wings_1", "label": "angel wings", "spans": [ { "text": "large white angel wings", "range": [ 67, 90 ] } ], "source_mask_link": null }, { "id": "blower_1", "label": "party blower", "spans": [ { "text": "a small blue party blower", "range": [ 139, 164 ] }, { "text": "it", "range": [ 189, 191 ] }, { "text": "it", "range": [ 227, 229 ] }, { "text": "the blower", "range": [ 242, 252 ] } ], "source_mask_link": null }, { "id": "mouth_1", "label": "mouth", "spans": [ { "text": "their mouth", "range": [ 168, 179 ] }, { "text": "their open mouth", "range": [ 267, 283 ] } ], "source_mask_link": null }, { "id": "arm_1", "label": "arm", "spans": [ { "text": "their arm", "range": [ 300, 309 ] } ], "source_mask_link": null } ], "actions": [ { "id": "action_1", "mentions": [ { "text": "smiles", "ranges": [ [ 92, 98 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "type": "simple", "subject_refs": [ "person_1" ], "object_refs": [], "confidence": "high", "temporal_status": "supported", "intervals": [ { "start_s": 0.0, "end_s": 3.0 } ], "evidence": "The masked person visibly smiles in the opening frames, especially at 0 and 2 seconds.", "mask_relation": "The action subject matches the supplied selected-person mask." }, { "id": "action_2", "mentions": [ { "text": "looking", "ranges": [ [ 105, 112 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "type": "prepositional", "subject_refs": [ "person_1" ], "object_refs": [], "confidence": "high", "temporal_status": "supported", "intervals": [ { "start_s": 0.0, "end_s": 2.0 } ], "evidence": "At 0 and 1 seconds, the masked person's head and gaze are turned toward her left.", "mask_relation": "The action subject matches the supplied selected-person mask." }, { "id": "action_3", "mentions": [ { "text": "raise", "ranges": [ [ 133, 138 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "type": "simple", "subject_refs": [ "person_1" ], "object_refs": [ "blower_1" ], "confidence": "high", "temporal_status": "supported", "intervals": [ { "start_s": 2.0, "end_s": 4.0 } ], "evidence": "The masked person's hand moves from a lowered position at 2 seconds to a raised position beside her face at 3 seconds.", "mask_relation": "The action subject matches the supplied selected-person mask." }, { "id": "action_4", "mentions": [ { "text": "hold", "ranges": [ [ 184, 188 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "type": "simple", "subject_refs": [ "person_1" ], "object_refs": [ "blower_1" ], "confidence": "high", "temporal_status": "supported", "intervals": [ { "start_s": 3.0, "end_s": 6.0 } ], "evidence": "From 3 through 5 seconds, the masked person keeps the small blower raised beside her mouth.", "mask_relation": "The action subject matches the supplied selected-person mask." }, { "id": "action_5", "mentions": [ { "text": "lowering", "ranges": [ [ 218, 226 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "type": "simple", "subject_refs": [ "person_1" ], "object_refs": [ "blower_1" ], "confidence": "high", "temporal_status": "supported", "intervals": [ { "start_s": 5.0, "end_s": 7.0 } ], "evidence": "The masked person's hand is raised at 5 seconds and lowered by 6 seconds.", "mask_relation": "The action subject matches the supplied selected-person mask." }, { "id": "action_6", "mentions": [ { "text": "raise", "ranges": [ [ 236, 241 ] ], "occurrence": 2, "particle": { "text": null, "occurrence": 1 } } ], "type": "simple", "subject_refs": [ "person_1" ], "object_refs": [ "blower_1" ], "confidence": "high", "temporal_status": "supported", "intervals": [ { "start_s": 6.0, "end_s": 8.0 } ], "evidence": "After having her hand down at 6 seconds, the masked person raises the blower beside her open mouth at 7 seconds.", "mask_relation": "The action subject matches the supplied selected-person mask." }, { "id": "action_7", "mentions": [ { "text": "bringing", "ranges": [ [ 291, 299 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "type": "simple", "subject_refs": [ "person_1" ], "object_refs": [ "arm_1" ], "confidence": "high", "temporal_status": "supported", "intervals": [ { "start_s": 9.0, "end_s": 11.0 } ], "evidence": "The masked person's arm is still raised at 9 seconds and is down by 10 seconds; it remains down at 11 seconds.", "mask_relation": "The action subject matches the supplied selected-person mask." } ], "notes": [ "The supplied mask consistently identifies the single person throughout the clip.", "No independent masks were supplied for the hair, bra, wings, party blower, mouth, or arm, so those entity groups have null source-mask links.", "The participle \"wearing\" describes a static clothing configuration rather than an eligible action event and was not annotated.", "In \"bringing their arm back down,\" \"back\" and \"down\" are directional modifiers and are excluded from the action span." ] }, "normalization_repairs": 0 }, { "source_video_id": "v_-DaX_1bBNAo", "caption_count_category": 2, "clip": { "caption_index": 2, "caption": "The person smiles while looking more forward, briefly bringing the hand holding the blower near their face. They then pose by placing both hands on their hips. Finally, the person leans forward and moves out of the frame to the left.", "source_interval_sec": [ 11.5, 19.69 ], "loom_clip_id": "v_-DaX_1bBNAo_345_590", "clip_frame_interval_4fps": [ 345, 590 ], "sample_count_1fps": 9, "sampled_frames": [ { "frame": "00000", "timestamp_sec": 11.5, "offset_sec": 0, "rgb_path": "assets/cat_02/v_-DaX_1bBNAo_345_590/0000s_00000_rgb.jpg", "mask_path": "assets/cat_02/v_-DaX_1bBNAo_345_590/0000s_00000_mask.png", "mask_status": "available" }, { "frame": "00028", "timestamp_sec": 12.5, "offset_sec": 1, "rgb_path": "assets/cat_02/v_-DaX_1bBNAo_345_590/0001s_00028_rgb.jpg", "mask_path": "assets/cat_02/v_-DaX_1bBNAo_345_590/0001s_00028_mask.png", "mask_status": "available" }, { "frame": "00060", "timestamp_sec": 13.5, "offset_sec": 2, "rgb_path": "assets/cat_02/v_-DaX_1bBNAo_345_590/0002s_00060_rgb.jpg", "mask_path": "assets/cat_02/v_-DaX_1bBNAo_345_590/0002s_00060_mask.png", "mask_status": "available" }, { "frame": "00088", "timestamp_sec": 14.5, "offset_sec": 3, "rgb_path": "assets/cat_02/v_-DaX_1bBNAo_345_590/0003s_00088_rgb.jpg", "mask_path": "assets/cat_02/v_-DaX_1bBNAo_345_590/0003s_00088_mask.png", "mask_status": "available" }, { "frame": "00120", "timestamp_sec": 15.5, "offset_sec": 4, "rgb_path": "assets/cat_02/v_-DaX_1bBNAo_345_590/0004s_00120_rgb.jpg", "mask_path": "assets/cat_02/v_-DaX_1bBNAo_345_590/0004s_00120_mask.png", "mask_status": "available" }, { "frame": "00148", "timestamp_sec": 16.5, "offset_sec": 5, "rgb_path": "assets/cat_02/v_-DaX_1bBNAo_345_590/0005s_00148_rgb.jpg", "mask_path": "assets/cat_02/v_-DaX_1bBNAo_345_590/0005s_00148_mask.png", "mask_status": "available" }, { "frame": "00180", "timestamp_sec": 17.5, "offset_sec": 6, "rgb_path": "assets/cat_02/v_-DaX_1bBNAo_345_590/0006s_00180_rgb.jpg", "mask_path": "assets/cat_02/v_-DaX_1bBNAo_345_590/0006s_00180_mask.png", "mask_status": "available" }, { "frame": "00208", "timestamp_sec": 18.5, "offset_sec": 7, "rgb_path": "assets/cat_02/v_-DaX_1bBNAo_345_590/0007s_00208_rgb.jpg", "mask_path": "assets/cat_02/v_-DaX_1bBNAo_345_590/0007s_00208_mask.png", "mask_status": "available" }, { "frame": "00240", "timestamp_sec": 19.5, "offset_sec": 8, "rgb_path": "assets/cat_02/v_-DaX_1bBNAo_345_590/0008s_00240_rgb.jpg", "mask_path": "assets/cat_02/v_-DaX_1bBNAo_345_590/0008s_00240_mask.png", "mask_status": "available" } ], "obj_id": [ 0 ], "anno_id": [ 44 ], "fps_inferred_from_clip_interval": 29.9145 }, "usage": { "input_tokens": 22312, "input_tokens_details": { "cache_write_tokens": 22309, "cached_tokens": 0 }, "output_tokens": 4998, "output_tokens_details": { "reasoning_tokens": 3106 }, "total_tokens": 27310 }, "model": "gpt-5.6-sol", "effort": "medium", "custom_id": "cat02_cap02_v_-DaX_1bBNAo_345_590", "annotation": { "caption": "The person smiles while looking more forward, briefly bringing the hand holding the blower near their face. They then pose by placing both hands on their hips. Finally, the person leans forward and moves out of the frame to the left.", "entity_groups": [ { "id": "entity_1", "label": "person", "spans": [ { "text": "The person", "range": [ 0, 10 ] }, { "text": "their", "range": [ 96, 101 ] }, { "text": "They", "range": [ 108, 112 ] }, { "text": "their", "range": [ 148, 153 ] }, { "text": "the person", "range": [ 169, 179 ] } ], "source_mask_link": "selected_person" }, { "id": "entity_2", "label": "hand", "spans": [ { "text": "the hand", "range": [ 63, 71 ] } ], "source_mask_link": null }, { "id": "entity_3", "label": "blower", "spans": [ { "text": "the blower", "range": [ 80, 90 ] } ], "source_mask_link": null }, { "id": "entity_4", "label": "face", "spans": [ { "text": "their face", "range": [ 96, 106 ] } ], "source_mask_link": null }, { "id": "entity_5", "label": "hands", "spans": [ { "text": "both hands", "range": [ 134, 144 ] } ], "source_mask_link": null }, { "id": "entity_6", "label": "hips", "spans": [ { "text": "their hips", "range": [ 148, 158 ] } ], "source_mask_link": null } ], "actions": [ { "id": "action_1", "mentions": [ { "text": "smiles", "ranges": [ [ 11, 17 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "type": "simple", "subject_refs": [ "entity_1" ], "object_refs": [], "confidence": "high", "temporal_status": "localized", "intervals": [ { "start_s": 0.0, "end_s": 8.190000000000001, "source_start_s": 11.5, "source_end_s": 19.69 } ], "evidence": "The masked person is visibly smiling throughout the sampled sequence, including the final sample.", "mask_relation": "subject matches the supplied selected-person mask", "source_subject_refs": "entity_1" }, { "id": "action_2", "mentions": [ { "text": "looking", "ranges": [ [ 24, 31 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "type": "simple", "subject_refs": [ "entity_1" ], "object_refs": [], "confidence": "medium", "temporal_status": "localized", "intervals": [ { "start_s": 0.0, "end_s": 2.0, "source_start_s": 11.5, "source_end_s": 13.5 } ], "evidence": "At 0.0 and 1.0 seconds, the masked person's head and gaze are directed generally forward.", "mask_relation": "subject matches the supplied selected-person mask", "source_subject_refs": "entity_1" }, { "id": "action_3", "mentions": [ { "text": "bringing", "ranges": [ [ 54, 62 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "type": "simple", "subject_refs": [ "entity_1" ], "object_refs": [ "entity_2" ], "confidence": "high", "temporal_status": "localized", "intervals": [ { "start_s": 3.0, "end_s": 5.0, "source_start_s": 14.5, "source_end_s": 16.5 } ], "evidence": "Between the 3.0- and 4.0-second samples, the masked person raises a hand and brings it close to her face.", "mask_relation": "subject matches the supplied selected-person mask", "source_subject_refs": "entity_1" }, { "id": "action_4", "mentions": [ { "text": "pose", "ranges": [ [ 118, 122 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "type": "posture", "subject_refs": [ "entity_1" ], "object_refs": [], "confidence": "high", "temporal_status": "localized", "intervals": [ { "start_s": 7.0, "end_s": 8.0, "source_start_s": 18.5, "source_end_s": 19.5 } ], "evidence": "At 7.0 seconds, the masked person visibly holds a hands-on-hips pose.", "mask_relation": "subject matches the supplied selected-person mask", "source_subject_refs": "entity_1" }, { "id": "action_5", "mentions": [ { "text": "placing", "ranges": [ [ 126, 133 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "type": "simple", "subject_refs": [ "entity_1" ], "object_refs": [ "entity_5" ], "confidence": "high", "temporal_status": "localized", "intervals": [ { "start_s": 6.0, "end_s": 8.0, "source_start_s": 17.5, "source_end_s": 19.5 } ], "evidence": "The masked person's arms move down around 6.0 seconds, and her hands are positioned on her hips by 7.0 seconds.", "mask_relation": "subject matches the supplied selected-person mask", "source_subject_refs": "entity_1" }, { "id": "action_6", "mentions": [ { "text": "leans", "ranges": [ [ 180, 185 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "type": "posture", "subject_refs": [ "entity_1" ], "object_refs": [], "confidence": "medium", "temporal_status": "localized", "intervals": [ { "start_s": 8.0, "end_s": 8.190000000000001, "source_start_s": 19.5, "source_end_s": 19.69 } ], "evidence": "In the final sample, the masked person is partially cropped at the left edge while her body is displaced from the preceding upright pose, providing sparse support for the lean.", "mask_relation": "subject matches the supplied selected-person mask", "source_subject_refs": "entity_1" }, { "id": "action_7", "mentions": [ { "text": "moves", "ranges": [ [ 198, 203 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "type": "simple", "subject_refs": [ "entity_1" ], "object_refs": [], "confidence": "high", "temporal_status": "localized", "intervals": [ { "start_s": 8.0, "end_s": 8.190000000000001, "source_start_s": 19.5, "source_end_s": 19.69 } ], "evidence": "At 8.0 seconds, only part of the masked person remains at the left edge, supporting movement out of the frame.", "mask_relation": "subject matches the supplied selected-person mask", "source_subject_refs": "entity_1" } ], "notes": [ "The supplied mask consistently identifies the person in entity_1; no separate masks were supplied for the hand, blower, face, hands, or hips.", "The predicate “holding” was omitted because its stated subject is “the hand,” which has no corresponding supplied source mask.", "“the frame” and “the left” were excluded as image-medium or spatial references rather than concrete entity instances.", "Temporal intervals are sparse support intervals derived from the sampled clean RGB frames; final supported intervals end at the clip boundary of 8.19 seconds." ] }, "model_output_raw": { "caption": "The person smiles while looking more forward, briefly bringing the hand holding the blower near their face. They then pose by placing both hands on their hips. Finally, the person leans forward and moves out of the frame to the left.", "entity_groups": [ { "id": "entity_1", "label": "person", "spans": [ { "text": "The person", "range": [ 0, 10 ] }, { "text": "their", "range": [ 96, 101 ] }, { "text": "They", "range": [ 108, 112 ] }, { "text": "their", "range": [ 148, 153 ] }, { "text": "the person", "range": [ 169, 179 ] } ], "source_mask": "selected_person_mask" }, { "id": "entity_2", "label": "hand", "spans": [ { "text": "the hand", "range": [ 63, 71 ] } ], "source_mask": null }, { "id": "entity_3", "label": "blower", "spans": [ { "text": "the blower", "range": [ 80, 90 ] } ], "source_mask": null }, { "id": "entity_4", "label": "face", "spans": [ { "text": "their face", "range": [ 96, 106 ] } ], "source_mask": null }, { "id": "entity_5", "label": "hands", "spans": [ { "text": "both hands", "range": [ 134, 144 ] } ], "source_mask": null }, { "id": "entity_6", "label": "hips", "spans": [ { "text": "their hips", "range": [ 148, 158 ] } ], "source_mask": null } ], "actions": [ { "mentions": [ { "text": "smiles", "ranges": [ [ 11, 17 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "type": "simple", "subject_entity_group": "entity_1", "object_entity_groups": [], "confidence": "high", "temporal_status": "localized", "intervals": [ { "start_s": 0.0, "end_s": 8.19 } ], "evidence": "The masked person is visibly smiling throughout the sampled sequence, including the final sample.", "mask_relation": "subject matches the supplied selected-person mask" }, { "mentions": [ { "text": "looking", "ranges": [ [ 24, 31 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "type": "simple", "subject_entity_group": "entity_1", "object_entity_groups": [], "confidence": "medium", "temporal_status": "localized", "intervals": [ { "start_s": 0.0, "end_s": 2.0 } ], "evidence": "At 0.0 and 1.0 seconds, the masked person's head and gaze are directed generally forward.", "mask_relation": "subject matches the supplied selected-person mask" }, { "mentions": [ { "text": "bringing", "ranges": [ [ 54, 62 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "type": "simple", "subject_entity_group": "entity_1", "object_entity_groups": [ "entity_2" ], "confidence": "high", "temporal_status": "localized", "intervals": [ { "start_s": 3.0, "end_s": 5.0 } ], "evidence": "Between the 3.0- and 4.0-second samples, the masked person raises a hand and brings it close to her face.", "mask_relation": "subject matches the supplied selected-person mask" }, { "mentions": [ { "text": "pose", "ranges": [ [ 118, 122 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "type": "posture", "subject_entity_group": "entity_1", "object_entity_groups": [], "confidence": "high", "temporal_status": "localized", "intervals": [ { "start_s": 7.0, "end_s": 8.0 } ], "evidence": "At 7.0 seconds, the masked person visibly holds a hands-on-hips pose.", "mask_relation": "subject matches the supplied selected-person mask" }, { "mentions": [ { "text": "placing", "ranges": [ [ 126, 133 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "type": "simple", "subject_entity_group": "entity_1", "object_entity_groups": [ "entity_5" ], "confidence": "high", "temporal_status": "localized", "intervals": [ { "start_s": 6.0, "end_s": 8.0 } ], "evidence": "The masked person's arms move down around 6.0 seconds, and her hands are positioned on her hips by 7.0 seconds.", "mask_relation": "subject matches the supplied selected-person mask" }, { "mentions": [ { "text": "leans", "ranges": [ [ 180, 185 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "type": "posture", "subject_entity_group": "entity_1", "object_entity_groups": [], "confidence": "medium", "temporal_status": "localized", "intervals": [ { "start_s": 8.0, "end_s": 8.19 } ], "evidence": "In the final sample, the masked person is partially cropped at the left edge while her body is displaced from the preceding upright pose, providing sparse support for the lean.", "mask_relation": "subject matches the supplied selected-person mask" }, { "mentions": [ { "text": "moves", "ranges": [ [ 198, 203 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "type": "simple", "subject_entity_group": "entity_1", "object_entity_groups": [], "confidence": "high", "temporal_status": "localized", "intervals": [ { "start_s": 8.0, "end_s": 8.19 } ], "evidence": "At 8.0 seconds, only part of the masked person remains at the left edge, supporting movement out of the frame.", "mask_relation": "subject matches the supplied selected-person mask" } ], "notes": [ "The supplied mask consistently identifies the person in entity_1; no separate masks were supplied for the hand, blower, face, hands, or hips.", "The predicate “holding” was omitted because its stated subject is “the hand,” which has no corresponding supplied source mask.", "“the frame” and “the left” were excluded as image-medium or spatial references rather than concrete entity instances.", "Temporal intervals are sparse support intervals derived from the sampled clean RGB frames; final supported intervals end at the clip boundary of 8.19 seconds." ] }, "normalization_repairs": 0 }, { "source_video_id": "v_-IkL9Kq9zfQ", "caption_count_category": 3, "clip": { "caption_index": 1, "caption": "A person wearing a red beanie, blue jacket, and black life vest sits in a red canoe, paddling on their right side. They use a black paddle, pulling it through the water to propel the canoe forward. The person completes one full stroke and begins another, maintaining a consistent rhythm as they navigate the flowing water.", "source_interval_sec": [ 0.0, 7.28 ], "loom_clip_id": "v_-IkL9Kq9zfQ_1_182", "clip_frame_interval_4fps": [ 1, 182 ], "sample_count_1fps": 8, "sampled_frames": [ { "frame": "00000", "timestamp_sec": 0.0, "offset_sec": 0, "rgb_path": "assets/cat_03/v_-IkL9Kq9zfQ_1_182/0000s_00000_rgb.jpg", "mask_path": "assets/cat_03/v_-IkL9Kq9zfQ_1_182/0000s_00000_mask.png", "mask_status": "available" }, { "frame": "00024", "timestamp_sec": 1.0, "offset_sec": 1, "rgb_path": "assets/cat_03/v_-IkL9Kq9zfQ_1_182/0001s_00024_rgb.jpg", "mask_path": "assets/cat_03/v_-IkL9Kq9zfQ_1_182/0001s_00024_mask.png", "mask_status": "available" }, { "frame": "00048", "timestamp_sec": 2.0, "offset_sec": 2, "rgb_path": "assets/cat_03/v_-IkL9Kq9zfQ_1_182/0002s_00048_rgb.jpg", "mask_path": "assets/cat_03/v_-IkL9Kq9zfQ_1_182/0002s_00048_mask.png", "mask_status": "available" }, { "frame": "00076", "timestamp_sec": 3.0, "offset_sec": 3, "rgb_path": "assets/cat_03/v_-IkL9Kq9zfQ_1_182/0003s_00076_rgb.jpg", "mask_path": "assets/cat_03/v_-IkL9Kq9zfQ_1_182/0003s_00076_mask.png", "mask_status": "available" }, { "frame": "00100", "timestamp_sec": 4.0, "offset_sec": 4, "rgb_path": "assets/cat_03/v_-IkL9Kq9zfQ_1_182/0004s_00100_rgb.jpg", "mask_path": "assets/cat_03/v_-IkL9Kq9zfQ_1_182/0004s_00100_mask.png", "mask_status": "available" }, { "frame": "00124", "timestamp_sec": 5.0, "offset_sec": 5, "rgb_path": "assets/cat_03/v_-IkL9Kq9zfQ_1_182/0005s_00124_rgb.jpg", "mask_path": "assets/cat_03/v_-IkL9Kq9zfQ_1_182/0005s_00124_mask.png", "mask_status": "available" }, { "frame": "00148", "timestamp_sec": 6.0, "offset_sec": 6, "rgb_path": "assets/cat_03/v_-IkL9Kq9zfQ_1_182/0006s_00148_rgb.jpg", "mask_path": "assets/cat_03/v_-IkL9Kq9zfQ_1_182/0006s_00148_mask.png", "mask_status": "available" }, { "frame": "00176", "timestamp_sec": 7.0, "offset_sec": 7, "rgb_path": "assets/cat_03/v_-IkL9Kq9zfQ_1_182/0007s_00176_rgb.jpg", "mask_path": "assets/cat_03/v_-IkL9Kq9zfQ_1_182/0007s_00176_mask.png", "mask_status": "available" } ], "obj_id": [ 0 ], "anno_id": [ 53 ], "fps_inferred_from_clip_interval": 24.8626 }, "usage": { "input_tokens": 21556, "input_tokens_details": { "cache_write_tokens": 21553, "cached_tokens": 0 }, "output_tokens": 7292, "output_tokens_details": { "reasoning_tokens": 4780 }, "total_tokens": 28848 }, "model": "gpt-5.6-sol", "effort": "medium", "custom_id": "cat03_cap01_v_-IkL9Kq9zfQ_1_182", "annotation": { "caption": "A person wearing a red beanie, blue jacket, and black life vest sits in a red canoe, paddling on their right side. They use a black paddle, pulling it through the water to propel the canoe forward. The person completes one full stroke and begins another, maintaining a consistent rhythm as they navigate the flowing water.", "entity_groups": [ { "id": "E1", "label": "person", "spans": [ { "text": "A person", "range": [ 0, 8 ] }, { "text": "their", "range": [ 97, 102 ] }, { "text": "They", "range": [ 115, 119 ] }, { "text": "The person", "range": [ 198, 208 ] }, { "text": "they", "range": [ 290, 294 ] } ], "source_mask_link": "selected_person" }, { "id": "E2", "label": "beanie", "spans": [ { "text": "a red beanie", "range": [ 17, 29 ] } ], "source_mask_link": null }, { "id": "E3", "label": "jacket", "spans": [ { "text": "blue jacket", "range": [ 31, 42 ] } ], "source_mask_link": null }, { "id": "E4", "label": "life vest", "spans": [ { "text": "black life vest", "range": [ 48, 63 ] } ], "source_mask_link": null }, { "id": "E5", "label": "canoe", "spans": [ { "text": "a red canoe", "range": [ 72, 83 ] }, { "text": "the canoe", "range": [ 179, 188 ] } ], "source_mask_link": null }, { "id": "E6", "label": "paddle", "spans": [ { "text": "a black paddle", "range": [ 124, 138 ] }, { "text": "it", "range": [ 148, 150 ] } ], "source_mask_link": null }, { "id": "E7", "label": "water", "spans": [ { "text": "the water", "range": [ 159, 168 ] }, { "text": "the flowing water", "range": [ 304, 321 ] } ], "source_mask_link": null } ], "actions": [ { "id": "A1", "mentions": [ { "text": "wearing", "ranges": [ [ 9, 16 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "type": "simple", "subject_refs": [ "E1" ], "object_refs": [ "E2", "E3", "E4" ], "confidence": "high", "temporal_status": "supported", "intervals": [ { "start_s": 0.0, "end_s": 7.28, "source_start_s": 0.0, "source_end_s": 7.28 } ], "evidence": "The masked person visibly wears the red beanie, blue jacket, and black vest throughout the sampled frames, including the partial final view.", "mask_relation": "subject matches the supplied selected-person mask", "source_subject_refs": "E1" }, { "id": "A2", "mentions": [ { "text": "sits", "ranges": [ [ 64, 68 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "type": "posture", "subject_refs": [ "E1" ], "object_refs": [], "confidence": "high", "temporal_status": "supported", "intervals": [ { "start_s": 0.0, "end_s": 7.28, "source_start_s": 0.0, "source_end_s": 7.28 } ], "evidence": "The masked person remains seated in the canoe across the clip.", "mask_relation": "subject matches the supplied selected-person mask", "source_subject_refs": "E1" }, { "id": "A3", "mentions": [ { "text": "paddling", "ranges": [ [ 85, 93 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "type": "simple", "subject_refs": [ "E1" ], "object_refs": [], "confidence": "high", "temporal_status": "supported", "intervals": [ { "start_s": 0.0, "end_s": 7.0, "source_start_s": 0.0, "source_end_s": 7.0 } ], "evidence": "Frames from 0 through 6 seconds show the masked person moving the paddle through successive stroke positions; the paddle is no longer visible at 7 seconds.", "mask_relation": "subject matches the supplied selected-person mask", "source_subject_refs": "E1" }, { "id": "A4", "mentions": [ { "text": "use", "ranges": [ [ 120, 123 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "type": "simple", "subject_refs": [ "E1" ], "object_refs": [ "E6" ], "confidence": "high", "temporal_status": "supported", "intervals": [ { "start_s": 0.0, "end_s": 7.0, "source_start_s": 0.0, "source_end_s": 7.0 } ], "evidence": "The masked person visibly grips and operates the black paddle in samples from 0 through 6 seconds.", "mask_relation": "subject matches the supplied selected-person mask", "source_subject_refs": "E1" }, { "id": "A5", "mentions": [ { "text": "pulling", "ranges": [ [ 140, 147 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "type": "simple", "subject_refs": [ "E1" ], "object_refs": [ "E6" ], "confidence": "high", "temporal_status": "supported", "intervals": [ { "start_s": 0.0, "end_s": 3.0, "source_start_s": 0.0, "source_end_s": 3.0 }, { "start_s": 4.0, "end_s": 7.0, "source_start_s": 4.0, "source_end_s": 7.0 } ], "evidence": "The paddle progresses through rearward pulling phases in samples 0–2 and again in samples 4–6, with a transition or recovery around 3 seconds.", "mask_relation": "subject matches the supplied selected-person mask", "source_subject_refs": "E1" }, { "id": "A6", "mentions": [ { "text": "propel", "ranges": [ [ 172, 178 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "type": "simple", "subject_refs": [ "E1" ], "object_refs": [ "E5" ], "confidence": "high", "temporal_status": "supported", "intervals": [ { "start_s": 0.0, "end_s": 7.28, "source_start_s": 0.0, "source_end_s": 7.28 } ], "evidence": "The canoe travels through visibly disturbed water while the masked person paddles, and it remains in forward motion through the final sample.", "mask_relation": "subject matches the supplied selected-person mask", "source_subject_refs": "E1" }, { "id": "A7", "mentions": [ { "text": "completes one full stroke", "ranges": [ [ 209, 234 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "type": "light_verb", "subject_refs": [ "E1" ], "object_refs": [], "confidence": "medium", "temporal_status": "supported", "intervals": [ { "start_s": 0.0, "end_s": 4.0, "source_start_s": 0.0, "source_end_s": 4.0 } ], "evidence": "Samples 0–3 show a forward/high paddle position, a vertical planted position, a rearward position, and then a transition consistent with completing a stroke.", "mask_relation": "subject matches the supplied selected-person mask", "source_subject_refs": "E1" }, { "id": "A8", "mentions": [ { "text": "maintaining", "ranges": [ [ 255, 266 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "type": "simple", "subject_refs": [ "E1" ], "object_refs": [], "confidence": "medium", "temporal_status": "supported", "intervals": [ { "start_s": 0.0, "end_s": 7.0, "source_start_s": 0.0, "source_end_s": 7.0 } ], "evidence": "The ordered paddle positions across samples 0–6 show repeated, consistent stroke phases.", "mask_relation": "subject matches the supplied selected-person mask", "source_subject_refs": "E1" }, { "id": "A9", "mentions": [ { "text": "navigate", "ranges": [ [ 295, 303 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "type": "simple", "subject_refs": [ "E1" ], "object_refs": [ "E7" ], "confidence": "high", "temporal_status": "supported", "intervals": [ { "start_s": 0.0, "end_s": 7.28, "source_start_s": 0.0, "source_end_s": 7.28 } ], "evidence": "The masked person controls the canoe while it proceeds through flowing, turbulent water across the full clip.", "mask_relation": "subject matches the supplied selected-person mask", "source_subject_refs": "E1" } ], "notes": [ "All action intervals are half-open and use clip-relative seconds; source-video times are identical because SOURCE_START_S is 0.0.", "Only the person has a supplied mask track. The beanie, jacket, life vest, canoe, paddle, and water are text-groundable entities but have no separate supplied masks.", "The water predicate “flowing” was not included as an action because its subject, the water, has no corresponding supplied mask.", "“begins” was excluded as an aspectual predicate framing the elided second stroke; no eligible overt embedded action verb appears in that phrase.", "The spatial phrase “right side” was not treated as a separate entity." ] }, "model_output_raw": { "caption": "A person wearing a red beanie, blue jacket, and black life vest sits in a red canoe, paddling on their right side. They use a black paddle, pulling it through the water to propel the canoe forward. The person completes one full stroke and begins another, maintaining a consistent rhythm as they navigate the flowing water.", "entity_groups": [ { "id": "E1", "label": "person", "spans": [ { "text": "A person", "range": [ 0, 8 ] }, { "text": "their", "range": [ 97, 102 ] }, { "text": "They", "range": [ 115, 119 ] }, { "text": "The person", "range": [ 198, 208 ] }, { "text": "they", "range": [ 290, 294 ] } ], "source_mask": "selected_person_mask" }, { "id": "E2", "label": "beanie", "spans": [ { "text": "a red beanie", "range": [ 17, 29 ] } ], "source_mask": null }, { "id": "E3", "label": "jacket", "spans": [ { "text": "blue jacket", "range": [ 31, 42 ] } ], "source_mask": null }, { "id": "E4", "label": "life vest", "spans": [ { "text": "black life vest", "range": [ 48, 63 ] } ], "source_mask": null }, { "id": "E5", "label": "canoe", "spans": [ { "text": "a red canoe", "range": [ 72, 83 ] }, { "text": "the canoe", "range": [ 179, 188 ] } ], "source_mask": null }, { "id": "E6", "label": "paddle", "spans": [ { "text": "a black paddle", "range": [ 124, 138 ] }, { "text": "it", "range": [ 148, 150 ] } ], "source_mask": null }, { "id": "E7", "label": "water", "spans": [ { "text": "the water", "range": [ 159, 168 ] }, { "text": "the flowing water", "range": [ 304, 321 ] } ], "source_mask": null } ], "actions": [ { "id": "A1", "mentions": [ { "text": "wearing", "ranges": [ [ 9, 16 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "type": "simple", "subject_ref": "E1", "object_refs": [ "E2", "E3", "E4" ], "confidence": "high", "mask_relation": "subject matches the supplied selected-person mask", "temporal_status": "supported", "intervals": [ { "start_s": 0.0, "end_s": 7.28 } ], "evidence": "The masked person visibly wears the red beanie, blue jacket, and black vest throughout the sampled frames, including the partial final view." }, { "id": "A2", "mentions": [ { "text": "sits", "ranges": [ [ 64, 68 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "type": "posture", "subject_ref": "E1", "object_refs": [], "confidence": "high", "mask_relation": "subject matches the supplied selected-person mask", "temporal_status": "supported", "intervals": [ { "start_s": 0.0, "end_s": 7.28 } ], "evidence": "The masked person remains seated in the canoe across the clip." }, { "id": "A3", "mentions": [ { "text": "paddling", "ranges": [ [ 85, 93 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "type": "simple", "subject_ref": "E1", "object_refs": [], "confidence": "high", "mask_relation": "subject matches the supplied selected-person mask", "temporal_status": "supported", "intervals": [ { "start_s": 0.0, "end_s": 7.0 } ], "evidence": "Frames from 0 through 6 seconds show the masked person moving the paddle through successive stroke positions; the paddle is no longer visible at 7 seconds." }, { "id": "A4", "mentions": [ { "text": "use", "ranges": [ [ 120, 123 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "type": "simple", "subject_ref": "E1", "object_refs": [ "E6" ], "confidence": "high", "mask_relation": "subject matches the supplied selected-person mask", "temporal_status": "supported", "intervals": [ { "start_s": 0.0, "end_s": 7.0 } ], "evidence": "The masked person visibly grips and operates the black paddle in samples from 0 through 6 seconds." }, { "id": "A5", "mentions": [ { "text": "pulling", "ranges": [ [ 140, 147 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "type": "simple", "subject_ref": "E1", "object_refs": [ "E6" ], "confidence": "high", "mask_relation": "subject matches the supplied selected-person mask", "temporal_status": "supported", "intervals": [ { "start_s": 0.0, "end_s": 3.0 }, { "start_s": 4.0, "end_s": 7.0 } ], "evidence": "The paddle progresses through rearward pulling phases in samples 0–2 and again in samples 4–6, with a transition or recovery around 3 seconds." }, { "id": "A6", "mentions": [ { "text": "propel", "ranges": [ [ 172, 178 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "type": "simple", "subject_ref": "E1", "object_refs": [ "E5" ], "confidence": "high", "mask_relation": "subject matches the supplied selected-person mask", "temporal_status": "supported", "intervals": [ { "start_s": 0.0, "end_s": 7.28 } ], "evidence": "The canoe travels through visibly disturbed water while the masked person paddles, and it remains in forward motion through the final sample." }, { "id": "A7", "mentions": [ { "text": "completes one full stroke", "ranges": [ [ 209, 234 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "type": "light_verb", "subject_ref": "E1", "object_refs": [], "confidence": "medium", "mask_relation": "subject matches the supplied selected-person mask", "temporal_status": "supported", "intervals": [ { "start_s": 0.0, "end_s": 4.0 } ], "evidence": "Samples 0–3 show a forward/high paddle position, a vertical planted position, a rearward position, and then a transition consistent with completing a stroke." }, { "id": "A8", "mentions": [ { "text": "maintaining", "ranges": [ [ 255, 266 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "type": "simple", "subject_ref": "E1", "object_refs": [], "confidence": "medium", "mask_relation": "subject matches the supplied selected-person mask", "temporal_status": "supported", "intervals": [ { "start_s": 0.0, "end_s": 7.0 } ], "evidence": "The ordered paddle positions across samples 0–6 show repeated, consistent stroke phases." }, { "id": "A9", "mentions": [ { "text": "navigate", "ranges": [ [ 295, 303 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "type": "simple", "subject_ref": "E1", "object_refs": [ "E7" ], "confidence": "high", "mask_relation": "subject matches the supplied selected-person mask", "temporal_status": "supported", "intervals": [ { "start_s": 0.0, "end_s": 7.28 } ], "evidence": "The masked person controls the canoe while it proceeds through flowing, turbulent water across the full clip." } ], "notes": [ "All action intervals are half-open and use clip-relative seconds; source-video times are identical because SOURCE_START_S is 0.0.", "Only the person has a supplied mask track. The beanie, jacket, life vest, canoe, paddle, and water are text-groundable entities but have no separate supplied masks.", "The water predicate “flowing” was not included as an action because its subject, the water, has no corresponding supplied mask.", "“begins” was excluded as an aspectual predicate framing the elided second stroke; no eligible overt embedded action verb appears in that phrase.", "The spatial phrase “right side” was not treated as a separate entity." ] }, "normalization_repairs": 0 }, { "source_video_id": "v_-IkL9Kq9zfQ", "caption_count_category": 3, "clip": { "caption_index": 2, "caption": "A person wearing a blue jacket, a black life vest, and a red beanie is seen from behind, paddling a red canoe. They consistently use a black paddle on the canoe's left side to navigate through the rapids. Their upper body and arms move in a steady, repetitive motion with each stroke.", "source_interval_sec": [ 11.64, 24.14 ], "loom_clip_id": "v_-IkL9Kq9zfQ_291_604", "clip_frame_interval_4fps": [ 291, 604 ], "sample_count_1fps": 13, "sampled_frames": [ { "frame": "00000", "timestamp_sec": 11.64, "offset_sec": 0, "rgb_path": "assets/cat_03/v_-IkL9Kq9zfQ_291_604/0000s_00000_rgb.jpg", "mask_path": "assets/cat_03/v_-IkL9Kq9zfQ_291_604/0000s_00000_mask.png", "mask_status": "available" }, { "frame": "00024", "timestamp_sec": 12.64, "offset_sec": 1, "rgb_path": "assets/cat_03/v_-IkL9Kq9zfQ_291_604/0001s_00024_rgb.jpg", "mask_path": "assets/cat_03/v_-IkL9Kq9zfQ_291_604/0001s_00024_mask.png", "mask_status": "available" }, { "frame": "00052", "timestamp_sec": 13.64, "offset_sec": 2, "rgb_path": "assets/cat_03/v_-IkL9Kq9zfQ_291_604/0002s_00052_rgb.jpg", "mask_path": "assets/cat_03/v_-IkL9Kq9zfQ_291_604/0002s_00052_mask.png", "mask_status": "available" }, { "frame": "00076", "timestamp_sec": 14.64, "offset_sec": 3, "rgb_path": "assets/cat_03/v_-IkL9Kq9zfQ_291_604/0003s_00076_rgb.jpg", "mask_path": "assets/cat_03/v_-IkL9Kq9zfQ_291_604/0003s_00076_mask.png", "mask_status": "available" }, { "frame": "00100", "timestamp_sec": 15.64, "offset_sec": 4, "rgb_path": "assets/cat_03/v_-IkL9Kq9zfQ_291_604/0004s_00100_rgb.jpg", "mask_path": "assets/cat_03/v_-IkL9Kq9zfQ_291_604/0004s_00100_mask.png", "mask_status": "available" }, { "frame": "00124", "timestamp_sec": 16.64, "offset_sec": 5, "rgb_path": "assets/cat_03/v_-IkL9Kq9zfQ_291_604/0005s_00124_rgb.jpg", "mask_path": "assets/cat_03/v_-IkL9Kq9zfQ_291_604/0005s_00124_mask.png", "mask_status": "available" }, { "frame": "00152", "timestamp_sec": 17.64, "offset_sec": 6, "rgb_path": "assets/cat_03/v_-IkL9Kq9zfQ_291_604/0006s_00152_rgb.jpg", "mask_path": "assets/cat_03/v_-IkL9Kq9zfQ_291_604/0006s_00152_mask.png", "mask_status": "available" }, { "frame": "00176", "timestamp_sec": 18.64, "offset_sec": 7, "rgb_path": "assets/cat_03/v_-IkL9Kq9zfQ_291_604/0007s_00176_rgb.jpg", "mask_path": "assets/cat_03/v_-IkL9Kq9zfQ_291_604/0007s_00176_mask.png", "mask_status": "available" }, { "frame": "00200", "timestamp_sec": 19.64, "offset_sec": 8, "rgb_path": "assets/cat_03/v_-IkL9Kq9zfQ_291_604/0008s_00200_rgb.jpg", "mask_path": "assets/cat_03/v_-IkL9Kq9zfQ_291_604/0008s_00200_mask.png", "mask_status": "available" }, { "frame": "00224", "timestamp_sec": 20.64, "offset_sec": 9, "rgb_path": "assets/cat_03/v_-IkL9Kq9zfQ_291_604/0009s_00224_rgb.jpg", "mask_path": "assets/cat_03/v_-IkL9Kq9zfQ_291_604/0009s_00224_mask.png", "mask_status": "available" }, { "frame": "00252", "timestamp_sec": 21.64, "offset_sec": 10, "rgb_path": "assets/cat_03/v_-IkL9Kq9zfQ_291_604/0010s_00252_rgb.jpg", "mask_path": "assets/cat_03/v_-IkL9Kq9zfQ_291_604/0010s_00252_mask.png", "mask_status": "available" }, { "frame": "00276", "timestamp_sec": 22.64, "offset_sec": 11, "rgb_path": "assets/cat_03/v_-IkL9Kq9zfQ_291_604/0011s_00276_rgb.jpg", "mask_path": "assets/cat_03/v_-IkL9Kq9zfQ_291_604/0011s_00276_mask.png", "mask_status": "available" }, { "frame": "00300", "timestamp_sec": 23.64, "offset_sec": 12, "rgb_path": "assets/cat_03/v_-IkL9Kq9zfQ_291_604/0012s_00300_rgb.jpg", "mask_path": "assets/cat_03/v_-IkL9Kq9zfQ_291_604/0012s_00300_mask.png", "mask_status": "available" } ], "obj_id": [ 0 ], "anno_id": [ 54 ], "fps_inferred_from_clip_interval": 25.04 }, "usage": { "input_tokens": 25428, "input_tokens_details": { "cache_write_tokens": 25425, "cached_tokens": 0 }, "output_tokens": 5928, "output_tokens_details": { "reasoning_tokens": 3756 }, "total_tokens": 31356 }, "model": "gpt-5.6-sol", "effort": "medium", "custom_id": "cat03_cap02_v_-IkL9Kq9zfQ_291_604", "annotation": { "caption": "A person wearing a blue jacket, a black life vest, and a red beanie is seen from behind, paddling a red canoe. They consistently use a black paddle on the canoe's left side to navigate through the rapids. Their upper body and arms move in a steady, repetitive motion with each stroke.", "entity_groups": [ { "id": "person_1", "label": "person", "spans": [ { "text": "A person", "range": [ 0, 8 ] }, { "text": "They", "range": [ 111, 115 ] }, { "text": "Their", "range": [ 205, 210 ] } ], "source_mask_link": "selected_person" }, { "id": "jacket_1", "label": "jacket", "spans": [ { "text": "a blue jacket", "range": [ 17, 30 ] } ], "source_mask_link": null }, { "id": "life_vest_1", "label": "life vest", "spans": [ { "text": "a black life vest", "range": [ 32, 49 ] } ], "source_mask_link": null }, { "id": "beanie_1", "label": "beanie", "spans": [ { "text": "a red beanie", "range": [ 55, 67 ] } ], "source_mask_link": null }, { "id": "canoe_1", "label": "canoe", "spans": [ { "text": "a red canoe", "range": [ 98, 109 ] }, { "text": "the canoe's", "range": [ 151, 162 ] } ], "source_mask_link": null }, { "id": "paddle_1", "label": "paddle", "spans": [ { "text": "a black paddle", "range": [ 133, 147 ] } ], "source_mask_link": null }, { "id": "rapids_1", "label": "rapids", "spans": [ { "text": "the rapids", "range": [ 193, 203 ] } ], "source_mask_link": null }, { "id": "upper_body_1", "label": "upper body", "spans": [ { "text": "Their upper body", "range": [ 205, 221 ] } ], "source_mask_link": null }, { "id": "arms_1", "label": "arms", "spans": [ { "text": "arms", "range": [ 226, 230 ] } ], "source_mask_link": null } ], "actions": [ { "id": "action_1", "mentions": [ { "text": "wearing", "ranges": [ [ 9, 16 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "type": "simple", "subject_refs": [ "person_1" ], "object_refs": [ "jacket_1", "life_vest_1", "beanie_1" ], "confidence": "high", "temporal_status": "localized", "intervals": [ { "start_s": 1.0, "end_s": 12.5, "source_start_s": 12.64, "source_end_s": 24.14 } ], "evidence": "Samples 2–13 show the mask-linked person continuously wearing the blue jacket, black life vest, and red beanie.", "mask_relation": "subject_matches_selected_person_mask", "source_subject_refs": [ "person_1" ] }, { "id": "action_2", "mentions": [ { "text": "paddling", "ranges": [ [ 89, 97 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "type": "simple", "subject_refs": [ "person_1" ], "object_refs": [ "canoe_1" ], "confidence": "high", "temporal_status": "localized", "intervals": [ { "start_s": 1.0, "end_s": 12.5, "source_start_s": 12.64, "source_end_s": 24.14 } ], "evidence": "Across samples 2–13, the mask-linked canoeist repeatedly changes arm and paddle position while traveling through the white water.", "mask_relation": "subject_matches_selected_person_mask", "source_subject_refs": [ "person_1" ] }, { "id": "action_3", "mentions": [ { "text": "use", "ranges": [ [ 129, 132 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "type": "simple", "subject_refs": [ "person_1" ], "object_refs": [ "paddle_1" ], "confidence": "high", "temporal_status": "localized", "intervals": [ { "start_s": 1.0, "end_s": 12.5, "source_start_s": 12.64, "source_end_s": 24.14 } ], "evidence": "Samples 2–13 show the selected person handling the black paddle through successive stroke positions; the paddle is intermittently occluded during the cycle.", "mask_relation": "subject_matches_selected_person_mask", "source_subject_refs": [ "person_1" ] }, { "id": "action_4", "mentions": [ { "text": "navigate", "ranges": [ [ 176, 184 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "type": "simple", "subject_refs": [ "person_1" ], "object_refs": [], "confidence": "high", "temporal_status": "localized", "intervals": [ { "start_s": 1.0, "end_s": 12.5, "source_start_s": 12.64, "source_end_s": 24.14 } ], "evidence": "Samples 2–13 show the selected canoeist progressing through turbulent rapids while controlling the canoe with repeated strokes.", "mask_relation": "subject_matches_selected_person_mask", "source_subject_refs": [ "person_1" ] }, { "id": "action_5", "mentions": [ { "text": "move", "ranges": [ [ 231, 235 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "type": "simple", "subject_refs": [ "person_1" ], "object_refs": [], "confidence": "high", "temporal_status": "localized", "intervals": [ { "start_s": 1.0, "end_s": 12.5, "source_start_s": 12.64, "source_end_s": 24.14 } ], "evidence": "The selected person's torso and arms cycle through extended and retracted positions across samples 2–13.", "mask_relation": "subject_matches_selected_person_mask", "source_subject_refs": [ "person_1", "upper_body_1", "arms_1" ] } ], "notes": [ "The supplied mask at sample 1 (clip-relative 0.000 s) marks a different canoeist farther ahead, not the captioned red-beanie person; that sample is excluded from action support.", "From sample 2 onward, the supplied masks consistently identify the captioned person in the red beanie, blue jacket, and black life vest.", "Only the person has a supplied masklet. The clothing, canoe, paddle, rapids, upper body, and arms have no separate source masks, so their source-mask links are null.", "\"is seen\" is excluded because it asserts visibility rather than a physical action. \"motion\" and \"stroke\" are nominal event descriptions rather than qualifying light-verb predicates." ] }, "model_output_raw": { "caption": "A person wearing a blue jacket, a black life vest, and a red beanie is seen from behind, paddling a red canoe. They consistently use a black paddle on the canoe's left side to navigate through the rapids. Their upper body and arms move in a steady, repetitive motion with each stroke.", "entity_groups": [ { "id": "person_1", "entity_type": "person", "spans": [ { "text": "A person", "start": 0, "end": 8 }, { "text": "They", "start": 111, "end": 115 }, { "text": "Their", "start": 205, "end": 210 } ], "source_mask_link": "selected_person_masklet" }, { "id": "jacket_1", "entity_type": "jacket", "spans": [ { "text": "a blue jacket", "start": 17, "end": 30 } ], "source_mask_link": null }, { "id": "life_vest_1", "entity_type": "life vest", "spans": [ { "text": "a black life vest", "start": 32, "end": 49 } ], "source_mask_link": null }, { "id": "beanie_1", "entity_type": "beanie", "spans": [ { "text": "a red beanie", "start": 55, "end": 67 } ], "source_mask_link": null }, { "id": "canoe_1", "entity_type": "canoe", "spans": [ { "text": "a red canoe", "start": 98, "end": 109 }, { "text": "the canoe's", "start": 151, "end": 162 } ], "source_mask_link": null }, { "id": "paddle_1", "entity_type": "paddle", "spans": [ { "text": "a black paddle", "start": 133, "end": 147 } ], "source_mask_link": null }, { "id": "rapids_1", "entity_type": "rapids", "spans": [ { "text": "the rapids", "start": 193, "end": 203 } ], "source_mask_link": null }, { "id": "upper_body_1", "entity_type": "upper body", "spans": [ { "text": "Their upper body", "start": 205, "end": 221 } ], "source_mask_link": null }, { "id": "arms_1", "entity_type": "arms", "spans": [ { "text": "arms", "start": 226, "end": 230 } ], "source_mask_link": null } ], "actions": [ { "id": "action_1", "mentions": [ { "text": "wearing", "occurrence": 1, "ranges": [ { "start": 9, "end": 16 } ], "particle": { "text": null, "occurrence": 1 } } ], "type": "simple", "subject_refs": [ "person_1" ], "object_refs": [ "jacket_1", "life_vest_1", "beanie_1" ], "confidence": "high", "mask_relation": "subject_matches_selected_person_mask", "temporal_status": "localized", "intervals": [ { "clip_relative": { "start_s": 1.0, "end_s": 12.5 }, "source_video": { "start_s": 12.64, "end_s": 24.14 } } ], "evidence": "Samples 2–13 show the mask-linked person continuously wearing the blue jacket, black life vest, and red beanie." }, { "id": "action_2", "mentions": [ { "text": "paddling", "occurrence": 1, "ranges": [ { "start": 89, "end": 97 } ], "particle": { "text": null, "occurrence": 1 } } ], "type": "simple", "subject_refs": [ "person_1" ], "object_refs": [ "canoe_1" ], "confidence": "high", "mask_relation": "subject_matches_selected_person_mask", "temporal_status": "localized", "intervals": [ { "clip_relative": { "start_s": 1.0, "end_s": 12.5 }, "source_video": { "start_s": 12.64, "end_s": 24.14 } } ], "evidence": "Across samples 2–13, the mask-linked canoeist repeatedly changes arm and paddle position while traveling through the white water." }, { "id": "action_3", "mentions": [ { "text": "use", "occurrence": 1, "ranges": [ { "start": 129, "end": 132 } ], "particle": { "text": null, "occurrence": 1 } } ], "type": "simple", "subject_refs": [ "person_1" ], "object_refs": [ "paddle_1" ], "confidence": "high", "mask_relation": "subject_matches_selected_person_mask", "temporal_status": "localized", "intervals": [ { "clip_relative": { "start_s": 1.0, "end_s": 12.5 }, "source_video": { "start_s": 12.64, "end_s": 24.14 } } ], "evidence": "Samples 2–13 show the selected person handling the black paddle through successive stroke positions; the paddle is intermittently occluded during the cycle." }, { "id": "action_4", "mentions": [ { "text": "navigate", "occurrence": 1, "ranges": [ { "start": 176, "end": 184 } ], "particle": { "text": null, "occurrence": 1 } } ], "type": "simple", "subject_refs": [ "person_1" ], "object_refs": [], "confidence": "high", "mask_relation": "subject_matches_selected_person_mask", "temporal_status": "localized", "intervals": [ { "clip_relative": { "start_s": 1.0, "end_s": 12.5 }, "source_video": { "start_s": 12.64, "end_s": 24.14 } } ], "evidence": "Samples 2–13 show the selected canoeist progressing through turbulent rapids while controlling the canoe with repeated strokes." }, { "id": "action_5", "mentions": [ { "text": "move", "occurrence": 1, "ranges": [ { "start": 231, "end": 235 } ], "particle": { "text": null, "occurrence": 1 } } ], "type": "simple", "subject_refs": [ "person_1", "upper_body_1", "arms_1" ], "object_refs": [], "confidence": "high", "mask_relation": "subject_matches_selected_person_mask", "temporal_status": "localized", "intervals": [ { "clip_relative": { "start_s": 1.0, "end_s": 12.5 }, "source_video": { "start_s": 12.64, "end_s": 24.14 } } ], "evidence": "The selected person's torso and arms cycle through extended and retracted positions across samples 2–13." } ], "notes": [ "The supplied mask at sample 1 (clip-relative 0.000 s) marks a different canoeist farther ahead, not the captioned red-beanie person; that sample is excluded from action support.", "From sample 2 onward, the supplied masks consistently identify the captioned person in the red beanie, blue jacket, and black life vest.", "Only the person has a supplied masklet. The clothing, canoe, paddle, rapids, upper body, and arms have no separate source masks, so their source-mask links are null.", "\"is seen\" is excluded because it asserts visibility rather than a physical action. \"motion\" and \"stroke\" are nominal event descriptions rather than qualifying light-verb predicates." ] }, "normalization_repairs": 0 }, { "source_video_id": "v_-IkL9Kq9zfQ", "caption_count_category": 3, "clip": { "caption_index": 3, "caption": "The person briefly stops paddling and raises their right hand to the side of their head. They quickly resume paddling on the left side of the canoe, continuing their journey downstream. The canoe moves further away through the rapids.", "source_interval_sec": [ 24.14, 41.36 ], "loom_clip_id": "v_-IkL9Kq9zfQ_604_1034", "clip_frame_interval_4fps": [ 604, 1034 ], "sample_count_1fps": 18, "sampled_frames": [ { "frame": "00000", "timestamp_sec": 24.14, "offset_sec": 0, "rgb_path": "assets/cat_03/v_-IkL9Kq9zfQ_604_1034/0000s_00000_rgb.jpg", "mask_path": "assets/cat_03/v_-IkL9Kq9zfQ_604_1034/0000s_00000_mask.png", "mask_status": "available" }, { "frame": "00024", "timestamp_sec": 25.14, "offset_sec": 1, "rgb_path": "assets/cat_03/v_-IkL9Kq9zfQ_604_1034/0001s_00024_rgb.jpg", "mask_path": "assets/cat_03/v_-IkL9Kq9zfQ_604_1034/0001s_00024_mask.png", "mask_status": "available" }, { "frame": "00048", "timestamp_sec": 26.14, "offset_sec": 2, "rgb_path": "assets/cat_03/v_-IkL9Kq9zfQ_604_1034/0002s_00048_rgb.jpg", "mask_path": "assets/cat_03/v_-IkL9Kq9zfQ_604_1034/0002s_00048_mask.png", "mask_status": "available" }, { "frame": "00076", "timestamp_sec": 27.14, "offset_sec": 3, "rgb_path": "assets/cat_03/v_-IkL9Kq9zfQ_604_1034/0003s_00076_rgb.jpg", "mask_path": "assets/cat_03/v_-IkL9Kq9zfQ_604_1034/0003s_00076_mask.png", "mask_status": "available" }, { "frame": "00100", "timestamp_sec": 28.14, "offset_sec": 4, "rgb_path": "assets/cat_03/v_-IkL9Kq9zfQ_604_1034/0004s_00100_rgb.jpg", "mask_path": "assets/cat_03/v_-IkL9Kq9zfQ_604_1034/0004s_00100_mask.png", "mask_status": "available" }, { "frame": "00124", "timestamp_sec": 29.14, "offset_sec": 5, "rgb_path": "assets/cat_03/v_-IkL9Kq9zfQ_604_1034/0005s_00124_rgb.jpg", "mask_path": "assets/cat_03/v_-IkL9Kq9zfQ_604_1034/0005s_00124_mask.png", "mask_status": "available" }, { "frame": "00148", "timestamp_sec": 30.14, "offset_sec": 6, "rgb_path": "assets/cat_03/v_-IkL9Kq9zfQ_604_1034/0006s_00148_rgb.jpg", "mask_path": "assets/cat_03/v_-IkL9Kq9zfQ_604_1034/0006s_00148_mask.png", "mask_status": "available" }, { "frame": "00176", "timestamp_sec": 31.14, "offset_sec": 7, "rgb_path": "assets/cat_03/v_-IkL9Kq9zfQ_604_1034/0007s_00176_rgb.jpg", "mask_path": "assets/cat_03/v_-IkL9Kq9zfQ_604_1034/0007s_00176_mask.png", "mask_status": "available" }, { "frame": "00200", "timestamp_sec": 32.14, "offset_sec": 8, "rgb_path": "assets/cat_03/v_-IkL9Kq9zfQ_604_1034/0008s_00200_rgb.jpg", "mask_path": "assets/cat_03/v_-IkL9Kq9zfQ_604_1034/0008s_00200_mask.png", "mask_status": "available" }, { "frame": "00224", "timestamp_sec": 33.14, "offset_sec": 9, "rgb_path": "assets/cat_03/v_-IkL9Kq9zfQ_604_1034/0009s_00224_rgb.jpg", "mask_path": "assets/cat_03/v_-IkL9Kq9zfQ_604_1034/0009s_00224_mask.png", "mask_status": "available" }, { "frame": "00248", "timestamp_sec": 34.14, "offset_sec": 10, "rgb_path": "assets/cat_03/v_-IkL9Kq9zfQ_604_1034/0010s_00248_rgb.jpg", "mask_path": "assets/cat_03/v_-IkL9Kq9zfQ_604_1034/0010s_00248_mask.png", "mask_status": "available" }, { "frame": "00276", "timestamp_sec": 35.14, "offset_sec": 11, "rgb_path": "assets/cat_03/v_-IkL9Kq9zfQ_604_1034/0011s_00276_rgb.jpg", "mask_path": "assets/cat_03/v_-IkL9Kq9zfQ_604_1034/0011s_00276_mask.png", "mask_status": "available" }, { "frame": "00300", "timestamp_sec": 36.14, "offset_sec": 12, "rgb_path": "assets/cat_03/v_-IkL9Kq9zfQ_604_1034/0012s_00300_rgb.jpg", "mask_path": "assets/cat_03/v_-IkL9Kq9zfQ_604_1034/0012s_00300_mask.png", "mask_status": "available" }, { "frame": "00324", "timestamp_sec": 37.14, "offset_sec": 13, "rgb_path": "assets/cat_03/v_-IkL9Kq9zfQ_604_1034/0013s_00324_rgb.jpg", "mask_path": "assets/cat_03/v_-IkL9Kq9zfQ_604_1034/0013s_00324_mask.png", "mask_status": "available" }, { "frame": "00348", "timestamp_sec": 38.14, "offset_sec": 14, "rgb_path": "assets/cat_03/v_-IkL9Kq9zfQ_604_1034/0014s_00348_rgb.jpg", "mask_path": "assets/cat_03/v_-IkL9Kq9zfQ_604_1034/0014s_00348_mask.png", "mask_status": "available" }, { "frame": "00376", "timestamp_sec": 39.14, "offset_sec": 15, "rgb_path": "assets/cat_03/v_-IkL9Kq9zfQ_604_1034/0015s_00376_rgb.jpg", "mask_path": "assets/cat_03/v_-IkL9Kq9zfQ_604_1034/0015s_00376_mask.png", "mask_status": "available" }, { "frame": "00400", "timestamp_sec": 40.14, "offset_sec": 16, "rgb_path": "assets/cat_03/v_-IkL9Kq9zfQ_604_1034/0016s_00400_rgb.jpg", "mask_path": "assets/cat_03/v_-IkL9Kq9zfQ_604_1034/0016s_00400_mask.png", "mask_status": "available" }, { "frame": "00424", "timestamp_sec": 41.14, "offset_sec": 17, "rgb_path": "assets/cat_03/v_-IkL9Kq9zfQ_604_1034/0017s_00424_rgb.jpg", "mask_path": "assets/cat_03/v_-IkL9Kq9zfQ_604_1034/0017s_00424_mask.png", "mask_status": "available" } ], "obj_id": [ 0 ], "anno_id": [ 55 ], "fps_inferred_from_clip_interval": 24.971 }, "usage": { "input_tokens": 29286, "input_tokens_details": { "cache_write_tokens": 29283, "cached_tokens": 0 }, "output_tokens": 4314, "output_tokens_details": { "reasoning_tokens": 3106 }, "total_tokens": 33600 }, "model": "gpt-5.6-sol", "effort": "medium", "custom_id": "cat03_cap03_v_-IkL9Kq9zfQ_604_1034", "annotation": { "caption": "The person briefly stops paddling and raises their right hand to the side of their head. They quickly resume paddling on the left side of the canoe, continuing their journey downstream. The canoe moves further away through the rapids.", "entity_groups": [ { "id": "person_1", "label": "person", "spans": [ { "text": "The person", "range": [ 0, 10 ] }, { "text": "their", "range": [ 45, 50 ] }, { "text": "their", "range": [ 77, 82 ] }, { "text": "They", "range": [ 89, 93 ] }, { "text": "their", "range": [ 160, 165 ] } ], "source_mask_link": "selected_person" }, { "id": "hand_1", "label": "hand", "spans": [ { "text": "their right hand", "range": [ 45, 61 ] } ], "source_mask_link": null }, { "id": "head_1", "label": "head", "spans": [ { "text": "their head", "range": [ 77, 87 ] } ], "source_mask_link": null }, { "id": "canoe_1", "label": "canoe", "spans": [ { "text": "the canoe", "range": [ 138, 147 ] }, { "text": "The canoe", "range": [ 186, 195 ] } ], "source_mask_link": null }, { "id": "rapids_1", "label": "rapids", "spans": [ { "text": "the rapids", "range": [ 223, 233 ] } ], "source_mask_link": null } ], "actions": [ { "id": "action_1", "mentions": [ { "text": "paddling", "ranges": [ [ 25, 33 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "type": "simple", "subject_refs": [ "person_1" ], "object_refs": [], "confidence": "high", "temporal_status": "supported", "intervals": [ { "start_s": 0.0, "end_s": 4.0, "source_start_s": 24.14, "source_end_s": 28.14 } ], "evidence": "At 0–3 seconds the masked person holds and works the paddle through successive stroke positions before the brief pause.", "mask_relation": "subject matches the supplied selected-person mask", "source_subject_refs": [ "person_1" ] }, { "id": "action_2", "mentions": [ { "text": "raises", "ranges": [ [ 38, 44 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "type": "simple", "subject_refs": [ "person_1" ], "object_refs": [ "hand_1" ], "confidence": "medium", "temporal_status": "supported", "intervals": [ { "start_s": 4.0, "end_s": 5.0, "source_start_s": 28.14, "source_end_s": 29.14 } ], "evidence": "At 4 seconds the masked person's right hand is raised beside the head, between visible paddling phases.", "mask_relation": "subject matches the supplied selected-person mask", "source_subject_refs": [ "person_1" ] }, { "id": "action_3", "mentions": [ { "text": "paddling", "ranges": [ [ 109, 117 ] ], "occurrence": 2, "particle": { "text": null, "occurrence": 1 } } ], "type": "simple", "subject_refs": [ "person_1" ], "object_refs": [], "confidence": "high", "temporal_status": "supported", "intervals": [ { "start_s": 5.0, "end_s": 10.0, "source_start_s": 29.14, "source_end_s": 34.14 }, { "start_s": 12.0, "end_s": 17.22, "source_start_s": 36.14, "source_end_s": 41.36 } ], "evidence": "From 5–9 seconds and again from 12 seconds through the final sample, the masked person shows repeated paddle-stroke and recovery postures while traveling downstream.", "mask_relation": "subject matches the supplied selected-person mask", "source_subject_refs": [ "person_1" ] } ], "notes": [ "The supplied masklet consistently identifies the person; no separate masks are supplied for the hand, head, canoe, or rapids.", "The aspectual verbs \"stops\" and \"resume\" are excluded from action spans; their embedded occurrences of \"paddling\" represent the two paddling phases.", "\"continuing their journey\" is not separately annotated because \"continuing\" aspectually frames the abstract journey expression rather than independently naming a physical action span.", "The captioned action \"moves\" has the unmasked canoe as its subject, so it is omitted under the selected-subject mask rule." ] }, "model_output_raw": { "caption": "The person briefly stops paddling and raises their right hand to the side of their head. They quickly resume paddling on the left side of the canoe, continuing their journey downstream. The canoe moves further away through the rapids.", "entity_groups": [ { "id": "person_1", "entity_type": "person", "spans": [ { "text": "The person", "range": [ 0, 10 ] }, { "text": "their", "range": [ 45, 50 ] }, { "text": "their", "range": [ 77, 82 ] }, { "text": "They", "range": [ 89, 93 ] }, { "text": "their", "range": [ 160, 165 ] } ], "source_mask": "selected_person_masklet" }, { "id": "hand_1", "entity_type": "hand", "spans": [ { "text": "their right hand", "range": [ 45, 61 ] } ], "source_mask": null }, { "id": "head_1", "entity_type": "head", "spans": [ { "text": "their head", "range": [ 77, 87 ] } ], "source_mask": null }, { "id": "canoe_1", "entity_type": "canoe", "spans": [ { "text": "the canoe", "range": [ 138, 147 ] }, { "text": "The canoe", "range": [ 186, 195 ] } ], "source_mask": null }, { "id": "rapids_1", "entity_type": "rapids", "spans": [ { "text": "the rapids", "range": [ 223, 233 ] } ], "source_mask": null } ], "actions": [ { "id": "action_1", "mentions": [ { "text": "paddling", "occurrence": 1, "ranges": [ [ 25, 33 ] ], "particle": { "text": null, "occurrence": 1 } } ], "type": "simple", "subject_refs": [ "person_1" ], "object_refs": [], "confidence": "high", "mask_relation": "subject matches the supplied selected-person mask", "temporal_status": "supported", "intervals": [ { "start_s": 0.0, "end_s": 4.0 } ], "evidence": "At 0–3 seconds the masked person holds and works the paddle through successive stroke positions before the brief pause." }, { "id": "action_2", "mentions": [ { "text": "raises", "occurrence": 1, "ranges": [ [ 38, 44 ] ], "particle": { "text": null, "occurrence": 1 } } ], "type": "simple", "subject_refs": [ "person_1" ], "object_refs": [ "hand_1" ], "confidence": "medium", "mask_relation": "subject matches the supplied selected-person mask", "temporal_status": "supported", "intervals": [ { "start_s": 4.0, "end_s": 5.0 } ], "evidence": "At 4 seconds the masked person's right hand is raised beside the head, between visible paddling phases." }, { "id": "action_3", "mentions": [ { "text": "paddling", "occurrence": 2, "ranges": [ [ 109, 117 ] ], "particle": { "text": null, "occurrence": 1 } } ], "type": "simple", "subject_refs": [ "person_1" ], "object_refs": [], "confidence": "high", "mask_relation": "subject matches the supplied selected-person mask", "temporal_status": "supported", "intervals": [ { "start_s": 5.0, "end_s": 10.0 }, { "start_s": 12.0, "end_s": 17.22 } ], "evidence": "From 5–9 seconds and again from 12 seconds through the final sample, the masked person shows repeated paddle-stroke and recovery postures while traveling downstream." } ], "notes": [ "The supplied masklet consistently identifies the person; no separate masks are supplied for the hand, head, canoe, or rapids.", "The aspectual verbs \"stops\" and \"resume\" are excluded from action spans; their embedded occurrences of \"paddling\" represent the two paddling phases.", "\"continuing their journey\" is not separately annotated because \"continuing\" aspectually frames the abstract journey expression rather than independently naming a physical action span.", "The captioned action \"moves\" has the unmasked canoe as its subject, so it is omitted under the selected-subject mask rule." ] }, "normalization_repairs": 0 }, { "source_video_id": "v_-4Q_zG9EChY", "caption_count_category": 4, "clip": { "caption_index": 1, "caption": "A shirtless person wearing blue pants balances on a slackline while holding a black object in their left hand. They begin walking forward slowly, using their arms to stabilize their movement as they take careful steps. Their body is oriented to the left.", "source_interval_sec": [ 6.01, 49.51 ], "loom_clip_id": "v_-4Q_zG9EChY_180_1484", "clip_frame_interval_4fps": [ 180, 1484 ], "sample_count_1fps": 44, "sampled_frames": [ { "frame": "00000", "timestamp_sec": 6.01, "offset_sec": 0, "rgb_path": "assets/cat_04/v_-4Q_zG9EChY_180_1484/0000s_00000_rgb.jpg", "mask_path": "assets/cat_04/v_-4Q_zG9EChY_180_1484/0000s_00000_mask.png", "mask_status": "available" }, { "frame": "00028", "timestamp_sec": 7.01, "offset_sec": 1, "rgb_path": "assets/cat_04/v_-4Q_zG9EChY_180_1484/0001s_00028_rgb.jpg", "mask_path": "assets/cat_04/v_-4Q_zG9EChY_180_1484/0001s_00028_mask.png", "mask_status": "available" }, { "frame": "00060", "timestamp_sec": 8.01, "offset_sec": 2, "rgb_path": "assets/cat_04/v_-4Q_zG9EChY_180_1484/0002s_00060_rgb.jpg", "mask_path": "assets/cat_04/v_-4Q_zG9EChY_180_1484/0002s_00060_mask.png", "mask_status": "available" }, { "frame": "00088", "timestamp_sec": 9.01, "offset_sec": 3, "rgb_path": "assets/cat_04/v_-4Q_zG9EChY_180_1484/0003s_00088_rgb.jpg", "mask_path": "assets/cat_04/v_-4Q_zG9EChY_180_1484/0003s_00088_mask.png", "mask_status": "available" }, { "frame": "00120", "timestamp_sec": 10.01, "offset_sec": 4, "rgb_path": "assets/cat_04/v_-4Q_zG9EChY_180_1484/0004s_00120_rgb.jpg", "mask_path": "assets/cat_04/v_-4Q_zG9EChY_180_1484/0004s_00120_mask.png", "mask_status": "available" }, { "frame": "00148", "timestamp_sec": 11.01, "offset_sec": 5, "rgb_path": "assets/cat_04/v_-4Q_zG9EChY_180_1484/0005s_00148_rgb.jpg", "mask_path": "assets/cat_04/v_-4Q_zG9EChY_180_1484/0005s_00148_mask.png", "mask_status": "available" }, { "frame": "00180", "timestamp_sec": 12.01, "offset_sec": 6, "rgb_path": "assets/cat_04/v_-4Q_zG9EChY_180_1484/0006s_00180_rgb.jpg", "mask_path": "assets/cat_04/v_-4Q_zG9EChY_180_1484/0006s_00180_mask.png", "mask_status": "available" }, { "frame": "00208", "timestamp_sec": 13.01, "offset_sec": 7, "rgb_path": "assets/cat_04/v_-4Q_zG9EChY_180_1484/0007s_00208_rgb.jpg", "mask_path": "assets/cat_04/v_-4Q_zG9EChY_180_1484/0007s_00208_mask.png", "mask_status": "available" }, { "frame": "00240", "timestamp_sec": 14.01, "offset_sec": 8, "rgb_path": "assets/cat_04/v_-4Q_zG9EChY_180_1484/0008s_00240_rgb.jpg", "mask_path": "assets/cat_04/v_-4Q_zG9EChY_180_1484/0008s_00240_mask.png", "mask_status": "available" }, { "frame": "00268", "timestamp_sec": 15.01, "offset_sec": 9, "rgb_path": "assets/cat_04/v_-4Q_zG9EChY_180_1484/0009s_00268_rgb.jpg", "mask_path": "assets/cat_04/v_-4Q_zG9EChY_180_1484/0009s_00268_mask.png", "mask_status": "available" }, { "frame": "00300", "timestamp_sec": 16.01, "offset_sec": 10, "rgb_path": "assets/cat_04/v_-4Q_zG9EChY_180_1484/0010s_00300_rgb.jpg", "mask_path": "assets/cat_04/v_-4Q_zG9EChY_180_1484/0010s_00300_mask.png", "mask_status": "available" }, { "frame": "00328", "timestamp_sec": 17.01, "offset_sec": 11, "rgb_path": "assets/cat_04/v_-4Q_zG9EChY_180_1484/0011s_00328_rgb.jpg", "mask_path": "assets/cat_04/v_-4Q_zG9EChY_180_1484/0011s_00328_mask.png", "mask_status": "available" }, { "frame": "00360", "timestamp_sec": 18.01, "offset_sec": 12, "rgb_path": "assets/cat_04/v_-4Q_zG9EChY_180_1484/0012s_00360_rgb.jpg", "mask_path": "assets/cat_04/v_-4Q_zG9EChY_180_1484/0012s_00360_mask.png", "mask_status": "available" }, { "frame": "00388", "timestamp_sec": 19.01, "offset_sec": 13, "rgb_path": "assets/cat_04/v_-4Q_zG9EChY_180_1484/0013s_00388_rgb.jpg", "mask_path": "assets/cat_04/v_-4Q_zG9EChY_180_1484/0013s_00388_mask.png", "mask_status": "available" }, { "frame": "00420", "timestamp_sec": 20.01, "offset_sec": 14, "rgb_path": "assets/cat_04/v_-4Q_zG9EChY_180_1484/0014s_00420_rgb.jpg", "mask_path": "assets/cat_04/v_-4Q_zG9EChY_180_1484/0014s_00420_mask.png", "mask_status": "available" }, { "frame": "00448", "timestamp_sec": 21.01, "offset_sec": 15, "rgb_path": "assets/cat_04/v_-4Q_zG9EChY_180_1484/0015s_00448_rgb.jpg", "mask_path": "assets/cat_04/v_-4Q_zG9EChY_180_1484/0015s_00448_mask.png", "mask_status": "available" }, { "frame": "00480", "timestamp_sec": 22.01, "offset_sec": 16, "rgb_path": "assets/cat_04/v_-4Q_zG9EChY_180_1484/0016s_00480_rgb.jpg", "mask_path": "assets/cat_04/v_-4Q_zG9EChY_180_1484/0016s_00480_mask.png", "mask_status": "available" }, { "frame": "00508", "timestamp_sec": 23.01, "offset_sec": 17, "rgb_path": "assets/cat_04/v_-4Q_zG9EChY_180_1484/0017s_00508_rgb.jpg", "mask_path": "assets/cat_04/v_-4Q_zG9EChY_180_1484/0017s_00508_mask.png", "mask_status": "available" }, { "frame": "00540", "timestamp_sec": 24.01, "offset_sec": 18, "rgb_path": "assets/cat_04/v_-4Q_zG9EChY_180_1484/0018s_00540_rgb.jpg", "mask_path": "assets/cat_04/v_-4Q_zG9EChY_180_1484/0018s_00540_mask.png", "mask_status": "available" }, { "frame": "00568", "timestamp_sec": 25.01, "offset_sec": 19, "rgb_path": "assets/cat_04/v_-4Q_zG9EChY_180_1484/0019s_00568_rgb.jpg", "mask_path": "assets/cat_04/v_-4Q_zG9EChY_180_1484/0019s_00568_mask.png", "mask_status": "available" }, { "frame": "00600", "timestamp_sec": 26.01, "offset_sec": 20, "rgb_path": "assets/cat_04/v_-4Q_zG9EChY_180_1484/0020s_00600_rgb.jpg", "mask_path": "assets/cat_04/v_-4Q_zG9EChY_180_1484/0020s_00600_mask.png", "mask_status": "available" }, { "frame": "00628", "timestamp_sec": 27.01, "offset_sec": 21, "rgb_path": "assets/cat_04/v_-4Q_zG9EChY_180_1484/0021s_00628_rgb.jpg", "mask_path": "assets/cat_04/v_-4Q_zG9EChY_180_1484/0021s_00628_mask.png", "mask_status": "available" }, { "frame": "00660", "timestamp_sec": 28.01, "offset_sec": 22, "rgb_path": "assets/cat_04/v_-4Q_zG9EChY_180_1484/0022s_00660_rgb.jpg", "mask_path": "assets/cat_04/v_-4Q_zG9EChY_180_1484/0022s_00660_mask.png", "mask_status": "available" }, { "frame": "00688", "timestamp_sec": 29.01, "offset_sec": 23, "rgb_path": "assets/cat_04/v_-4Q_zG9EChY_180_1484/0023s_00688_rgb.jpg", "mask_path": "assets/cat_04/v_-4Q_zG9EChY_180_1484/0023s_00688_mask.png", "mask_status": "available" }, { "frame": "00720", "timestamp_sec": 30.01, "offset_sec": 24, "rgb_path": "assets/cat_04/v_-4Q_zG9EChY_180_1484/0024s_00720_rgb.jpg", "mask_path": "assets/cat_04/v_-4Q_zG9EChY_180_1484/0024s_00720_mask.png", "mask_status": "available" }, { "frame": "00748", "timestamp_sec": 31.01, "offset_sec": 25, "rgb_path": "assets/cat_04/v_-4Q_zG9EChY_180_1484/0025s_00748_rgb.jpg", "mask_path": "assets/cat_04/v_-4Q_zG9EChY_180_1484/0025s_00748_mask.png", "mask_status": "available" }, { "frame": "00780", "timestamp_sec": 32.01, "offset_sec": 26, "rgb_path": "assets/cat_04/v_-4Q_zG9EChY_180_1484/0026s_00780_rgb.jpg", "mask_path": "assets/cat_04/v_-4Q_zG9EChY_180_1484/0026s_00780_mask.png", "mask_status": "available" }, { "frame": "00808", "timestamp_sec": 33.01, "offset_sec": 27, "rgb_path": "assets/cat_04/v_-4Q_zG9EChY_180_1484/0027s_00808_rgb.jpg", "mask_path": "assets/cat_04/v_-4Q_zG9EChY_180_1484/0027s_00808_mask.png", "mask_status": "available" }, { "frame": "00840", "timestamp_sec": 34.01, "offset_sec": 28, "rgb_path": "assets/cat_04/v_-4Q_zG9EChY_180_1484/0028s_00840_rgb.jpg", "mask_path": "assets/cat_04/v_-4Q_zG9EChY_180_1484/0028s_00840_mask.png", "mask_status": "available" }, { "frame": "00868", "timestamp_sec": 35.01, "offset_sec": 29, "rgb_path": "assets/cat_04/v_-4Q_zG9EChY_180_1484/0029s_00868_rgb.jpg", "mask_path": "assets/cat_04/v_-4Q_zG9EChY_180_1484/0029s_00868_mask.png", "mask_status": "available" }, { "frame": "00900", "timestamp_sec": 36.01, "offset_sec": 30, "rgb_path": "assets/cat_04/v_-4Q_zG9EChY_180_1484/0030s_00900_rgb.jpg", "mask_path": "assets/cat_04/v_-4Q_zG9EChY_180_1484/0030s_00900_mask.png", "mask_status": "available" }, { "frame": "00928", "timestamp_sec": 37.01, "offset_sec": 31, "rgb_path": "assets/cat_04/v_-4Q_zG9EChY_180_1484/0031s_00928_rgb.jpg", "mask_path": "assets/cat_04/v_-4Q_zG9EChY_180_1484/0031s_00928_mask.png", "mask_status": "available" }, { "frame": "00960", "timestamp_sec": 38.01, "offset_sec": 32, "rgb_path": "assets/cat_04/v_-4Q_zG9EChY_180_1484/0032s_00960_rgb.jpg", "mask_path": "assets/cat_04/v_-4Q_zG9EChY_180_1484/0032s_00960_mask.png", "mask_status": "available" }, { "frame": "00988", "timestamp_sec": 39.01, "offset_sec": 33, "rgb_path": "assets/cat_04/v_-4Q_zG9EChY_180_1484/0033s_00988_rgb.jpg", "mask_path": "assets/cat_04/v_-4Q_zG9EChY_180_1484/0033s_00988_mask.png", "mask_status": "available" }, { "frame": "01020", "timestamp_sec": 40.01, "offset_sec": 34, "rgb_path": "assets/cat_04/v_-4Q_zG9EChY_180_1484/0034s_01020_rgb.jpg", "mask_path": "assets/cat_04/v_-4Q_zG9EChY_180_1484/0034s_01020_mask.png", "mask_status": "available" }, { "frame": "01048", "timestamp_sec": 41.01, "offset_sec": 35, "rgb_path": "assets/cat_04/v_-4Q_zG9EChY_180_1484/0035s_01048_rgb.jpg", "mask_path": "assets/cat_04/v_-4Q_zG9EChY_180_1484/0035s_01048_mask.png", "mask_status": "available" }, { "frame": "01080", "timestamp_sec": 42.01, "offset_sec": 36, "rgb_path": "assets/cat_04/v_-4Q_zG9EChY_180_1484/0036s_01080_rgb.jpg", "mask_path": "assets/cat_04/v_-4Q_zG9EChY_180_1484/0036s_01080_mask.png", "mask_status": "available" }, { "frame": "01108", "timestamp_sec": 43.01, "offset_sec": 37, "rgb_path": "assets/cat_04/v_-4Q_zG9EChY_180_1484/0037s_01108_rgb.jpg", "mask_path": "assets/cat_04/v_-4Q_zG9EChY_180_1484/0037s_01108_mask.png", "mask_status": "available" }, { "frame": "01140", "timestamp_sec": 44.01, "offset_sec": 38, "rgb_path": "assets/cat_04/v_-4Q_zG9EChY_180_1484/0038s_01140_rgb.jpg", "mask_path": "assets/cat_04/v_-4Q_zG9EChY_180_1484/0038s_01140_mask.png", "mask_status": "available" }, { "frame": "01168", "timestamp_sec": 45.01, "offset_sec": 39, "rgb_path": "assets/cat_04/v_-4Q_zG9EChY_180_1484/0039s_01168_rgb.jpg", "mask_path": "assets/cat_04/v_-4Q_zG9EChY_180_1484/0039s_01168_mask.png", "mask_status": "available" }, { "frame": "01200", "timestamp_sec": 46.01, "offset_sec": 40, "rgb_path": "assets/cat_04/v_-4Q_zG9EChY_180_1484/0040s_01200_rgb.jpg", "mask_path": "assets/cat_04/v_-4Q_zG9EChY_180_1484/0040s_01200_mask.png", "mask_status": "available" }, { "frame": "01228", "timestamp_sec": 47.01, "offset_sec": 41, "rgb_path": "assets/cat_04/v_-4Q_zG9EChY_180_1484/0041s_01228_rgb.jpg", "mask_path": "assets/cat_04/v_-4Q_zG9EChY_180_1484/0041s_01228_mask.png", "mask_status": "available" }, { "frame": "01260", "timestamp_sec": 48.01, "offset_sec": 42, "rgb_path": "assets/cat_04/v_-4Q_zG9EChY_180_1484/0042s_01260_rgb.jpg", "mask_path": "assets/cat_04/v_-4Q_zG9EChY_180_1484/0042s_01260_mask.png", "mask_status": "available" }, { "frame": "01288", "timestamp_sec": 49.01, "offset_sec": 43, "rgb_path": "assets/cat_04/v_-4Q_zG9EChY_180_1484/0043s_01288_rgb.jpg", "mask_path": "assets/cat_04/v_-4Q_zG9EChY_180_1484/0043s_01288_mask.png", "mask_status": "available" } ], "obj_id": [ 0 ], "anno_id": [ 28 ], "fps_inferred_from_clip_interval": 29.977 }, "usage": { "input_tokens": 179327, "input_tokens_details": { "cache_write_tokens": 179324, "cached_tokens": 0 }, "output_tokens": 4741, "output_tokens_details": { "reasoning_tokens": 2588 }, "total_tokens": 184068 }, "model": "gpt-5.6-sol", "effort": "medium", "custom_id": "cat04_cap01_v_-4Q_zG9EChY_180_1484", "annotation": { "caption": "A shirtless person wearing blue pants balances on a slackline while holding a black object in their left hand. They begin walking forward slowly, using their arms to stabilize their movement as they take careful steps. Their body is oriented to the left.", "entity_groups": [ { "id": "person_1", "label": "person", "spans": [ { "text": "A shirtless person", "range": [ 0, 18 ] }, { "text": "their", "range": [ 94, 99 ] }, { "text": "They", "range": [ 111, 115 ] }, { "text": "their", "range": [ 152, 157 ] }, { "text": "their", "range": [ 176, 181 ] }, { "text": "they", "range": [ 194, 198 ] }, { "text": "Their", "range": [ 219, 224 ] } ], "source_mask_link": "selected_person" }, { "id": "pants_1", "label": "pants", "spans": [ { "text": "blue pants", "range": [ 27, 37 ] } ], "source_mask_link": null }, { "id": "slackline_1", "label": "slackline", "spans": [ { "text": "a slackline", "range": [ 50, 61 ] } ], "source_mask_link": null }, { "id": "object_1", "label": "object", "spans": [ { "text": "a black object", "range": [ 76, 90 ] } ], "source_mask_link": null }, { "id": "hand_1", "label": "hand", "spans": [ { "text": "their left hand", "range": [ 94, 109 ] } ], "source_mask_link": null }, { "id": "arms_1", "label": "arms", "spans": [ { "text": "their arms", "range": [ 152, 162 ] } ], "source_mask_link": null }, { "id": "body_1", "label": "body", "spans": [ { "text": "Their body", "range": [ 219, 229 ] } ], "source_mask_link": null } ], "actions": [ { "id": "action_1", "mentions": [ { "text": "wearing", "ranges": [ [ 19, 26 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "type": "simple", "subject_refs": [ "person_1" ], "object_refs": [ "pants_1" ], "confidence": "high", "temporal_status": "localized", "intervals": [ { "start_s": 1.0, "end_s": 6.0, "source_start_s": 7.01, "source_end_s": 12.01 }, { "start_s": 7.0, "end_s": 43.5, "source_start_s": 13.01, "source_end_s": 49.51 } ], "evidence": "The mask-linked person visibly wears blue pants throughout the unobstructed sampled frames; sample 6 is a transition.", "mask_relation": "The action subject matches the supplied selected-person mask.", "source_subject_refs": [ "person_1" ] }, { "id": "action_2", "mentions": [ { "text": "balances", "ranges": [ [ 38, 46 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "type": "simple", "subject_refs": [ "person_1" ], "object_refs": [ "slackline_1" ], "confidence": "high", "temporal_status": "localized", "intervals": [ { "start_s": 7.0, "end_s": 43.5, "source_start_s": 13.01, "source_end_s": 49.51 } ], "evidence": "From sample 8 onward, the masked person remains upright by the slackline while repeatedly shifting posture and arm position to maintain balance.", "mask_relation": "The action subject matches the supplied selected-person mask.", "source_subject_refs": [ "person_1" ] }, { "id": "action_3", "mentions": [ { "text": "holding", "ranges": [ [ 68, 75 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "type": "simple", "subject_refs": [ "person_1" ], "object_refs": [ "object_1" ], "confidence": "medium", "temporal_status": "localized", "intervals": [ { "start_s": 1.0, "end_s": 6.0, "source_start_s": 7.01, "source_end_s": 12.01 } ], "evidence": "Samples 2–6 show the masked person with a dark object at the forward hand area; the object is not clearly visible after the transition.", "mask_relation": "The action subject matches the supplied selected-person mask.", "source_subject_refs": [ "person_1" ] }, { "id": "action_4", "mentions": [ { "text": "walking", "ranges": [ [ 122, 129 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } }, { "text": "take careful steps", "ranges": [ [ 199, 217 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "type": "simple", "subject_refs": [ "person_1" ], "object_refs": [], "confidence": "high", "temporal_status": "localized", "intervals": [ { "start_s": 7.0, "end_s": 43.5, "source_start_s": 13.01, "source_end_s": 49.51 } ], "evidence": "Across samples 8–44, the mask-linked person progresses to the right with alternating leg and foot positions. “walking” and “take careful steps” describe the same event.", "mask_relation": "The action subject matches the supplied selected-person mask.", "source_subject_refs": [ "person_1" ] }, { "id": "action_5", "mentions": [ { "text": "using", "ranges": [ [ 146, 151 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "type": "simple", "subject_refs": [ "person_1" ], "object_refs": [ "arms_1" ], "confidence": "high", "temporal_status": "localized", "intervals": [ { "start_s": 7.0, "end_s": 43.5, "source_start_s": 13.01, "source_end_s": 49.51 } ], "evidence": "The masked person repeatedly extends, lowers, and raises the arms while moving along the slackline.", "mask_relation": "The action subject matches the supplied selected-person mask.", "source_subject_refs": [ "person_1" ] }, { "id": "action_6", "mentions": [ { "text": "stabilize", "ranges": [ [ 166, 175 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "type": "simple", "subject_refs": [ "person_1" ], "object_refs": [], "confidence": "high", "temporal_status": "localized", "intervals": [ { "start_s": 7.0, "end_s": 43.5, "source_start_s": 13.01, "source_end_s": 49.51 } ], "evidence": "Repeated lateral body adjustments and changing arm positions visibly support active stabilization by the mask-linked person.", "mask_relation": "The action subject matches the supplied selected-person mask.", "source_subject_refs": [ "person_1" ] } ], "notes": [ "The supplied masklet tracks only the shirtless person; all other entity groups therefore have null source-mask links.", "Sample 1 is substantially obscured by an introductory graphic, and sample 7 is a transition, so neither was used as clean temporal support.", "The aspectual predicate “begin” is excluded; the embedded action “walking” is retained.", "“is oriented” is a static orientation relation rather than the V5 lexical-face exception, so it is not annotated as an action.", "“movement” and “careful steps” name an abstract process and an action within a light-verb construction, respectively; they are not separate concrete entity groups." ] }, "model_output_raw": { "caption": "A shirtless person wearing blue pants balances on a slackline while holding a black object in their left hand. They begin walking forward slowly, using their arms to stabilize their movement as they take careful steps. Their body is oriented to the left.", "entity_groups": [ { "id": "person_1", "label": "person", "spans": [ { "text": "A shirtless person", "range": [ 0, 18 ] }, { "text": "their", "range": [ 94, 99 ] }, { "text": "They", "range": [ 111, 115 ] }, { "text": "their", "range": [ 152, 157 ] }, { "text": "their", "range": [ 176, 181 ] }, { "text": "they", "range": [ 194, 198 ] }, { "text": "Their", "range": [ 219, 224 ] } ], "source_mask": { "track_id": "selected_person", "description": "The supplied masklet identifies the shirtless person in blue pants." } }, { "id": "pants_1", "label": "pants", "spans": [ { "text": "blue pants", "range": [ 27, 37 ] } ], "source_mask": null }, { "id": "slackline_1", "label": "slackline", "spans": [ { "text": "a slackline", "range": [ 50, 61 ] } ], "source_mask": null }, { "id": "object_1", "label": "object", "spans": [ { "text": "a black object", "range": [ 76, 90 ] } ], "source_mask": null }, { "id": "hand_1", "label": "hand", "spans": [ { "text": "their left hand", "range": [ 94, 109 ] } ], "source_mask": null }, { "id": "arms_1", "label": "arms", "spans": [ { "text": "their arms", "range": [ 152, 162 ] } ], "source_mask": null }, { "id": "body_1", "label": "body", "spans": [ { "text": "Their body", "range": [ 219, 229 ] } ], "source_mask": null } ], "actions": [ { "id": "action_1", "mentions": [ { "text": "wearing", "ranges": [ [ 19, 26 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "type": "simple", "subject_refs": [ "person_1" ], "object_refs": [ "pants_1" ], "confidence": "high", "temporal_status": "localized", "intervals": [ { "clip_relative": [ 1.0, 6.0 ], "source_video": [ 7.01, 12.01 ] }, { "clip_relative": [ 7.0, 43.5 ], "source_video": [ 13.01, 49.51 ] } ], "evidence": "The mask-linked person visibly wears blue pants throughout the unobstructed sampled frames; sample 6 is a transition.", "mask_relation": "The action subject matches the supplied selected-person mask." }, { "id": "action_2", "mentions": [ { "text": "balances", "ranges": [ [ 38, 46 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "type": "simple", "subject_refs": [ "person_1" ], "object_refs": [ "slackline_1" ], "confidence": "high", "temporal_status": "localized", "intervals": [ { "clip_relative": [ 7.0, 43.5 ], "source_video": [ 13.01, 49.51 ] } ], "evidence": "From sample 8 onward, the masked person remains upright by the slackline while repeatedly shifting posture and arm position to maintain balance.", "mask_relation": "The action subject matches the supplied selected-person mask." }, { "id": "action_3", "mentions": [ { "text": "holding", "ranges": [ [ 68, 75 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "type": "simple", "subject_refs": [ "person_1" ], "object_refs": [ "object_1" ], "confidence": "medium", "temporal_status": "localized", "intervals": [ { "clip_relative": [ 1.0, 6.0 ], "source_video": [ 7.01, 12.01 ] } ], "evidence": "Samples 2–6 show the masked person with a dark object at the forward hand area; the object is not clearly visible after the transition.", "mask_relation": "The action subject matches the supplied selected-person mask." }, { "id": "action_4", "mentions": [ { "text": "walking", "ranges": [ [ 122, 129 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } }, { "text": "take careful steps", "ranges": [ [ 199, 217 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "type": "simple", "subject_refs": [ "person_1" ], "object_refs": [], "confidence": "high", "temporal_status": "localized", "intervals": [ { "clip_relative": [ 7.0, 43.5 ], "source_video": [ 13.01, 49.51 ] } ], "evidence": "Across samples 8–44, the mask-linked person progresses to the right with alternating leg and foot positions. “walking” and “take careful steps” describe the same event.", "mask_relation": "The action subject matches the supplied selected-person mask." }, { "id": "action_5", "mentions": [ { "text": "using", "ranges": [ [ 146, 151 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "type": "simple", "subject_refs": [ "person_1" ], "object_refs": [ "arms_1" ], "confidence": "high", "temporal_status": "localized", "intervals": [ { "clip_relative": [ 7.0, 43.5 ], "source_video": [ 13.01, 49.51 ] } ], "evidence": "The masked person repeatedly extends, lowers, and raises the arms while moving along the slackline.", "mask_relation": "The action subject matches the supplied selected-person mask." }, { "id": "action_6", "mentions": [ { "text": "stabilize", "ranges": [ [ 166, 175 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "type": "simple", "subject_refs": [ "person_1" ], "object_refs": [], "confidence": "high", "temporal_status": "localized", "intervals": [ { "clip_relative": [ 7.0, 43.5 ], "source_video": [ 13.01, 49.51 ] } ], "evidence": "Repeated lateral body adjustments and changing arm positions visibly support active stabilization by the mask-linked person.", "mask_relation": "The action subject matches the supplied selected-person mask." } ], "notes": [ "The supplied masklet tracks only the shirtless person; all other entity groups therefore have null source-mask links.", "Sample 1 is substantially obscured by an introductory graphic, and sample 7 is a transition, so neither was used as clean temporal support.", "The aspectual predicate “begin” is excluded; the embedded action “walking” is retained.", "“is oriented” is a static orientation relation rather than the V5 lexical-face exception, so it is not annotated as an action.", "“movement” and “careful steps” name an abstract process and an action within a light-verb construction, respectively; they are not separate concrete entity groups." ] }, "normalization_repairs": 0 }, { "source_video_id": "v_-4Q_zG9EChY", "caption_count_category": 4, "clip": { "caption_index": 2, "caption": "The person tosses the black object away to their left, briefly squats to regain balance, and then continues walking forward. They extend their arms out to the sides as they move steadily toward the far end of the slackline. Their pace remains slow and controlled.", "source_interval_sec": [ 49.51, 100.51 ], "loom_clip_id": "v_-4Q_zG9EChY_1484_3012", "clip_frame_interval_4fps": [ 1484, 3012 ], "sample_count_1fps": 52, "sampled_frames": [ { "frame": "00000", "timestamp_sec": 49.51, "offset_sec": 0, "rgb_path": "assets/cat_04/v_-4Q_zG9EChY_1484_3012/0000s_00000_rgb.jpg", "mask_path": "assets/cat_04/v_-4Q_zG9EChY_1484_3012/0000s_00000_mask.png", "mask_status": "available" }, { "frame": "00028", "timestamp_sec": 50.51, "offset_sec": 1, "rgb_path": "assets/cat_04/v_-4Q_zG9EChY_1484_3012/0001s_00028_rgb.jpg", "mask_path": "assets/cat_04/v_-4Q_zG9EChY_1484_3012/0001s_00028_mask.png", "mask_status": "available" }, { "frame": "00060", "timestamp_sec": 51.51, "offset_sec": 2, "rgb_path": "assets/cat_04/v_-4Q_zG9EChY_1484_3012/0002s_00060_rgb.jpg", "mask_path": "assets/cat_04/v_-4Q_zG9EChY_1484_3012/0002s_00060_mask.png", "mask_status": "available" }, { "frame": "00088", "timestamp_sec": 52.51, "offset_sec": 3, "rgb_path": "assets/cat_04/v_-4Q_zG9EChY_1484_3012/0003s_00088_rgb.jpg", "mask_path": "assets/cat_04/v_-4Q_zG9EChY_1484_3012/0003s_00088_mask.png", "mask_status": "available" }, { "frame": "00120", "timestamp_sec": 53.51, "offset_sec": 4, "rgb_path": "assets/cat_04/v_-4Q_zG9EChY_1484_3012/0004s_00120_rgb.jpg", "mask_path": "assets/cat_04/v_-4Q_zG9EChY_1484_3012/0004s_00120_mask.png", "mask_status": "available" }, { "frame": "00148", "timestamp_sec": 54.51, "offset_sec": 5, "rgb_path": "assets/cat_04/v_-4Q_zG9EChY_1484_3012/0005s_00148_rgb.jpg", "mask_path": "assets/cat_04/v_-4Q_zG9EChY_1484_3012/0005s_00148_mask.png", "mask_status": "available" }, { "frame": "00180", "timestamp_sec": 55.51, "offset_sec": 6, "rgb_path": "assets/cat_04/v_-4Q_zG9EChY_1484_3012/0006s_00180_rgb.jpg", "mask_path": "assets/cat_04/v_-4Q_zG9EChY_1484_3012/0006s_00180_mask.png", "mask_status": "available" }, { "frame": "00208", "timestamp_sec": 56.51, "offset_sec": 7, "rgb_path": "assets/cat_04/v_-4Q_zG9EChY_1484_3012/0007s_00208_rgb.jpg", "mask_path": "assets/cat_04/v_-4Q_zG9EChY_1484_3012/0007s_00208_mask.png", "mask_status": "available" }, { "frame": "00240", "timestamp_sec": 57.51, "offset_sec": 8, "rgb_path": "assets/cat_04/v_-4Q_zG9EChY_1484_3012/0008s_00240_rgb.jpg", "mask_path": "assets/cat_04/v_-4Q_zG9EChY_1484_3012/0008s_00240_mask.png", "mask_status": "available" }, { "frame": "00268", "timestamp_sec": 58.51, "offset_sec": 9, "rgb_path": "assets/cat_04/v_-4Q_zG9EChY_1484_3012/0009s_00268_rgb.jpg", "mask_path": "assets/cat_04/v_-4Q_zG9EChY_1484_3012/0009s_00268_mask.png", "mask_status": "available" }, { "frame": "00300", "timestamp_sec": 59.51, "offset_sec": 10, "rgb_path": "assets/cat_04/v_-4Q_zG9EChY_1484_3012/0010s_00300_rgb.jpg", "mask_path": "assets/cat_04/v_-4Q_zG9EChY_1484_3012/0010s_00300_mask.png", "mask_status": "available" }, { "frame": "00328", "timestamp_sec": 60.51, "offset_sec": 11, "rgb_path": "assets/cat_04/v_-4Q_zG9EChY_1484_3012/0011s_00328_rgb.jpg", "mask_path": "assets/cat_04/v_-4Q_zG9EChY_1484_3012/0011s_00328_mask.png", "mask_status": "available" }, { "frame": "00360", "timestamp_sec": 61.51, "offset_sec": 12, "rgb_path": "assets/cat_04/v_-4Q_zG9EChY_1484_3012/0012s_00360_rgb.jpg", "mask_path": "assets/cat_04/v_-4Q_zG9EChY_1484_3012/0012s_00360_mask.png", "mask_status": "available" }, { "frame": "00388", "timestamp_sec": 62.51, "offset_sec": 13, "rgb_path": "assets/cat_04/v_-4Q_zG9EChY_1484_3012/0013s_00388_rgb.jpg", "mask_path": "assets/cat_04/v_-4Q_zG9EChY_1484_3012/0013s_00388_mask.png", "mask_status": "available" }, { "frame": "00420", "timestamp_sec": 63.51, "offset_sec": 14, "rgb_path": "assets/cat_04/v_-4Q_zG9EChY_1484_3012/0014s_00420_rgb.jpg", "mask_path": "assets/cat_04/v_-4Q_zG9EChY_1484_3012/0014s_00420_mask.png", "mask_status": "available" }, { "frame": "00448", "timestamp_sec": 64.51, "offset_sec": 15, "rgb_path": "assets/cat_04/v_-4Q_zG9EChY_1484_3012/0015s_00448_rgb.jpg", "mask_path": "assets/cat_04/v_-4Q_zG9EChY_1484_3012/0015s_00448_mask.png", "mask_status": "available" }, { "frame": "00480", "timestamp_sec": 65.51, "offset_sec": 16, "rgb_path": "assets/cat_04/v_-4Q_zG9EChY_1484_3012/0016s_00480_rgb.jpg", "mask_path": "assets/cat_04/v_-4Q_zG9EChY_1484_3012/0016s_00480_mask.png", "mask_status": "available" }, { "frame": "00508", "timestamp_sec": 66.51, "offset_sec": 17, "rgb_path": "assets/cat_04/v_-4Q_zG9EChY_1484_3012/0017s_00508_rgb.jpg", "mask_path": "assets/cat_04/v_-4Q_zG9EChY_1484_3012/0017s_00508_mask.png", "mask_status": "available" }, { "frame": "00540", "timestamp_sec": 67.51, "offset_sec": 18, "rgb_path": "assets/cat_04/v_-4Q_zG9EChY_1484_3012/0018s_00540_rgb.jpg", "mask_path": "assets/cat_04/v_-4Q_zG9EChY_1484_3012/0018s_00540_mask.png", "mask_status": "available" }, { "frame": "00568", "timestamp_sec": 68.51, "offset_sec": 19, "rgb_path": "assets/cat_04/v_-4Q_zG9EChY_1484_3012/0019s_00568_rgb.jpg", "mask_path": "assets/cat_04/v_-4Q_zG9EChY_1484_3012/0019s_00568_mask.png", "mask_status": "available" }, { "frame": "00600", "timestamp_sec": 69.51, "offset_sec": 20, "rgb_path": "assets/cat_04/v_-4Q_zG9EChY_1484_3012/0020s_00600_rgb.jpg", "mask_path": "assets/cat_04/v_-4Q_zG9EChY_1484_3012/0020s_00600_mask.png", "mask_status": "available" }, { "frame": "00628", "timestamp_sec": 70.51, "offset_sec": 21, "rgb_path": "assets/cat_04/v_-4Q_zG9EChY_1484_3012/0021s_00628_rgb.jpg", "mask_path": "assets/cat_04/v_-4Q_zG9EChY_1484_3012/0021s_00628_mask.png", "mask_status": "available" }, { "frame": "00660", "timestamp_sec": 71.51, "offset_sec": 22, "rgb_path": "assets/cat_04/v_-4Q_zG9EChY_1484_3012/0022s_00660_rgb.jpg", "mask_path": "assets/cat_04/v_-4Q_zG9EChY_1484_3012/0022s_00660_mask.png", "mask_status": "available" }, { "frame": "00688", "timestamp_sec": 72.51, "offset_sec": 23, "rgb_path": "assets/cat_04/v_-4Q_zG9EChY_1484_3012/0023s_00688_rgb.jpg", "mask_path": "assets/cat_04/v_-4Q_zG9EChY_1484_3012/0023s_00688_mask.png", "mask_status": "available" }, { "frame": "00720", "timestamp_sec": 73.51, "offset_sec": 24, "rgb_path": "assets/cat_04/v_-4Q_zG9EChY_1484_3012/0024s_00720_rgb.jpg", "mask_path": "assets/cat_04/v_-4Q_zG9EChY_1484_3012/0024s_00720_mask.png", "mask_status": "available" }, { "frame": "00748", "timestamp_sec": 74.51, "offset_sec": 25, "rgb_path": "assets/cat_04/v_-4Q_zG9EChY_1484_3012/0025s_00748_rgb.jpg", "mask_path": "assets/cat_04/v_-4Q_zG9EChY_1484_3012/0025s_00748_mask.png", "mask_status": "available" }, { "frame": "00780", "timestamp_sec": 75.51, "offset_sec": 26, "rgb_path": "assets/cat_04/v_-4Q_zG9EChY_1484_3012/0026s_00780_rgb.jpg", "mask_path": "assets/cat_04/v_-4Q_zG9EChY_1484_3012/0026s_00780_mask.png", "mask_status": "available" }, { "frame": "00808", "timestamp_sec": 76.51, "offset_sec": 27, "rgb_path": "assets/cat_04/v_-4Q_zG9EChY_1484_3012/0027s_00808_rgb.jpg", "mask_path": "assets/cat_04/v_-4Q_zG9EChY_1484_3012/0027s_00808_mask.png", "mask_status": "available" }, { "frame": "00840", "timestamp_sec": 77.51, "offset_sec": 28, "rgb_path": "assets/cat_04/v_-4Q_zG9EChY_1484_3012/0028s_00840_rgb.jpg", "mask_path": "assets/cat_04/v_-4Q_zG9EChY_1484_3012/0028s_00840_mask.png", "mask_status": "available" }, { "frame": "00868", "timestamp_sec": 78.51, "offset_sec": 29, "rgb_path": "assets/cat_04/v_-4Q_zG9EChY_1484_3012/0029s_00868_rgb.jpg", "mask_path": "assets/cat_04/v_-4Q_zG9EChY_1484_3012/0029s_00868_mask.png", "mask_status": "available" }, { "frame": "00900", "timestamp_sec": 79.51, "offset_sec": 30, "rgb_path": "assets/cat_04/v_-4Q_zG9EChY_1484_3012/0030s_00900_rgb.jpg", "mask_path": "assets/cat_04/v_-4Q_zG9EChY_1484_3012/0030s_00900_mask.png", "mask_status": "available" }, { "frame": "00928", "timestamp_sec": 80.51, "offset_sec": 31, "rgb_path": "assets/cat_04/v_-4Q_zG9EChY_1484_3012/0031s_00928_rgb.jpg", "mask_path": "assets/cat_04/v_-4Q_zG9EChY_1484_3012/0031s_00928_mask.png", "mask_status": "available" }, { "frame": "00960", "timestamp_sec": 81.51, "offset_sec": 32, "rgb_path": "assets/cat_04/v_-4Q_zG9EChY_1484_3012/0032s_00960_rgb.jpg", "mask_path": "assets/cat_04/v_-4Q_zG9EChY_1484_3012/0032s_00960_mask.png", "mask_status": "available" }, { "frame": "00988", "timestamp_sec": 82.51, "offset_sec": 33, "rgb_path": "assets/cat_04/v_-4Q_zG9EChY_1484_3012/0033s_00988_rgb.jpg", "mask_path": "assets/cat_04/v_-4Q_zG9EChY_1484_3012/0033s_00988_mask.png", "mask_status": "available" }, { "frame": "01020", "timestamp_sec": 83.51, "offset_sec": 34, "rgb_path": "assets/cat_04/v_-4Q_zG9EChY_1484_3012/0034s_01020_rgb.jpg", "mask_path": "assets/cat_04/v_-4Q_zG9EChY_1484_3012/0034s_01020_mask.png", "mask_status": "available" }, { "frame": "01048", "timestamp_sec": 84.51, "offset_sec": 35, "rgb_path": "assets/cat_04/v_-4Q_zG9EChY_1484_3012/0035s_01048_rgb.jpg", "mask_path": "assets/cat_04/v_-4Q_zG9EChY_1484_3012/0035s_01048_mask.png", "mask_status": "available" }, { "frame": "01080", "timestamp_sec": 85.51, "offset_sec": 36, "rgb_path": "assets/cat_04/v_-4Q_zG9EChY_1484_3012/0036s_01080_rgb.jpg", "mask_path": "assets/cat_04/v_-4Q_zG9EChY_1484_3012/0036s_01080_mask.png", "mask_status": "available" }, { "frame": "01108", "timestamp_sec": 86.51, "offset_sec": 37, "rgb_path": "assets/cat_04/v_-4Q_zG9EChY_1484_3012/0037s_01108_rgb.jpg", "mask_path": "assets/cat_04/v_-4Q_zG9EChY_1484_3012/0037s_01108_mask.png", "mask_status": "available" }, { "frame": "01140", "timestamp_sec": 87.51, "offset_sec": 38, "rgb_path": "assets/cat_04/v_-4Q_zG9EChY_1484_3012/0038s_01140_rgb.jpg", "mask_path": "assets/cat_04/v_-4Q_zG9EChY_1484_3012/0038s_01140_mask.png", "mask_status": "available" }, { "frame": "01168", "timestamp_sec": 88.51, "offset_sec": 39, "rgb_path": "assets/cat_04/v_-4Q_zG9EChY_1484_3012/0039s_01168_rgb.jpg", "mask_path": "assets/cat_04/v_-4Q_zG9EChY_1484_3012/0039s_01168_mask.png", "mask_status": "available" }, { "frame": "01200", "timestamp_sec": 89.51, "offset_sec": 40, "rgb_path": "assets/cat_04/v_-4Q_zG9EChY_1484_3012/0040s_01200_rgb.jpg", "mask_path": "assets/cat_04/v_-4Q_zG9EChY_1484_3012/0040s_01200_mask.png", "mask_status": "available" }, { "frame": "01228", "timestamp_sec": 90.51, "offset_sec": 41, "rgb_path": "assets/cat_04/v_-4Q_zG9EChY_1484_3012/0041s_01228_rgb.jpg", "mask_path": "assets/cat_04/v_-4Q_zG9EChY_1484_3012/0041s_01228_mask.png", "mask_status": "available" }, { "frame": "01260", "timestamp_sec": 91.51, "offset_sec": 42, "rgb_path": "assets/cat_04/v_-4Q_zG9EChY_1484_3012/0042s_01260_rgb.jpg", "mask_path": "assets/cat_04/v_-4Q_zG9EChY_1484_3012/0042s_01260_mask.png", "mask_status": "available" }, { "frame": "01288", "timestamp_sec": 92.51, "offset_sec": 43, "rgb_path": "assets/cat_04/v_-4Q_zG9EChY_1484_3012/0043s_01288_rgb.jpg", "mask_path": "assets/cat_04/v_-4Q_zG9EChY_1484_3012/0043s_01288_mask.png", "mask_status": "available" }, { "frame": "01320", "timestamp_sec": 93.51, "offset_sec": 44, "rgb_path": "assets/cat_04/v_-4Q_zG9EChY_1484_3012/0044s_01320_rgb.jpg", "mask_path": "assets/cat_04/v_-4Q_zG9EChY_1484_3012/0044s_01320_mask.png", "mask_status": "available" }, { "frame": "01348", "timestamp_sec": 94.51, "offset_sec": 45, "rgb_path": "assets/cat_04/v_-4Q_zG9EChY_1484_3012/0045s_01348_rgb.jpg", "mask_path": "assets/cat_04/v_-4Q_zG9EChY_1484_3012/0045s_01348_mask.png", "mask_status": "available" }, { "frame": "01380", "timestamp_sec": 95.51, "offset_sec": 46, "rgb_path": "assets/cat_04/v_-4Q_zG9EChY_1484_3012/0046s_01380_rgb.jpg", "mask_path": "assets/cat_04/v_-4Q_zG9EChY_1484_3012/0046s_01380_mask.png", "mask_status": "available" }, { "frame": "01408", "timestamp_sec": 96.51, "offset_sec": 47, "rgb_path": "assets/cat_04/v_-4Q_zG9EChY_1484_3012/0047s_01408_rgb.jpg", "mask_path": "assets/cat_04/v_-4Q_zG9EChY_1484_3012/0047s_01408_mask.png", "mask_status": "available" }, { "frame": "01440", "timestamp_sec": 97.51, "offset_sec": 48, "rgb_path": "assets/cat_04/v_-4Q_zG9EChY_1484_3012/0048s_01440_rgb.jpg", "mask_path": "assets/cat_04/v_-4Q_zG9EChY_1484_3012/0048s_01440_mask.png", "mask_status": "available" }, { "frame": "01468", "timestamp_sec": 98.51, "offset_sec": 49, "rgb_path": "assets/cat_04/v_-4Q_zG9EChY_1484_3012/0049s_01468_rgb.jpg", "mask_path": "assets/cat_04/v_-4Q_zG9EChY_1484_3012/0049s_01468_mask.png", "mask_status": "available" }, { "frame": "01500", "timestamp_sec": 99.51, "offset_sec": 50, "rgb_path": "assets/cat_04/v_-4Q_zG9EChY_1484_3012/0050s_01500_rgb.jpg", "mask_path": "assets/cat_04/v_-4Q_zG9EChY_1484_3012/0050s_01500_mask.png", "mask_status": "available" }, { "frame": "01528", "timestamp_sec": 100.51, "offset_sec": 51.0, "rgb_path": "assets/cat_04/v_-4Q_zG9EChY_1484_3012/0051s_01528_rgb.jpg", "mask_path": "assets/cat_04/v_-4Q_zG9EChY_1484_3012/0051s_01528_mask.png", "mask_status": "available" } ], "obj_id": [ 0 ], "anno_id": [ 27 ], "fps_inferred_from_clip_interval": 29.9608 }, "usage": { "input_tokens": 209147, "input_tokens_details": { "cache_write_tokens": 209144, "cached_tokens": 0 }, "output_tokens": 5009, "output_tokens_details": { "reasoning_tokens": 3139 }, "total_tokens": 214156 }, "model": "gpt-5.6-sol", "effort": "medium", "custom_id": "cat04_cap02_v_-4Q_zG9EChY_1484_3012", "annotation": { "caption": "The person tosses the black object away to their left, briefly squats to regain balance, and then continues walking forward. They extend their arms out to the sides as they move steadily toward the far end of the slackline. Their pace remains slow and controlled.", "entity_groups": [ { "id": "entity_1", "label": "person", "spans": [ { "text": "The person", "range": [ 0, 10 ] }, { "text": "their", "range": [ 43, 48 ] }, { "text": "They", "range": [ 125, 129 ] }, { "text": "their", "range": [ 137, 142 ] }, { "text": "they", "range": [ 168, 172 ] }, { "text": "Their", "range": [ 224, 229 ] } ], "source_mask_link": "selected_person" }, { "id": "entity_2", "label": "black object", "spans": [ { "text": "the black object", "range": [ 18, 34 ] } ], "source_mask_link": null }, { "id": "entity_3", "label": "arms", "spans": [ { "text": "their arms", "range": [ 137, 147 ] } ], "source_mask_link": null }, { "id": "entity_4", "label": "slackline", "spans": [ { "text": "the slackline", "range": [ 209, 222 ] } ], "source_mask_link": null } ], "actions": [ { "id": "action_1", "mentions": [ { "text": "tosses", "ranges": [ [ 11, 17 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "type": "simple", "subject_refs": [ "entity_1" ], "object_refs": [ "entity_2" ], "confidence": "high", "temporal_status": "uncertain", "intervals": [], "evidence": "The selected person is visible and mask-linked, but no sampled clean frame clearly shows the black object being released or moving away.", "mask_relation": "matches_selected_person_mask", "source_subject_refs": [ "entity_1" ] }, { "id": "action_2", "mentions": [ { "text": "squats", "ranges": [ [ 63, 69 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "type": "posture", "subject_refs": [ "entity_1" ], "object_refs": [], "confidence": "high", "temporal_status": "localized", "intervals": [ { "start_s": 1.0, "end_s": 2.0, "source_start_s": 50.51, "source_end_s": 51.51 } ], "evidence": "At clip time 1.0 seconds, the masked person has markedly bent knees and a lowered body while balancing.", "mask_relation": "matches_selected_person_mask", "source_subject_refs": [ "entity_1" ] }, { "id": "action_3", "mentions": [ { "text": "regain balance", "ranges": [ [ 73, 87 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "type": "other", "subject_refs": [ "entity_1" ], "object_refs": [], "confidence": "high", "temporal_status": "uncertain", "intervals": [], "evidence": "Arm and torso adjustments are visible, but sparse still frames do not independently establish the specific balance-recovery event.", "mask_relation": "matches_selected_person_mask", "source_subject_refs": [ "entity_1" ] }, { "id": "action_4", "mentions": [ { "text": "walking", "ranges": [ [ 108, 115 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } }, { "text": "move", "ranges": [ [ 173, 177 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "type": "simple", "subject_refs": [ "entity_1" ], "object_refs": [], "confidence": "high", "temporal_status": "localized", "intervals": [ { "start_s": 0.0, "end_s": 51.00000000000001, "source_start_s": 49.51, "source_end_s": 100.51 } ], "evidence": "Across the sampled sequence, the masked person repeatedly changes foot placement and advances slowly toward the right-hand end of the slackline.", "mask_relation": "matches_selected_person_mask", "source_subject_refs": [ "entity_1" ] }, { "id": "action_5", "mentions": [ { "text": "extend", "ranges": [ [ 130, 136 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "type": "simple", "subject_refs": [ "entity_1" ], "object_refs": [ "entity_3" ], "confidence": "high", "temporal_status": "localized", "intervals": [ { "start_s": 0.0, "end_s": 2.0, "source_start_s": 49.51, "source_end_s": 51.51 }, { "start_s": 3.0, "end_s": 7.0, "source_start_s": 52.51, "source_end_s": 56.51 }, { "start_s": 10.0, "end_s": 18.0, "source_start_s": 59.51, "source_end_s": 67.50999999999999 }, { "start_s": 20.0, "end_s": 24.0, "source_start_s": 69.50999999999999, "source_end_s": 73.50999999999999 }, { "start_s": 26.0, "end_s": 31.0, "source_start_s": 75.50999999999999, "source_end_s": 80.50999999999999 }, { "start_s": 32.0, "end_s": 36.0, "source_start_s": 81.50999999999999, "source_end_s": 85.50999999999999 }, { "start_s": 40.0, "end_s": 46.0, "source_start_s": 89.50999999999999, "source_end_s": 95.50999999999999 }, { "start_s": 48.0, "end_s": 51.00000000000001, "source_start_s": 97.50999999999999, "source_end_s": 100.51 } ], "evidence": "The clean frames repeatedly show the masked person holding or moving one or both arms outward and upward for balance while progressing along the line.", "mask_relation": "matches_selected_person_mask", "source_subject_refs": [ "entity_1" ] } ], "notes": [ "The supplied mask consistently identifies the single person throughout the sampled sequence.", "No separate source masks were supplied for the black object, arms, or slackline; the arms are included within the person's mask rather than provided as an independent body-part mask.", "The black object and its toss are not clearly visible in the sampled clean frames, so the toss has no temporal interval.", "The description of the pace as slow and controlled is not annotated as a separate action.", "The final supported intervals end at the clip boundary of 51.0 seconds and do not extend beyond the original source interval." ] }, "model_output_raw": { "caption": "The person tosses the black object away to their left, briefly squats to regain balance, and then continues walking forward. They extend their arms out to the sides as they move steadily toward the far end of the slackline. Their pace remains slow and controlled.", "entity_groups": [ { "id": "entity_1", "label": "person", "spans": [ { "text": "The person", "range": [ 0, 10 ] }, { "text": "their", "range": [ 43, 48 ] }, { "text": "They", "range": [ 125, 129 ] }, { "text": "their", "range": [ 137, 142 ] }, { "text": "they", "range": [ 168, 172 ] }, { "text": "Their", "range": [ 224, 229 ] } ], "source_mask": "selected_person_mask" }, { "id": "entity_2", "label": "black object", "spans": [ { "text": "the black object", "range": [ 18, 34 ] } ], "source_mask": null }, { "id": "entity_3", "label": "arms", "spans": [ { "text": "their arms", "range": [ 137, 147 ] } ], "source_mask": null }, { "id": "entity_4", "label": "slackline", "spans": [ { "text": "the slackline", "range": [ 209, 222 ] } ], "source_mask": null } ], "actions": [ { "id": "action_1", "mentions": [ { "text": "tosses", "ranges": [ [ 11, 17 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "type": "simple", "subject_refs": [ "entity_1" ], "object_refs": [ "entity_2" ], "confidence": "high", "mask_relation": "matches_selected_person_mask", "temporal_status": "uncertain", "intervals": [], "evidence": "The selected person is visible and mask-linked, but no sampled clean frame clearly shows the black object being released or moving away." }, { "id": "action_2", "mentions": [ { "text": "squats", "ranges": [ [ 63, 69 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "type": "posture", "subject_refs": [ "entity_1" ], "object_refs": [], "confidence": "high", "mask_relation": "matches_selected_person_mask", "temporal_status": "localized", "intervals": [ { "start_s": 1.0, "end_s": 2.0, "source_start_s": 50.51, "source_end_s": 51.51 } ], "evidence": "At clip time 1.0 seconds, the masked person has markedly bent knees and a lowered body while balancing." }, { "id": "action_3", "mentions": [ { "text": "regain balance", "ranges": [ [ 73, 87 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "type": "other", "subject_refs": [ "entity_1" ], "object_refs": [], "confidence": "high", "mask_relation": "matches_selected_person_mask", "temporal_status": "uncertain", "intervals": [], "evidence": "Arm and torso adjustments are visible, but sparse still frames do not independently establish the specific balance-recovery event." }, { "id": "action_4", "mentions": [ { "text": "walking", "ranges": [ [ 108, 115 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } }, { "text": "move", "ranges": [ [ 173, 177 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "type": "simple", "subject_refs": [ "entity_1" ], "object_refs": [], "confidence": "high", "mask_relation": "matches_selected_person_mask", "temporal_status": "localized", "intervals": [ { "start_s": 0.0, "end_s": 51.0, "source_start_s": 49.51, "source_end_s": 100.51 } ], "evidence": "Across the sampled sequence, the masked person repeatedly changes foot placement and advances slowly toward the right-hand end of the slackline." }, { "id": "action_5", "mentions": [ { "text": "extend", "ranges": [ [ 130, 136 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "type": "simple", "subject_refs": [ "entity_1" ], "object_refs": [ "entity_3" ], "confidence": "high", "mask_relation": "matches_selected_person_mask", "temporal_status": "localized", "intervals": [ { "start_s": 0.0, "end_s": 2.0, "source_start_s": 49.51, "source_end_s": 51.51 }, { "start_s": 3.0, "end_s": 7.0, "source_start_s": 52.51, "source_end_s": 56.51 }, { "start_s": 10.0, "end_s": 18.0, "source_start_s": 59.51, "source_end_s": 67.51 }, { "start_s": 20.0, "end_s": 24.0, "source_start_s": 69.51, "source_end_s": 73.51 }, { "start_s": 26.0, "end_s": 31.0, "source_start_s": 75.51, "source_end_s": 80.51 }, { "start_s": 32.0, "end_s": 36.0, "source_start_s": 81.51, "source_end_s": 85.51 }, { "start_s": 40.0, "end_s": 46.0, "source_start_s": 89.51, "source_end_s": 95.51 }, { "start_s": 48.0, "end_s": 51.0, "source_start_s": 97.51, "source_end_s": 100.51 } ], "evidence": "The clean frames repeatedly show the masked person holding or moving one or both arms outward and upward for balance while progressing along the line." } ], "notes": [ "The supplied mask consistently identifies the single person throughout the sampled sequence.", "No separate source masks were supplied for the black object, arms, or slackline; the arms are included within the person's mask rather than provided as an independent body-part mask.", "The black object and its toss are not clearly visible in the sampled clean frames, so the toss has no temporal interval.", "The description of the pace as slow and controlled is not annotated as a separate action.", "The final supported intervals end at the clip boundary of 51.0 seconds and do not extend beyond the original source interval." ] }, "normalization_repairs": 0 }, { "source_video_id": "v_-4Q_zG9EChY", "caption_count_category": 4, "clip": { "caption_index": 3, "caption": "Upon reaching the far end of the slackline near a wooden post, the person stops walking and plants their feet. They carefully pivot 180 degrees on the narrow line, using their arms for stabilization, to face the direction they came from.", "source_interval_sec": [ 100.51, 114.51 ], "loom_clip_id": "v_-4Q_zG9EChY_3012_3432", "clip_frame_interval_4fps": [ 3012, 3432 ], "sample_count_1fps": 14, "sampled_frames": [ { "frame": "00000", "timestamp_sec": 100.51, "offset_sec": 0, "rgb_path": "assets/cat_04/v_-4Q_zG9EChY_3012_3432/0000s_00000_rgb.jpg", "mask_path": "assets/cat_04/v_-4Q_zG9EChY_3012_3432/0000s_00000_mask.png", "mask_status": "available" }, { "frame": "00028", "timestamp_sec": 101.51, "offset_sec": 1, "rgb_path": "assets/cat_04/v_-4Q_zG9EChY_3012_3432/0001s_00028_rgb.jpg", "mask_path": "assets/cat_04/v_-4Q_zG9EChY_3012_3432/0001s_00028_mask.png", "mask_status": "available" }, { "frame": "00060", "timestamp_sec": 102.51, "offset_sec": 2, "rgb_path": "assets/cat_04/v_-4Q_zG9EChY_3012_3432/0002s_00060_rgb.jpg", "mask_path": "assets/cat_04/v_-4Q_zG9EChY_3012_3432/0002s_00060_mask.png", "mask_status": "available" }, { "frame": "00088", "timestamp_sec": 103.51, "offset_sec": 3, "rgb_path": "assets/cat_04/v_-4Q_zG9EChY_3012_3432/0003s_00088_rgb.jpg", "mask_path": "assets/cat_04/v_-4Q_zG9EChY_3012_3432/0003s_00088_mask.png", "mask_status": "available" }, { "frame": "00120", "timestamp_sec": 104.51, "offset_sec": 4, "rgb_path": "assets/cat_04/v_-4Q_zG9EChY_3012_3432/0004s_00120_rgb.jpg", "mask_path": "assets/cat_04/v_-4Q_zG9EChY_3012_3432/0004s_00120_mask.png", "mask_status": "available" }, { "frame": "00148", "timestamp_sec": 105.51, "offset_sec": 5, "rgb_path": "assets/cat_04/v_-4Q_zG9EChY_3012_3432/0005s_00148_rgb.jpg", "mask_path": "assets/cat_04/v_-4Q_zG9EChY_3012_3432/0005s_00148_mask.png", "mask_status": "available" }, { "frame": "00180", "timestamp_sec": 106.51, "offset_sec": 6, "rgb_path": "assets/cat_04/v_-4Q_zG9EChY_3012_3432/0006s_00180_rgb.jpg", "mask_path": "assets/cat_04/v_-4Q_zG9EChY_3012_3432/0006s_00180_mask.png", "mask_status": "available" }, { "frame": "00208", "timestamp_sec": 107.51, "offset_sec": 7, "rgb_path": "assets/cat_04/v_-4Q_zG9EChY_3012_3432/0007s_00208_rgb.jpg", "mask_path": "assets/cat_04/v_-4Q_zG9EChY_3012_3432/0007s_00208_mask.png", "mask_status": "available" }, { "frame": "00240", "timestamp_sec": 108.51, "offset_sec": 8, "rgb_path": "assets/cat_04/v_-4Q_zG9EChY_3012_3432/0008s_00240_rgb.jpg", "mask_path": "assets/cat_04/v_-4Q_zG9EChY_3012_3432/0008s_00240_mask.png", "mask_status": "available" }, { "frame": "00268", "timestamp_sec": 109.51, "offset_sec": 9, "rgb_path": "assets/cat_04/v_-4Q_zG9EChY_3012_3432/0009s_00268_rgb.jpg", "mask_path": "assets/cat_04/v_-4Q_zG9EChY_3012_3432/0009s_00268_mask.png", "mask_status": "available" }, { "frame": "00300", "timestamp_sec": 110.51, "offset_sec": 10, "rgb_path": "assets/cat_04/v_-4Q_zG9EChY_3012_3432/0010s_00300_rgb.jpg", "mask_path": "assets/cat_04/v_-4Q_zG9EChY_3012_3432/0010s_00300_mask.png", "mask_status": "available" }, { "frame": "00328", "timestamp_sec": 111.51, "offset_sec": 11, "rgb_path": "assets/cat_04/v_-4Q_zG9EChY_3012_3432/0011s_00328_rgb.jpg", "mask_path": "assets/cat_04/v_-4Q_zG9EChY_3012_3432/0011s_00328_mask.png", "mask_status": "available" }, { "frame": "00360", "timestamp_sec": 112.51, "offset_sec": 12, "rgb_path": "assets/cat_04/v_-4Q_zG9EChY_3012_3432/0012s_00360_rgb.jpg", "mask_path": "assets/cat_04/v_-4Q_zG9EChY_3012_3432/0012s_00360_mask.png", "mask_status": "available" }, { "frame": "00388", "timestamp_sec": 113.51, "offset_sec": 13, "rgb_path": "assets/cat_04/v_-4Q_zG9EChY_3012_3432/0013s_00388_rgb.jpg", "mask_path": "assets/cat_04/v_-4Q_zG9EChY_3012_3432/0013s_00388_mask.png", "mask_status": "available" } ], "obj_id": [ 0 ], "anno_id": [ 29 ], "fps_inferred_from_clip_interval": 30.0 }, "usage": { "input_tokens": 67516, "input_tokens_details": { "cache_write_tokens": 67513, "cached_tokens": 0 }, "output_tokens": 4882, "output_tokens_details": { "reasoning_tokens": 3181 }, "total_tokens": 72398 }, "model": "gpt-5.6-sol", "effort": "medium", "custom_id": "cat04_cap03_v_-4Q_zG9EChY_3012_3432", "annotation": { "caption": "Upon reaching the far end of the slackline near a wooden post, the person stops walking and plants their feet. They carefully pivot 180 degrees on the narrow line, using their arms for stabilization, to face the direction they came from.", "entity_groups": [ { "id": "E1", "label": "person", "spans": [ { "text": "the person", "range": [ 63, 73 ] }, { "text": "their", "range": [ 99, 104 ] }, { "text": "They", "range": [ 111, 115 ] }, { "text": "their", "range": [ 170, 175 ] }, { "text": "they", "range": [ 222, 226 ] } ], "source_mask_link": "selected_person" }, { "id": "E2", "label": "slackline", "spans": [ { "text": "the slackline", "range": [ 29, 42 ] }, { "text": "the narrow line", "range": [ 147, 162 ] } ], "source_mask_link": null }, { "id": "E3", "label": "wooden post", "spans": [ { "text": "a wooden post", "range": [ 48, 61 ] } ], "source_mask_link": null }, { "id": "E4", "label": "feet", "spans": [ { "text": "their feet", "range": [ 99, 109 ] } ], "source_mask_link": null }, { "id": "E5", "label": "arms", "spans": [ { "text": "their arms", "range": [ 170, 180 ] } ], "source_mask_link": null } ], "actions": [ { "id": "A1", "mentions": [ { "text": "reaching", "ranges": [ [ 5, 13 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "type": "simple", "subject_refs": [ "E1" ], "object_refs": [], "confidence": "high", "temporal_status": "uncertain", "intervals": [], "evidence": "The masked person is already positioned at an end region of the slackline in the samples, but no sampled frame clearly shows the arrival itself.", "mask_relation": "subject_matches_selected_person_mask", "source_subject_refs": "E1" }, { "id": "A2", "mentions": [ { "text": "walking", "ranges": [ [ 80, 87 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } }, { "text": "came", "ranges": [ [ 227, 231 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "type": "simple", "subject_refs": [ "E1" ], "object_refs": [], "confidence": "high", "temporal_status": "uncertain", "intervals": [], "evidence": "The two mentions describe the person's preceding approach along the line, but the sparse frames do not clearly show a sustained walking sequence.", "mask_relation": "subject_matches_selected_person_mask", "source_subject_refs": "E1" }, { "id": "A3", "mentions": [ { "text": "plants", "ranges": [ [ 92, 98 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "type": "simple", "subject_refs": [ "E1" ], "object_refs": [ "E4" ], "confidence": "high", "temporal_status": "uncertain", "intervals": [], "evidence": "The masked person's feet remain supported on the line, but the discrete act of planting them cannot be isolated in the sampled frames.", "mask_relation": "subject_matches_selected_person_mask", "source_subject_refs": "E1" }, { "id": "A4", "mentions": [ { "text": "pivot", "ranges": [ [ 126, 131 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "type": "simple", "subject_refs": [ "E1" ], "object_refs": [], "confidence": "high", "temporal_status": "localized", "intervals": [ { "start_s": 0.0, "end_s": 14.0, "source_start_s": 100.51, "source_end_s": 114.51 } ], "evidence": "Across the sampled sequence, the masked person slowly changes body orientation while remaining on the slackline; the changing torso, leg, and arm configurations support an ongoing careful pivot through the clip boundary.", "mask_relation": "subject_matches_selected_person_mask", "source_subject_refs": "E1" }, { "id": "A5", "mentions": [ { "text": "using", "ranges": [ [ 164, 169 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "type": "simple", "subject_refs": [ "E1" ], "object_refs": [ "E5" ], "confidence": "high", "temporal_status": "localized", "intervals": [ { "start_s": 0.0, "end_s": 14.0, "source_start_s": 100.51, "source_end_s": 114.51 } ], "evidence": "In every sampled frame, the masked person raises or extends one or both arms while balancing and turning on the line.", "mask_relation": "subject_matches_selected_person_mask", "source_subject_refs": "E1" }, { "id": "A6", "mentions": [ { "text": "face", "ranges": [ [ 203, 207 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "type": "posture", "subject_refs": [ "E1" ], "object_refs": [], "confidence": "high", "temporal_status": "uncertain", "intervals": [], "evidence": "The person's orientation changes, but the sampled frames do not establish which visible direction is specifically the direction from which the person previously came.", "mask_relation": "subject_matches_selected_person_mask", "source_subject_refs": "E1" } ], "notes": [ "The supplied masklet consistently identifies the sole person described by the caption, so all listed action subjects match the selected-person mask.", "No separate masks were supplied for the slackline, wooden post, feet, or arms; their source-mask links are therefore null.", "The aspectual verb \"stops\" is excluded from the walking action span under the RDCap V5 rules." ] }, "model_output_raw": { "caption": "Upon reaching the far end of the slackline near a wooden post, the person stops walking and plants their feet. They carefully pivot 180 degrees on the narrow line, using their arms for stabilization, to face the direction they came from.", "entity_groups": [ { "id": "E1", "entity": "person", "spans": [ { "text": "the person", "range": [ 63, 73 ] }, { "text": "their", "range": [ 99, 104 ] }, { "text": "They", "range": [ 111, 115 ] }, { "text": "their", "range": [ 170, 175 ] }, { "text": "they", "range": [ 222, 226 ] } ], "source_mask_link": "selected_person_masklet" }, { "id": "E2", "entity": "slackline", "spans": [ { "text": "the slackline", "range": [ 29, 42 ] }, { "text": "the narrow line", "range": [ 147, 162 ] } ], "source_mask_link": null }, { "id": "E3", "entity": "wooden post", "spans": [ { "text": "a wooden post", "range": [ 48, 61 ] } ], "source_mask_link": null }, { "id": "E4", "entity": "feet", "spans": [ { "text": "their feet", "range": [ 99, 109 ] } ], "source_mask_link": null }, { "id": "E5", "entity": "arms", "spans": [ { "text": "their arms", "range": [ 170, 180 ] } ], "source_mask_link": null } ], "actions": [ { "id": "A1", "mentions": [ { "text": "reaching", "ranges": [ [ 5, 13 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "type": "simple", "subject_ref": "E1", "object_refs": [], "related_entity_refs": [ "E2" ], "confidence": "high", "temporal_status": "uncertain", "segments": [], "evidence": "The masked person is already positioned at an end region of the slackline in the samples, but no sampled frame clearly shows the arrival itself.", "mask_relation": "subject_matches_selected_person_mask" }, { "id": "A2", "mentions": [ { "text": "walking", "ranges": [ [ 80, 87 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } }, { "text": "came", "ranges": [ [ 227, 231 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "type": "simple", "subject_ref": "E1", "object_refs": [], "related_entity_refs": [ "E2" ], "confidence": "high", "temporal_status": "uncertain", "segments": [], "evidence": "The two mentions describe the person's preceding approach along the line, but the sparse frames do not clearly show a sustained walking sequence.", "mask_relation": "subject_matches_selected_person_mask" }, { "id": "A3", "mentions": [ { "text": "plants", "ranges": [ [ 92, 98 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "type": "simple", "subject_ref": "E1", "object_refs": [ "E4" ], "confidence": "high", "temporal_status": "uncertain", "segments": [], "evidence": "The masked person's feet remain supported on the line, but the discrete act of planting them cannot be isolated in the sampled frames.", "mask_relation": "subject_matches_selected_person_mask" }, { "id": "A4", "mentions": [ { "text": "pivot", "ranges": [ [ 126, 131 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "type": "simple", "subject_ref": "E1", "object_refs": [], "related_entity_refs": [ "E2" ], "confidence": "high", "temporal_status": "localized", "segments": [ { "start_s": 0.0, "end_s": 14.0, "source_start_s": 100.51, "source_end_s": 114.51 } ], "evidence": "Across the sampled sequence, the masked person slowly changes body orientation while remaining on the slackline; the changing torso, leg, and arm configurations support an ongoing careful pivot through the clip boundary.", "mask_relation": "subject_matches_selected_person_mask" }, { "id": "A5", "mentions": [ { "text": "using", "ranges": [ [ 164, 169 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "type": "simple", "subject_ref": "E1", "object_refs": [ "E5" ], "confidence": "high", "temporal_status": "localized", "segments": [ { "start_s": 0.0, "end_s": 14.0, "source_start_s": 100.51, "source_end_s": 114.51 } ], "evidence": "In every sampled frame, the masked person raises or extends one or both arms while balancing and turning on the line.", "mask_relation": "subject_matches_selected_person_mask" }, { "id": "A6", "mentions": [ { "text": "face", "ranges": [ [ 203, 207 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "type": "posture", "subject_ref": "E1", "object_refs": [], "confidence": "high", "temporal_status": "uncertain", "segments": [], "evidence": "The person's orientation changes, but the sampled frames do not establish which visible direction is specifically the direction from which the person previously came.", "mask_relation": "subject_matches_selected_person_mask" } ], "notes": [ "The supplied masklet consistently identifies the sole person described by the caption, so all listed action subjects match the selected-person mask.", "No separate masks were supplied for the slackline, wooden post, feet, or arms; their source-mask links are therefore null.", "The aspectual verb \"stops\" is excluded from the walking action span under the RDCap V5 rules." ] }, "normalization_repairs": 0 }, { "source_video_id": "v_-4Q_zG9EChY", "caption_count_category": 4, "clip": { "caption_index": 4, "caption": "Now facing the opposite direction, the person walks back along the slackline toward the starting point. They maintain their balance with their arms outstretched. After reaching the end, they jump off the line onto the ground.", "source_interval_sec": [ 114.51, 137.07 ], "loom_clip_id": "v_-4Q_zG9EChY_3432_4108", "clip_frame_interval_4fps": [ 3432, 4108 ], "sample_count_1fps": 23, "sampled_frames": [ { "frame": "00000", "timestamp_sec": 114.51, "offset_sec": 0, "rgb_path": "assets/cat_04/v_-4Q_zG9EChY_3432_4108/0000s_00000_rgb.jpg", "mask_path": "assets/cat_04/v_-4Q_zG9EChY_3432_4108/0000s_00000_mask.png", "mask_status": "available" }, { "frame": "00028", "timestamp_sec": 115.51, "offset_sec": 1, "rgb_path": "assets/cat_04/v_-4Q_zG9EChY_3432_4108/0001s_00028_rgb.jpg", "mask_path": "assets/cat_04/v_-4Q_zG9EChY_3432_4108/0001s_00028_mask.png", "mask_status": "available" }, { "frame": "00060", "timestamp_sec": 116.51, "offset_sec": 2, "rgb_path": "assets/cat_04/v_-4Q_zG9EChY_3432_4108/0002s_00060_rgb.jpg", "mask_path": "assets/cat_04/v_-4Q_zG9EChY_3432_4108/0002s_00060_mask.png", "mask_status": "available" }, { "frame": "00088", "timestamp_sec": 117.51, "offset_sec": 3, "rgb_path": "assets/cat_04/v_-4Q_zG9EChY_3432_4108/0003s_00088_rgb.jpg", "mask_path": "assets/cat_04/v_-4Q_zG9EChY_3432_4108/0003s_00088_mask.png", "mask_status": "available" }, { "frame": "00120", "timestamp_sec": 118.51, "offset_sec": 4, "rgb_path": "assets/cat_04/v_-4Q_zG9EChY_3432_4108/0004s_00120_rgb.jpg", "mask_path": "assets/cat_04/v_-4Q_zG9EChY_3432_4108/0004s_00120_mask.png", "mask_status": "available" }, { "frame": "00148", "timestamp_sec": 119.51, "offset_sec": 5, "rgb_path": "assets/cat_04/v_-4Q_zG9EChY_3432_4108/0005s_00148_rgb.jpg", "mask_path": "assets/cat_04/v_-4Q_zG9EChY_3432_4108/0005s_00148_mask.png", "mask_status": "available" }, { "frame": "00180", "timestamp_sec": 120.51, "offset_sec": 6, "rgb_path": "assets/cat_04/v_-4Q_zG9EChY_3432_4108/0006s_00180_rgb.jpg", "mask_path": "assets/cat_04/v_-4Q_zG9EChY_3432_4108/0006s_00180_mask.png", "mask_status": "available" }, { "frame": "00208", "timestamp_sec": 121.51, "offset_sec": 7, "rgb_path": "assets/cat_04/v_-4Q_zG9EChY_3432_4108/0007s_00208_rgb.jpg", "mask_path": "assets/cat_04/v_-4Q_zG9EChY_3432_4108/0007s_00208_mask.png", "mask_status": "available" }, { "frame": "00240", "timestamp_sec": 122.51, "offset_sec": 8, "rgb_path": "assets/cat_04/v_-4Q_zG9EChY_3432_4108/0008s_00240_rgb.jpg", "mask_path": "assets/cat_04/v_-4Q_zG9EChY_3432_4108/0008s_00240_mask.png", "mask_status": "available" }, { "frame": "00268", "timestamp_sec": 123.51, "offset_sec": 9, "rgb_path": "assets/cat_04/v_-4Q_zG9EChY_3432_4108/0009s_00268_rgb.jpg", "mask_path": "assets/cat_04/v_-4Q_zG9EChY_3432_4108/0009s_00268_mask.png", "mask_status": "available" }, { "frame": "00300", "timestamp_sec": 124.51, "offset_sec": 10, "rgb_path": "assets/cat_04/v_-4Q_zG9EChY_3432_4108/0010s_00300_rgb.jpg", "mask_path": "assets/cat_04/v_-4Q_zG9EChY_3432_4108/0010s_00300_mask.png", "mask_status": "available" }, { "frame": "00328", "timestamp_sec": 125.51, "offset_sec": 11, "rgb_path": "assets/cat_04/v_-4Q_zG9EChY_3432_4108/0011s_00328_rgb.jpg", "mask_path": "assets/cat_04/v_-4Q_zG9EChY_3432_4108/0011s_00328_mask.png", "mask_status": "available" }, { "frame": "00360", "timestamp_sec": 126.51, "offset_sec": 12, "rgb_path": "assets/cat_04/v_-4Q_zG9EChY_3432_4108/0012s_00360_rgb.jpg", "mask_path": "assets/cat_04/v_-4Q_zG9EChY_3432_4108/0012s_00360_mask.png", "mask_status": "available" }, { "frame": "00388", "timestamp_sec": 127.51, "offset_sec": 13, "rgb_path": "assets/cat_04/v_-4Q_zG9EChY_3432_4108/0013s_00388_rgb.jpg", "mask_path": "assets/cat_04/v_-4Q_zG9EChY_3432_4108/0013s_00388_mask.png", "mask_status": "available" }, { "frame": "00420", "timestamp_sec": 128.51, "offset_sec": 14, "rgb_path": "assets/cat_04/v_-4Q_zG9EChY_3432_4108/0014s_00420_rgb.jpg", "mask_path": "assets/cat_04/v_-4Q_zG9EChY_3432_4108/0014s_00420_mask.png", "mask_status": "available" }, { "frame": "00448", "timestamp_sec": 129.51, "offset_sec": 15, "rgb_path": "assets/cat_04/v_-4Q_zG9EChY_3432_4108/0015s_00448_rgb.jpg", "mask_path": "assets/cat_04/v_-4Q_zG9EChY_3432_4108/0015s_00448_mask.png", "mask_status": "available" }, { "frame": "00480", "timestamp_sec": 130.51, "offset_sec": 16, "rgb_path": "assets/cat_04/v_-4Q_zG9EChY_3432_4108/0016s_00480_rgb.jpg", "mask_path": "assets/cat_04/v_-4Q_zG9EChY_3432_4108/0016s_00480_mask.png", "mask_status": "available" }, { "frame": "00508", "timestamp_sec": 131.51, "offset_sec": 17, "rgb_path": "assets/cat_04/v_-4Q_zG9EChY_3432_4108/0017s_00508_rgb.jpg", "mask_path": "assets/cat_04/v_-4Q_zG9EChY_3432_4108/0017s_00508_mask.png", "mask_status": "available" }, { "frame": "00540", "timestamp_sec": 132.51, "offset_sec": 18, "rgb_path": "assets/cat_04/v_-4Q_zG9EChY_3432_4108/0018s_00540_rgb.jpg", "mask_path": "assets/cat_04/v_-4Q_zG9EChY_3432_4108/0018s_00540_mask.png", "mask_status": "available" }, { "frame": "00568", "timestamp_sec": 133.51, "offset_sec": 19, "rgb_path": "assets/cat_04/v_-4Q_zG9EChY_3432_4108/0019s_00568_rgb.jpg", "mask_path": "assets/cat_04/v_-4Q_zG9EChY_3432_4108/0019s_00568_mask.png", "mask_status": "available" }, { "frame": "00600", "timestamp_sec": 134.51, "offset_sec": 20, "rgb_path": "assets/cat_04/v_-4Q_zG9EChY_3432_4108/0020s_00600_rgb.jpg", "mask_path": "assets/cat_04/v_-4Q_zG9EChY_3432_4108/0020s_00600_mask.png", "mask_status": "available" }, { "frame": "00628", "timestamp_sec": 135.51, "offset_sec": 21, "rgb_path": "assets/cat_04/v_-4Q_zG9EChY_3432_4108/0021s_00628_rgb.jpg", "mask_path": "assets/cat_04/v_-4Q_zG9EChY_3432_4108/0021s_00628_mask.png", "mask_status": "available" }, { "frame": "00660", "timestamp_sec": 136.51, "offset_sec": 22, "rgb_path": "assets/cat_04/v_-4Q_zG9EChY_3432_4108/0022s_00660_rgb.jpg", "mask_path": "assets/cat_04/v_-4Q_zG9EChY_3432_4108/0022s_00660_mask.png", "mask_status": "available" } ], "obj_id": [ 0 ], "anno_id": [ 30 ], "fps_inferred_from_clip_interval": 29.9645 }, "usage": { "input_tokens": 101055, "input_tokens_details": { "cache_write_tokens": 101052, "cached_tokens": 0 }, "output_tokens": 4564, "output_tokens_details": { "reasoning_tokens": 2907 }, "total_tokens": 105619 }, "model": "gpt-5.6-sol", "effort": "medium", "custom_id": "cat04_cap04_v_-4Q_zG9EChY_3432_4108", "annotation": { "caption": "Now facing the opposite direction, the person walks back along the slackline toward the starting point. They maintain their balance with their arms outstretched. After reaching the end, they jump off the line onto the ground.", "entity_groups": [ { "id": "person_1", "label": "person", "spans": [ { "text": "the person", "range": [ 35, 45 ] }, { "text": "They", "range": [ 104, 108 ] }, { "text": "their", "range": [ 118, 123 ] }, { "text": "their", "range": [ 137, 142 ] }, { "text": "they", "range": [ 186, 190 ] } ], "source_mask_link": "selected_person" }, { "id": "slackline_1", "label": "slackline", "spans": [ { "text": "the slackline", "range": [ 63, 76 ] }, { "text": "the line", "range": [ 200, 208 ] } ], "source_mask_link": null }, { "id": "arms_1", "label": "arms", "spans": [ { "text": "their arms", "range": [ 137, 147 ] } ], "source_mask_link": null }, { "id": "ground_1", "label": "ground", "spans": [ { "text": "the ground", "range": [ 214, 224 ] } ], "source_mask_link": null } ], "actions": [ { "id": "action_1", "mentions": [ { "text": "facing", "ranges": [ [ 4, 10 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "type": "posture", "subject_refs": [ "person_1" ], "object_refs": [], "confidence": "medium", "temporal_status": "supported_sparse", "intervals": [ { "start_s": 0.0, "end_s": 21.0, "source_start_s": 114.51, "source_end_s": 135.51 } ], "evidence": "Samples 0–20 show the masked person consistently oriented toward the slackline's right-hand endpoint; the prior orientation needed to verify 'opposite' is not shown.", "mask_relation": "subject matches the supplied selected-person mask", "source_subject_refs": "person_1" }, { "id": "action_2", "mentions": [ { "text": "walks", "ranges": [ [ 46, 51 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "type": "simple", "subject_refs": [ "person_1" ], "object_refs": [ "slackline_1" ], "confidence": "high", "temporal_status": "supported_sparse", "intervals": [ { "start_s": 0.0, "end_s": 20.0, "source_start_s": 114.51, "source_end_s": 134.51 } ], "evidence": "Across samples 0–19, the masked person advances step by step along the slackline toward its right-hand end.", "mask_relation": "subject matches the supplied selected-person mask", "source_subject_refs": "person_1" }, { "id": "action_3", "mentions": [ { "text": "maintain their balance", "ranges": [ [ 109, 131 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "type": "other", "subject_refs": [ "person_1" ], "object_refs": [], "confidence": "high", "temporal_status": "supported_sparse", "intervals": [ { "start_s": 0.0, "end_s": 21.0, "source_start_s": 114.51, "source_end_s": 135.51 } ], "evidence": "Samples 0–20 show the masked person balancing on the narrow line with repeated arm and torso adjustments.", "mask_relation": "subject matches the supplied selected-person mask", "source_subject_refs": "person_1" }, { "id": "action_4", "mentions": [ { "text": "reaching", "ranges": [ [ 168, 176 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "type": "simple", "subject_refs": [ "person_1" ], "object_refs": [ "slackline_1" ], "confidence": "high", "temporal_status": "supported_sparse", "intervals": [ { "start_s": 19.0, "end_s": 21.0, "source_start_s": 133.51, "source_end_s": 135.51 } ], "evidence": "In samples 19–20, the masked person arrives at the slackline's right-hand endpoint.", "mask_relation": "subject matches the supplied selected-person mask", "source_subject_refs": "person_1" }, { "id": "action_5", "mentions": [ { "text": "jump", "ranges": [ [ 191, 195 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "type": "simple", "subject_refs": [ "person_1" ], "object_refs": [ "slackline_1", "ground_1" ], "confidence": "medium", "temporal_status": "inferred_from_sparse_transition", "intervals": [ { "start_s": 20.0, "end_s": 22.0, "source_start_s": 134.51, "source_end_s": 136.51 } ], "evidence": "Sample 20 shows the masked person at the line's endpoint, while sample 21 shows him off the line on the ground in a landing or departure posture; no sampled frame captures the airborne phase.", "mask_relation": "subject matches the supplied selected-person mask", "source_subject_refs": "person_1" } ], "notes": [ "The supplied masklet consistently identifies the caption's person.", "No separate masks were supplied for the slackline, arms, or ground.", "The possessive references 'their' are grouped with the person; 'their arms' forms a separate body-part entity group.", "The spatial expressions 'the opposite direction,' 'the starting point,' and 'the end,' and the abstract noun 'balance,' are not entity groups.", "The arm state 'outstretched' was not added as a separate action because its grammatical subject is the arms, which have no corresponding supplied instance mask.", "In 'jump off the line,' 'off' introduces the source phrase 'off the line' and is excluded from the action span as an ordinary directional preposition." ] }, "model_output_raw": { "caption": "Now facing the opposite direction, the person walks back along the slackline toward the starting point. They maintain their balance with their arms outstretched. After reaching the end, they jump off the line onto the ground.", "entity_groups": [ { "id": "person_1", "label": "person", "spans": [ { "text": "the person", "range": [ 35, 45 ] }, { "text": "They", "range": [ 104, 108 ] }, { "text": "their", "range": [ 118, 123 ] }, { "text": "their", "range": [ 137, 142 ] }, { "text": "they", "range": [ 185, 189 ] } ], "source_mask": "selected_person_masklet" }, { "id": "slackline_1", "label": "slackline", "spans": [ { "text": "the slackline", "range": [ 63, 76 ] }, { "text": "the line", "range": [ 199, 207 ] } ], "source_mask": null }, { "id": "arms_1", "label": "arms", "spans": [ { "text": "their arms", "range": [ 137, 147 ] } ], "source_mask": null }, { "id": "ground_1", "label": "ground", "spans": [ { "text": "the ground", "range": [ 213, 223 ] } ], "source_mask": null } ], "actions": [ { "id": "action_1", "mentions": [ { "text": "facing", "occurrence": 1, "ranges": [ [ 4, 10 ] ], "particle": { "text": null, "occurrence": 1 } } ], "type": "posture", "subject_ref": "person_1", "object_refs": [], "confidence": "medium", "mask_relation": "subject matches the supplied selected-person mask", "temporal_status": "supported_sparse", "intervals": [ { "clip_relative_s": [ 0.0, 21.0 ], "source_video_s": [ 114.51, 135.51 ] } ], "evidence": "Samples 0–20 show the masked person consistently oriented toward the slackline's right-hand endpoint; the prior orientation needed to verify 'opposite' is not shown." }, { "id": "action_2", "mentions": [ { "text": "walks", "occurrence": 1, "ranges": [ [ 46, 51 ] ], "particle": { "text": null, "occurrence": 1 } } ], "type": "simple", "subject_ref": "person_1", "object_refs": [ "slackline_1" ], "confidence": "high", "mask_relation": "subject matches the supplied selected-person mask", "temporal_status": "supported_sparse", "intervals": [ { "clip_relative_s": [ 0.0, 20.0 ], "source_video_s": [ 114.51, 134.51 ] } ], "evidence": "Across samples 0–19, the masked person advances step by step along the slackline toward its right-hand end." }, { "id": "action_3", "mentions": [ { "text": "maintain their balance", "occurrence": 1, "ranges": [ [ 109, 131 ] ], "particle": { "text": null, "occurrence": 1 } } ], "type": "other", "subject_ref": "person_1", "object_refs": [], "confidence": "high", "mask_relation": "subject matches the supplied selected-person mask", "temporal_status": "supported_sparse", "intervals": [ { "clip_relative_s": [ 0.0, 21.0 ], "source_video_s": [ 114.51, 135.51 ] } ], "evidence": "Samples 0–20 show the masked person balancing on the narrow line with repeated arm and torso adjustments." }, { "id": "action_4", "mentions": [ { "text": "reaching", "occurrence": 1, "ranges": [ [ 167, 175 ] ], "particle": { "text": null, "occurrence": 1 } } ], "type": "simple", "subject_ref": "person_1", "object_refs": [ "slackline_1" ], "confidence": "high", "mask_relation": "subject matches the supplied selected-person mask", "temporal_status": "supported_sparse", "intervals": [ { "clip_relative_s": [ 19.0, 21.0 ], "source_video_s": [ 133.51, 135.51 ] } ], "evidence": "In samples 19–20, the masked person arrives at the slackline's right-hand endpoint." }, { "id": "action_5", "mentions": [ { "text": "jump", "occurrence": 1, "ranges": [ [ 190, 194 ] ], "particle": { "text": null, "occurrence": 1 } } ], "type": "simple", "subject_ref": "person_1", "object_refs": [ "slackline_1", "ground_1" ], "confidence": "medium", "mask_relation": "subject matches the supplied selected-person mask", "temporal_status": "inferred_from_sparse_transition", "intervals": [ { "clip_relative_s": [ 20.0, 22.0 ], "source_video_s": [ 134.51, 136.51 ] } ], "evidence": "Sample 20 shows the masked person at the line's endpoint, while sample 21 shows him off the line on the ground in a landing or departure posture; no sampled frame captures the airborne phase." } ], "notes": [ "The supplied masklet consistently identifies the caption's person.", "No separate masks were supplied for the slackline, arms, or ground.", "The possessive references 'their' are grouped with the person; 'their arms' forms a separate body-part entity group.", "The spatial expressions 'the opposite direction,' 'the starting point,' and 'the end,' and the abstract noun 'balance,' are not entity groups.", "The arm state 'outstretched' was not added as a separate action because its grammatical subject is the arms, which have no corresponding supplied instance mask.", "In 'jump off the line,' 'off' introduces the source phrase 'off the line' and is excluded from the action span as an ordinary directional preposition." ] }, "normalization_repairs": 5 }, { "source_video_id": "v_-3Ss9jSRbKk", "caption_count_category": 5, "clip": { "caption_index": 1, "caption": "A person wearing a black jacket, colorful leggings, and a black baseball cap is kneeling on the floor. She is facing forward and speaking, with her hands resting on her thighs. She leans her torso forward slightly.", "source_interval_sec": [ 3.44, 9.44 ], "loom_clip_id": "v_-3Ss9jSRbKk_103_283", "clip_frame_interval_4fps": [ 103, 283 ], "sample_count_1fps": 6, "sampled_frames": [ { "frame": "00000", "timestamp_sec": 3.44, "offset_sec": 0, "rgb_path": "assets/cat_05/v_-3Ss9jSRbKk_103_283/0000s_00000_rgb.jpg", "mask_path": "assets/cat_05/v_-3Ss9jSRbKk_103_283/0000s_00000_mask.png", "mask_status": "available" }, { "frame": "00028", "timestamp_sec": 4.44, "offset_sec": 1, "rgb_path": "assets/cat_05/v_-3Ss9jSRbKk_103_283/0001s_00028_rgb.jpg", "mask_path": "assets/cat_05/v_-3Ss9jSRbKk_103_283/0001s_00028_mask.png", "mask_status": "available" }, { "frame": "00060", "timestamp_sec": 5.44, "offset_sec": 2, "rgb_path": "assets/cat_05/v_-3Ss9jSRbKk_103_283/0002s_00060_rgb.jpg", "mask_path": "assets/cat_05/v_-3Ss9jSRbKk_103_283/0002s_00060_mask.png", "mask_status": "available" }, { "frame": "00088", "timestamp_sec": 6.44, "offset_sec": 3, "rgb_path": "assets/cat_05/v_-3Ss9jSRbKk_103_283/0003s_00088_rgb.jpg", "mask_path": "assets/cat_05/v_-3Ss9jSRbKk_103_283/0003s_00088_mask.png", "mask_status": "available" }, { "frame": "00120", "timestamp_sec": 7.44, "offset_sec": 4, "rgb_path": "assets/cat_05/v_-3Ss9jSRbKk_103_283/0004s_00120_rgb.jpg", "mask_path": "assets/cat_05/v_-3Ss9jSRbKk_103_283/0004s_00120_mask.png", "mask_status": "available" }, { "frame": "00148", "timestamp_sec": 8.44, "offset_sec": 5, "rgb_path": "assets/cat_05/v_-3Ss9jSRbKk_103_283/0005s_00148_rgb.jpg", "mask_path": "assets/cat_05/v_-3Ss9jSRbKk_103_283/0005s_00148_mask.png", "mask_status": "available" } ], "obj_id": [ 0 ], "anno_id": [ 22 ], "fps_inferred_from_clip_interval": 30.0 }, "usage": { "input_tokens": 45041, "input_tokens_details": { "cache_write_tokens": 45038, "cached_tokens": 0 }, "output_tokens": 3521, "output_tokens_details": { "reasoning_tokens": 2747 }, "total_tokens": 48562 }, "model": "gpt-5.6-sol", "effort": "medium", "custom_id": "cat05_cap01_v_-3Ss9jSRbKk_103_283", "annotation": { "caption": "A person wearing a black jacket, colorful leggings, and a black baseball cap is kneeling on the floor. She is facing forward and speaking, with her hands resting on her thighs. She leans her torso forward slightly.", "entity_groups": [ { "id": "entity_1", "label": "person", "spans": [ { "text": "A person", "range": [ 0, 8 ] }, { "text": "She", "range": [ 103, 106 ] }, { "text": "her", "range": [ 144, 147 ] }, { "text": "her", "range": [ 165, 168 ] }, { "text": "She", "range": [ 177, 180 ] }, { "text": "her", "range": [ 187, 190 ] } ], "source_mask_link": null }, { "id": "entity_2", "label": "jacket", "spans": [ { "text": "a black jacket", "range": [ 17, 31 ] } ], "source_mask_link": null }, { "id": "entity_3", "label": "leggings", "spans": [ { "text": "colorful leggings", "range": [ 33, 50 ] } ], "source_mask_link": null }, { "id": "entity_4", "label": "baseball cap", "spans": [ { "text": "a black baseball cap", "range": [ 56, 76 ] } ], "source_mask_link": null }, { "id": "entity_5", "label": "floor", "spans": [ { "text": "the floor", "range": [ 92, 101 ] } ], "source_mask_link": null }, { "id": "entity_6", "label": "hands", "spans": [ { "text": "her hands", "range": [ 144, 153 ] } ], "source_mask_link": null }, { "id": "entity_7", "label": "thighs", "spans": [ { "text": "her thighs", "range": [ 165, 175 ] } ], "source_mask_link": null }, { "id": "entity_8", "label": "torso", "spans": [ { "text": "her torso", "range": [ 187, 196 ] } ], "source_mask_link": null } ], "actions": [], "notes": [ "The supplied mask is empty at 0.000 seconds and marks only a foreground bent leg or knee at later samples. That region does not confidently belong to the centrally visible, colorful-legginged person described by the caption.", "Accordingly, the captioned person cannot be linked confidently to the selected-person mask. The stated predicates wearing, is kneeling, is facing, speaking, resting, and leans are omitted under the requirement that an action subject must have a corresponding supplied mask.", "No separate source masks are supplied for the clothing, floor, hands, thighs, or torso." ] }, "model_output_raw": { "caption": "A person wearing a black jacket, colorful leggings, and a black baseball cap is kneeling on the floor. She is facing forward and speaking, with her hands resting on her thighs. She leans her torso forward slightly.", "entity_groups": [ { "id": "entity_1", "label": "person", "spans": [ { "text": "A person", "range": [ 0, 8 ] }, { "text": "She", "range": [ 103, 106 ] }, { "text": "her", "range": [ 144, 147 ] }, { "text": "her", "range": [ 165, 168 ] }, { "text": "She", "range": [ 177, 180 ] }, { "text": "her", "range": [ 187, 190 ] } ], "source_mask": null }, { "id": "entity_2", "label": "jacket", "spans": [ { "text": "a black jacket", "range": [ 17, 31 ] } ], "source_mask": null }, { "id": "entity_3", "label": "leggings", "spans": [ { "text": "colorful leggings", "range": [ 33, 50 ] } ], "source_mask": null }, { "id": "entity_4", "label": "baseball cap", "spans": [ { "text": "a black baseball cap", "range": [ 56, 76 ] } ], "source_mask": null }, { "id": "entity_5", "label": "floor", "spans": [ { "text": "the floor", "range": [ 92, 101 ] } ], "source_mask": null }, { "id": "entity_6", "label": "hands", "spans": [ { "text": "her hands", "range": [ 144, 153 ] } ], "source_mask": null }, { "id": "entity_7", "label": "thighs", "spans": [ { "text": "her thighs", "range": [ 165, 175 ] } ], "source_mask": null }, { "id": "entity_8", "label": "torso", "spans": [ { "text": "her torso", "range": [ 187, 196 ] } ], "source_mask": null } ], "actions": [], "notes": [ "The supplied mask is empty at 0.000 seconds and marks only a foreground bent leg or knee at later samples. That region does not confidently belong to the centrally visible, colorful-legginged person described by the caption.", "Accordingly, the captioned person cannot be linked confidently to the selected-person mask. The stated predicates wearing, is kneeling, is facing, speaking, resting, and leans are omitted under the requirement that an action subject must have a corresponding supplied mask.", "No separate source masks are supplied for the clothing, floor, hands, thighs, or torso." ] }, "normalization_repairs": 0 }, { "source_video_id": "v_-3Ss9jSRbKk", "caption_count_category": 5, "clip": { "caption_index": 2, "caption": "A woman in a black shirt and black pants is lying on her back on a red mat. Her knees are bent with her feet flat on the floor. She moves her arms from her sides to clasp her hands behind her head and holds this position.", "source_interval_sec": [ 9.44, 26.94 ], "loom_clip_id": "v_-3Ss9jSRbKk_283_807", "clip_frame_interval_4fps": [ 283, 807 ], "sample_count_1fps": 18, "sampled_frames": [ { "frame": "00000", "timestamp_sec": 9.44, "offset_sec": 0, "rgb_path": "assets/cat_05/v_-3Ss9jSRbKk_283_807/0000s_00000_rgb.jpg", "mask_path": "assets/cat_05/v_-3Ss9jSRbKk_283_807/0000s_00000_mask.png", "mask_status": "available" }, { "frame": "00028", "timestamp_sec": 10.44, "offset_sec": 1, "rgb_path": "assets/cat_05/v_-3Ss9jSRbKk_283_807/0001s_00028_rgb.jpg", "mask_path": "assets/cat_05/v_-3Ss9jSRbKk_283_807/0001s_00028_mask.png", "mask_status": "available" }, { "frame": "00060", "timestamp_sec": 11.44, "offset_sec": 2, "rgb_path": "assets/cat_05/v_-3Ss9jSRbKk_283_807/0002s_00060_rgb.jpg", "mask_path": "assets/cat_05/v_-3Ss9jSRbKk_283_807/0002s_00060_mask.png", "mask_status": "available" }, { "frame": "00088", "timestamp_sec": 12.44, "offset_sec": 3, "rgb_path": "assets/cat_05/v_-3Ss9jSRbKk_283_807/0003s_00088_rgb.jpg", "mask_path": "assets/cat_05/v_-3Ss9jSRbKk_283_807/0003s_00088_mask.png", "mask_status": "available" }, { "frame": "00120", "timestamp_sec": 13.44, "offset_sec": 4, "rgb_path": "assets/cat_05/v_-3Ss9jSRbKk_283_807/0004s_00120_rgb.jpg", "mask_path": "assets/cat_05/v_-3Ss9jSRbKk_283_807/0004s_00120_mask.png", "mask_status": "available" }, { "frame": "00148", "timestamp_sec": 14.44, "offset_sec": 5, "rgb_path": "assets/cat_05/v_-3Ss9jSRbKk_283_807/0005s_00148_rgb.jpg", "mask_path": "assets/cat_05/v_-3Ss9jSRbKk_283_807/0005s_00148_mask.png", "mask_status": "available" }, { "frame": "00180", "timestamp_sec": 15.44, "offset_sec": 6, "rgb_path": "assets/cat_05/v_-3Ss9jSRbKk_283_807/0006s_00180_rgb.jpg", "mask_path": "assets/cat_05/v_-3Ss9jSRbKk_283_807/0006s_00180_mask.png", "mask_status": "available" }, { "frame": "00208", "timestamp_sec": 16.44, "offset_sec": 7, "rgb_path": "assets/cat_05/v_-3Ss9jSRbKk_283_807/0007s_00208_rgb.jpg", "mask_path": "assets/cat_05/v_-3Ss9jSRbKk_283_807/0007s_00208_mask.png", "mask_status": "available" }, { "frame": "00240", "timestamp_sec": 17.44, "offset_sec": 8, "rgb_path": "assets/cat_05/v_-3Ss9jSRbKk_283_807/0008s_00240_rgb.jpg", "mask_path": "assets/cat_05/v_-3Ss9jSRbKk_283_807/0008s_00240_mask.png", "mask_status": "available" }, { "frame": "00268", "timestamp_sec": 18.44, "offset_sec": 9, "rgb_path": "assets/cat_05/v_-3Ss9jSRbKk_283_807/0009s_00268_rgb.jpg", "mask_path": "assets/cat_05/v_-3Ss9jSRbKk_283_807/0009s_00268_mask.png", "mask_status": "available" }, { "frame": "00300", "timestamp_sec": 19.44, "offset_sec": 10, "rgb_path": "assets/cat_05/v_-3Ss9jSRbKk_283_807/0010s_00300_rgb.jpg", "mask_path": "assets/cat_05/v_-3Ss9jSRbKk_283_807/0010s_00300_mask.png", "mask_status": "available" }, { "frame": "00328", "timestamp_sec": 20.44, "offset_sec": 11, "rgb_path": "assets/cat_05/v_-3Ss9jSRbKk_283_807/0011s_00328_rgb.jpg", "mask_path": "assets/cat_05/v_-3Ss9jSRbKk_283_807/0011s_00328_mask.png", "mask_status": "available" }, { "frame": "00360", "timestamp_sec": 21.44, "offset_sec": 12, "rgb_path": "assets/cat_05/v_-3Ss9jSRbKk_283_807/0012s_00360_rgb.jpg", "mask_path": "assets/cat_05/v_-3Ss9jSRbKk_283_807/0012s_00360_mask.png", "mask_status": "available" }, { "frame": "00388", "timestamp_sec": 22.44, "offset_sec": 13, "rgb_path": "assets/cat_05/v_-3Ss9jSRbKk_283_807/0013s_00388_rgb.jpg", "mask_path": "assets/cat_05/v_-3Ss9jSRbKk_283_807/0013s_00388_mask.png", "mask_status": "available" }, { "frame": "00420", "timestamp_sec": 23.44, "offset_sec": 14, "rgb_path": "assets/cat_05/v_-3Ss9jSRbKk_283_807/0014s_00420_rgb.jpg", "mask_path": "assets/cat_05/v_-3Ss9jSRbKk_283_807/0014s_00420_mask.png", "mask_status": "available" }, { "frame": "00448", "timestamp_sec": 24.44, "offset_sec": 15, "rgb_path": "assets/cat_05/v_-3Ss9jSRbKk_283_807/0015s_00448_rgb.jpg", "mask_path": "assets/cat_05/v_-3Ss9jSRbKk_283_807/0015s_00448_mask.png", "mask_status": "available" }, { "frame": "00480", "timestamp_sec": 25.44, "offset_sec": 16, "rgb_path": "assets/cat_05/v_-3Ss9jSRbKk_283_807/0016s_00480_rgb.jpg", "mask_path": "assets/cat_05/v_-3Ss9jSRbKk_283_807/0016s_00480_mask.png", "mask_status": "available" }, { "frame": "00508", "timestamp_sec": 26.44, "offset_sec": 17, "rgb_path": "assets/cat_05/v_-3Ss9jSRbKk_283_807/0017s_00508_rgb.jpg", "mask_path": "assets/cat_05/v_-3Ss9jSRbKk_283_807/0017s_00508_mask.png", "mask_status": "available" } ], "obj_id": [ 0 ], "anno_id": [ 25 ], "fps_inferred_from_clip_interval": 29.9429 }, "usage": { "input_tokens": 104458, "input_tokens_details": { "cache_write_tokens": 104455, "cached_tokens": 0 }, "output_tokens": 4834, "output_tokens_details": { "reasoning_tokens": 3045 }, "total_tokens": 109292 }, "model": "gpt-5.6-sol", "effort": "medium", "custom_id": "cat05_cap02_v_-3Ss9jSRbKk_283_807", "annotation": { "caption": "A woman in a black shirt and black pants is lying on her back on a red mat. Her knees are bent with her feet flat on the floor. She moves her arms from her sides to clasp her hands behind her head and holds this position.", "entity_groups": [ { "id": "E1", "label": "woman", "spans": [ { "text": "A woman", "range": [ 0, 7 ] }, { "text": "her", "range": [ 53, 56 ] }, { "text": "Her", "range": [ 76, 79 ] }, { "text": "her", "range": [ 100, 103 ] }, { "text": "She", "range": [ 128, 131 ] }, { "text": "her", "range": [ 138, 141 ] }, { "text": "her", "range": [ 152, 155 ] }, { "text": "her", "range": [ 171, 174 ] }, { "text": "her", "range": [ 188, 191 ] } ], "source_mask_link": "selected_person" }, { "id": "E2", "label": "shirt", "spans": [ { "text": "a black shirt", "range": [ 11, 24 ] } ], "source_mask_link": null }, { "id": "E3", "label": "pants", "spans": [ { "text": "black pants", "range": [ 29, 40 ] } ], "source_mask_link": null }, { "id": "E4", "label": "back", "spans": [ { "text": "her back", "range": [ 53, 61 ] } ], "source_mask_link": null }, { "id": "E5", "label": "mat", "spans": [ { "text": "a red mat", "range": [ 65, 74 ] } ], "source_mask_link": null }, { "id": "E6", "label": "knees", "spans": [ { "text": "Her knees", "range": [ 76, 85 ] } ], "source_mask_link": null }, { "id": "E7", "label": "feet", "spans": [ { "text": "her feet", "range": [ 100, 108 ] } ], "source_mask_link": null }, { "id": "E8", "label": "floor", "spans": [ { "text": "the floor", "range": [ 117, 126 ] } ], "source_mask_link": null }, { "id": "E9", "label": "arms", "spans": [ { "text": "her arms", "range": [ 138, 146 ] } ], "source_mask_link": null }, { "id": "E10", "label": "sides", "spans": [ { "text": "her sides", "range": [ 152, 161 ] } ], "source_mask_link": null }, { "id": "E11", "label": "hands", "spans": [ { "text": "her hands", "range": [ 171, 180 ] } ], "source_mask_link": null }, { "id": "E12", "label": "head", "spans": [ { "text": "her head", "range": [ 188, 196 ] } ], "source_mask_link": null } ], "actions": [ { "id": "A1", "mentions": [ { "text": "is lying", "ranges": [ [ 41, 49 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "type": "posture", "subject_refs": [ "E1" ], "object_refs": [], "confidence": "high", "temporal_status": "supported", "intervals": [ { "start_s": 0.0, "end_s": 17.5, "source_start_s": 9.44, "source_end_s": 26.94 } ], "evidence": "The mask-linked woman is lying on her back in every sampled frame from 0.000 through 17.000 seconds.", "mask_relation": "subject matches the supplied selected-person mask", "source_subject_refs": [ "E1" ] }, { "id": "A2", "mentions": [ { "text": "moves", "ranges": [ [ 132, 137 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "type": "simple", "subject_refs": [ "E1" ], "object_refs": [ "E9" ], "confidence": "high", "temporal_status": "not_observed_in_sampled_frames", "intervals": [], "evidence": "The mask-linked woman already has her arms positioned behind her head at the first sample, and no sampled frame shows movement from her sides.", "mask_relation": "subject matches the supplied selected-person mask", "source_subject_refs": [ "E1" ] }, { "id": "A3", "mentions": [ { "text": "clasp", "ranges": [ [ 165, 170 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "type": "simple", "subject_refs": [ "E1" ], "object_refs": [ "E11" ], "confidence": "high", "temporal_status": "not_observed_in_sampled_frames", "intervals": [], "evidence": "Her hands remain behind her head throughout the samples, but the clasping transition itself is not visible and the hands are partly occluded.", "mask_relation": "subject matches the supplied selected-person mask", "source_subject_refs": [ "E1" ] }, { "id": "A4", "mentions": [ { "text": "holds", "ranges": [ [ 201, 206 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "type": "simple", "subject_refs": [ "E1" ], "object_refs": [], "confidence": "high", "temporal_status": "supported", "intervals": [ { "start_s": 0.0, "end_s": 17.5, "source_start_s": 9.44, "source_end_s": 26.94 } ], "evidence": "Across all sampled frames, the mask-linked woman maintains the same supine pose with bent knees and her hands behind her head.", "mask_relation": "subject matches the supplied selected-person mask", "source_subject_refs": [ "E1" ] } ], "notes": [ "The supplied mask consistently identifies the woman lying on the mat, not the kneeling instructor visible in the RGB frames.", "Only the woman has a supplied instance mask. Clothing, body parts, the mat, and the floor therefore have null source-mask links.", "\"are bent\" describes the knees' resulting bodily configuration and was not added as a separate event; the knees also have no dedicated supplied mask.", "\"this position\" is abstract and is not an entity group." ] }, "model_output_raw": { "caption": "A woman in a black shirt and black pants is lying on her back on a red mat. Her knees are bent with her feet flat on the floor. She moves her arms from her sides to clasp her hands behind her head and holds this position.", "entity_groups": [ { "id": "E1", "label": "woman", "spans": [ { "text": "A woman", "range": [ 0, 7 ] }, { "text": "her", "range": [ 53, 56 ] }, { "text": "Her", "range": [ 76, 79 ] }, { "text": "her", "range": [ 100, 103 ] }, { "text": "She", "range": [ 128, 131 ] }, { "text": "her", "range": [ 138, 141 ] }, { "text": "her", "range": [ 152, 155 ] }, { "text": "her", "range": [ 171, 174 ] }, { "text": "her", "range": [ 188, 191 ] } ], "source_mask": { "id": "selected_person_mask" } }, { "id": "E2", "label": "shirt", "spans": [ { "text": "a black shirt", "range": [ 11, 24 ] } ], "source_mask": null }, { "id": "E3", "label": "pants", "spans": [ { "text": "black pants", "range": [ 29, 40 ] } ], "source_mask": null }, { "id": "E4", "label": "back", "spans": [ { "text": "her back", "range": [ 53, 61 ] } ], "source_mask": null }, { "id": "E5", "label": "mat", "spans": [ { "text": "a red mat", "range": [ 65, 74 ] } ], "source_mask": null }, { "id": "E6", "label": "knees", "spans": [ { "text": "Her knees", "range": [ 76, 85 ] } ], "source_mask": null }, { "id": "E7", "label": "feet", "spans": [ { "text": "her feet", "range": [ 100, 108 ] } ], "source_mask": null }, { "id": "E8", "label": "floor", "spans": [ { "text": "the floor", "range": [ 117, 126 ] } ], "source_mask": null }, { "id": "E9", "label": "arms", "spans": [ { "text": "her arms", "range": [ 138, 146 ] } ], "source_mask": null }, { "id": "E10", "label": "sides", "spans": [ { "text": "her sides", "range": [ 152, 161 ] } ], "source_mask": null }, { "id": "E11", "label": "hands", "spans": [ { "text": "her hands", "range": [ 171, 180 ] } ], "source_mask": null }, { "id": "E12", "label": "head", "spans": [ { "text": "her head", "range": [ 188, 196 ] } ], "source_mask": null } ], "actions": [ { "id": "A1", "mentions": [ { "text": "is lying", "ranges": [ [ 41, 49 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "type": "posture", "subject_refs": [ "E1" ], "object_refs": [], "confidence": "high", "temporal_status": "supported", "intervals": [ { "clip_relative_s": [ 0.0, 17.5 ], "source_video_s": [ 9.44, 26.94 ] } ], "evidence": "The mask-linked woman is lying on her back in every sampled frame from 0.000 through 17.000 seconds.", "mask_relation": "subject matches the supplied selected-person mask" }, { "id": "A2", "mentions": [ { "text": "moves", "ranges": [ [ 132, 137 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "type": "simple", "subject_refs": [ "E1" ], "object_refs": [ "E9" ], "confidence": "high", "temporal_status": "not_observed_in_sampled_frames", "intervals": [], "evidence": "The mask-linked woman already has her arms positioned behind her head at the first sample, and no sampled frame shows movement from her sides.", "mask_relation": "subject matches the supplied selected-person mask" }, { "id": "A3", "mentions": [ { "text": "clasp", "ranges": [ [ 165, 170 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "type": "simple", "subject_refs": [ "E1" ], "object_refs": [ "E11" ], "confidence": "high", "temporal_status": "not_observed_in_sampled_frames", "intervals": [], "evidence": "Her hands remain behind her head throughout the samples, but the clasping transition itself is not visible and the hands are partly occluded.", "mask_relation": "subject matches the supplied selected-person mask" }, { "id": "A4", "mentions": [ { "text": "holds", "ranges": [ [ 201, 206 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "type": "simple", "subject_refs": [ "E1" ], "object_refs": [], "confidence": "high", "temporal_status": "supported", "intervals": [ { "clip_relative_s": [ 0.0, 17.5 ], "source_video_s": [ 9.44, 26.94 ] } ], "evidence": "Across all sampled frames, the mask-linked woman maintains the same supine pose with bent knees and her hands behind her head.", "mask_relation": "subject matches the supplied selected-person mask" } ], "notes": [ "The supplied mask consistently identifies the woman lying on the mat, not the kneeling instructor visible in the RGB frames.", "Only the woman has a supplied instance mask. Clothing, body parts, the mat, and the floor therefore have null source-mask links.", "\"are bent\" describes the knees' resulting bodily configuration and was not added as a separate event; the knees also have no dedicated supplied mask.", "\"this position\" is abstract and is not an entity group." ] }, "normalization_repairs": 0 }, { "source_video_id": "v_-3Ss9jSRbKk", "caption_count_category": 5, "clip": { "caption_index": 3, "caption": "The person performs a sit-up by lifting her head and shoulders off the mat. She holds this raised position for a few moments. She then slowly lowers her upper body back down to the mat.", "source_interval_sec": [ 26.94, 36.94 ], "loom_clip_id": "v_-3Ss9jSRbKk_807_1107", "clip_frame_interval_4fps": [ 807, 1107 ], "sample_count_1fps": 10, "sampled_frames": [ { "frame": "00000", "timestamp_sec": 26.94, "offset_sec": 0, "rgb_path": "assets/cat_05/v_-3Ss9jSRbKk_807_1107/0000s_00000_rgb.jpg", "mask_path": "assets/cat_05/v_-3Ss9jSRbKk_807_1107/0000s_00000_mask.png", "mask_status": "available" }, { "frame": "00032", "timestamp_sec": 27.94, "offset_sec": 1, "rgb_path": "assets/cat_05/v_-3Ss9jSRbKk_807_1107/0001s_00032_rgb.jpg", "mask_path": "assets/cat_05/v_-3Ss9jSRbKk_807_1107/0001s_00032_mask.png", "mask_status": "available" }, { "frame": "00060", "timestamp_sec": 28.94, "offset_sec": 2, "rgb_path": "assets/cat_05/v_-3Ss9jSRbKk_807_1107/0002s_00060_rgb.jpg", "mask_path": "assets/cat_05/v_-3Ss9jSRbKk_807_1107/0002s_00060_mask.png", "mask_status": "available" }, { "frame": "00092", "timestamp_sec": 29.94, "offset_sec": 3, "rgb_path": "assets/cat_05/v_-3Ss9jSRbKk_807_1107/0003s_00092_rgb.jpg", "mask_path": "assets/cat_05/v_-3Ss9jSRbKk_807_1107/0003s_00092_mask.png", "mask_status": "available" }, { "frame": "00120", "timestamp_sec": 30.94, "offset_sec": 4, "rgb_path": "assets/cat_05/v_-3Ss9jSRbKk_807_1107/0004s_00120_rgb.jpg", "mask_path": "assets/cat_05/v_-3Ss9jSRbKk_807_1107/0004s_00120_mask.png", "mask_status": "available" }, { "frame": "00152", "timestamp_sec": 31.94, "offset_sec": 5, "rgb_path": "assets/cat_05/v_-3Ss9jSRbKk_807_1107/0005s_00152_rgb.jpg", "mask_path": "assets/cat_05/v_-3Ss9jSRbKk_807_1107/0005s_00152_mask.png", "mask_status": "available" }, { "frame": "00180", "timestamp_sec": 32.94, "offset_sec": 6, "rgb_path": "assets/cat_05/v_-3Ss9jSRbKk_807_1107/0006s_00180_rgb.jpg", "mask_path": "assets/cat_05/v_-3Ss9jSRbKk_807_1107/0006s_00180_mask.png", "mask_status": "available" }, { "frame": "00212", "timestamp_sec": 33.94, "offset_sec": 7, "rgb_path": "assets/cat_05/v_-3Ss9jSRbKk_807_1107/0007s_00212_rgb.jpg", "mask_path": "assets/cat_05/v_-3Ss9jSRbKk_807_1107/0007s_00212_mask.png", "mask_status": "available" }, { "frame": "00240", "timestamp_sec": 34.94, "offset_sec": 8, "rgb_path": "assets/cat_05/v_-3Ss9jSRbKk_807_1107/0008s_00240_rgb.jpg", "mask_path": "assets/cat_05/v_-3Ss9jSRbKk_807_1107/0008s_00240_mask.png", "mask_status": "available" }, { "frame": "00272", "timestamp_sec": 35.94, "offset_sec": 9, "rgb_path": "assets/cat_05/v_-3Ss9jSRbKk_807_1107/0009s_00272_rgb.jpg", "mask_path": "assets/cat_05/v_-3Ss9jSRbKk_807_1107/0009s_00272_mask.png", "mask_status": "available" } ], "obj_id": [ 0 ], "anno_id": [ 26 ], "fps_inferred_from_clip_interval": 30.0 }, "usage": { "input_tokens": 64839, "input_tokens_details": { "cache_write_tokens": 64836, "cached_tokens": 0 }, "output_tokens": 2980, "output_tokens_details": { "reasoning_tokens": 1552 }, "total_tokens": 67819 }, "model": "gpt-5.6-sol", "effort": "medium", "custom_id": "cat05_cap03_v_-3Ss9jSRbKk_807_1107", "annotation": { "caption": "The person performs a sit-up by lifting her head and shoulders off the mat. She holds this raised position for a few moments. She then slowly lowers her upper body back down to the mat.", "entity_groups": [ { "id": "person_1", "label": "person", "spans": [ { "text": "The person", "range": [ 0, 10 ] }, { "text": "her", "range": [ 40, 43 ] }, { "text": "She", "range": [ 76, 79 ] }, { "text": "She", "range": [ 126, 129 ] }, { "text": "her", "range": [ 149, 152 ] } ], "source_mask_link": "selected_person" }, { "id": "head_1", "label": "head", "spans": [ { "text": "her head", "range": [ 40, 48 ] } ], "source_mask_link": null }, { "id": "shoulders_1", "label": "shoulders", "spans": [ { "text": "shoulders", "range": [ 53, 62 ] } ], "source_mask_link": null }, { "id": "mat_1", "label": "mat", "spans": [ { "text": "the mat", "range": [ 67, 74 ] }, { "text": "the mat", "range": [ 177, 184 ] } ], "source_mask_link": null }, { "id": "upper_body_1", "label": "upper body", "spans": [ { "text": "her upper body", "range": [ 149, 163 ] } ], "source_mask_link": null } ], "actions": [ { "id": "action_1", "mentions": [ { "text": "performs a sit-up", "ranges": [ [ 11, 28 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "type": "light_verb", "subject_refs": [ "person_1" ], "object_refs": [], "confidence": "high", "temporal_status": "supported", "intervals": [ { "start_s": 0.0, "end_s": 8.0, "source_start_s": 26.94, "source_end_s": 34.94 } ], "evidence": "The masked reclining person rises from the mat between samples 0 and 1, maintains the raised sit-up posture through sample 6, and is lowered again by sample 7.", "mask_relation": "The action subject matches the supplied selected-person mask.", "source_subject_refs": "person_1" }, { "id": "action_2", "mentions": [ { "text": "lifting", "ranges": [ [ 32, 39 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "type": "simple", "subject_refs": [ "person_1" ], "object_refs": [], "confidence": "high", "temporal_status": "supported", "intervals": [ { "start_s": 0.0, "end_s": 2.0, "source_start_s": 26.94, "source_end_s": 28.94 } ], "evidence": "The masked person's head and shoulders are near the mat at sample 0 and visibly elevated at sample 1.", "mask_relation": "The action subject matches the supplied selected-person mask.", "source_subject_refs": "person_1" }, { "id": "action_3", "mentions": [ { "text": "holds", "ranges": [ [ 80, 85 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "type": "posture", "subject_refs": [ "person_1" ], "object_refs": [], "confidence": "high", "temporal_status": "supported", "intervals": [ { "start_s": 1.0, "end_s": 7.0, "source_start_s": 27.94, "source_end_s": 33.94 } ], "evidence": "Across samples 1 through 6, the masked person maintains a raised head-and-shoulders posture with little change.", "mask_relation": "The action subject matches the supplied selected-person mask.", "source_subject_refs": "person_1" }, { "id": "action_4", "mentions": [ { "text": "lowers", "ranges": [ [ 142, 148 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "type": "simple", "subject_refs": [ "person_1" ], "object_refs": [], "confidence": "high", "temporal_status": "supported", "intervals": [ { "start_s": 6.0, "end_s": 8.0, "source_start_s": 32.94, "source_end_s": 34.94 } ], "evidence": "The masked person's upper body is raised at sample 6 and lies back near the mat at sample 7.", "mask_relation": "The action subject matches the supplied selected-person mask.", "source_subject_refs": "person_1" } ], "notes": [ "The supplied mask consistently identifies the reclining person performing the exercise, not the standing instructor.", "No separate masks were supplied for the person's head, shoulders, upper body, or the mat; those entity groups therefore have null source-mask links.", "Samples 8 and 9 show the masked person rising again, but the caption does not explicitly state a repeated sit-up, so that later repetition is not assigned to a new caption action." ] }, "model_output_raw": { "caption": "The person performs a sit-up by lifting her head and shoulders off the mat. She holds this raised position for a few moments. She then slowly lowers her upper body back down to the mat.", "entity_groups": [ { "id": "person_1", "entity_type": "person", "spans": [ { "text": "The person", "range": [ 0, 10 ] }, { "text": "her", "range": [ 40, 43 ] }, { "text": "She", "range": [ 76, 79 ] }, { "text": "She", "range": [ 126, 129 ] }, { "text": "her", "range": [ 149, 152 ] } ], "source_mask": "selected_person_masklet" }, { "id": "head_1", "entity_type": "head", "spans": [ { "text": "her head", "range": [ 40, 48 ] } ], "source_mask": null }, { "id": "shoulders_1", "entity_type": "shoulders", "spans": [ { "text": "shoulders", "range": [ 53, 62 ] } ], "source_mask": null }, { "id": "mat_1", "entity_type": "mat", "spans": [ { "text": "the mat", "range": [ 67, 74 ] }, { "text": "the mat", "range": [ 177, 184 ] } ], "source_mask": null }, { "id": "upper_body_1", "entity_type": "upper body", "spans": [ { "text": "her upper body", "range": [ 149, 163 ] } ], "source_mask": null } ], "actions": [ { "id": "action_1", "mentions": [ { "text": "performs a sit-up", "ranges": [ [ 11, 28 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "type": "light_verb", "subject_reference": "person_1", "object_references": [], "confidence": "high", "temporal_status": "supported", "intervals": [ { "clip_relative_s": [ 0.0, 8.0 ], "source_video_s": [ 26.94, 34.94 ] } ], "evidence": "The masked reclining person rises from the mat between samples 0 and 1, maintains the raised sit-up posture through sample 6, and is lowered again by sample 7.", "mask_relation": "The action subject matches the supplied selected-person mask." }, { "id": "action_2", "mentions": [ { "text": "lifting", "ranges": [ [ 32, 39 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "type": "simple", "subject_reference": "person_1", "object_references": [ "head_1", "shoulders_1" ], "confidence": "high", "temporal_status": "supported", "intervals": [ { "clip_relative_s": [ 0.0, 2.0 ], "source_video_s": [ 26.94, 28.94 ] } ], "evidence": "The masked person's head and shoulders are near the mat at sample 0 and visibly elevated at sample 1.", "mask_relation": "The action subject matches the supplied selected-person mask." }, { "id": "action_3", "mentions": [ { "text": "holds", "ranges": [ [ 80, 85 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "type": "posture", "subject_reference": "person_1", "object_references": [], "confidence": "high", "temporal_status": "supported", "intervals": [ { "clip_relative_s": [ 1.0, 7.0 ], "source_video_s": [ 27.94, 33.94 ] } ], "evidence": "Across samples 1 through 6, the masked person maintains a raised head-and-shoulders posture with little change.", "mask_relation": "The action subject matches the supplied selected-person mask." }, { "id": "action_4", "mentions": [ { "text": "lowers", "ranges": [ [ 142, 148 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "type": "simple", "subject_reference": "person_1", "object_references": [ "upper_body_1" ], "confidence": "high", "temporal_status": "supported", "intervals": [ { "clip_relative_s": [ 6.0, 8.0 ], "source_video_s": [ 32.94, 34.94 ] } ], "evidence": "The masked person's upper body is raised at sample 6 and lies back near the mat at sample 7.", "mask_relation": "The action subject matches the supplied selected-person mask." } ], "notes": [ "The supplied mask consistently identifies the reclining person performing the exercise, not the standing instructor.", "No separate masks were supplied for the person's head, shoulders, upper body, or the mat; those entity groups therefore have null source-mask links.", "Samples 8 and 9 show the masked person rising again, but the caption does not explicitly state a repeated sit-up, so that later repetition is not assigned to a new caption action." ] }, "normalization_repairs": 0 }, { "source_video_id": "v_-3Ss9jSRbKk", "caption_count_category": 5, "clip": { "caption_index": 4, "caption": "The person performs a second sit-up, lifting her head and upper back from the mat and holding the position. She then lowers her torso back down to the mat. She remains in the starting position with her hands behind her head.", "source_interval_sec": [ 36.94, 53.44 ], "loom_clip_id": "v_-3Ss9jSRbKk_1107_1602", "clip_frame_interval_4fps": [ 1107, 1602 ], "sample_count_1fps": 17, "sampled_frames": [ { "frame": "00000", "timestamp_sec": 36.94, "offset_sec": 0, "rgb_path": "assets/cat_05/v_-3Ss9jSRbKk_1107_1602/0000s_00000_rgb.jpg", "mask_path": "assets/cat_05/v_-3Ss9jSRbKk_1107_1602/0000s_00000_mask.png", "mask_status": "available" }, { "frame": "00028", "timestamp_sec": 37.94, "offset_sec": 1, "rgb_path": "assets/cat_05/v_-3Ss9jSRbKk_1107_1602/0001s_00028_rgb.jpg", "mask_path": "assets/cat_05/v_-3Ss9jSRbKk_1107_1602/0001s_00028_mask.png", "mask_status": "available" }, { "frame": "00060", "timestamp_sec": 38.94, "offset_sec": 2, "rgb_path": "assets/cat_05/v_-3Ss9jSRbKk_1107_1602/0002s_00060_rgb.jpg", "mask_path": "assets/cat_05/v_-3Ss9jSRbKk_1107_1602/0002s_00060_mask.png", "mask_status": "available" }, { "frame": "00088", "timestamp_sec": 39.94, "offset_sec": 3, "rgb_path": "assets/cat_05/v_-3Ss9jSRbKk_1107_1602/0003s_00088_rgb.jpg", "mask_path": "assets/cat_05/v_-3Ss9jSRbKk_1107_1602/0003s_00088_mask.png", "mask_status": "available" }, { "frame": "00120", "timestamp_sec": 40.94, "offset_sec": 4, "rgb_path": "assets/cat_05/v_-3Ss9jSRbKk_1107_1602/0004s_00120_rgb.jpg", "mask_path": "assets/cat_05/v_-3Ss9jSRbKk_1107_1602/0004s_00120_mask.png", "mask_status": "available" }, { "frame": "00148", "timestamp_sec": 41.94, "offset_sec": 5, "rgb_path": "assets/cat_05/v_-3Ss9jSRbKk_1107_1602/0005s_00148_rgb.jpg", "mask_path": "assets/cat_05/v_-3Ss9jSRbKk_1107_1602/0005s_00148_mask.png", "mask_status": "available" }, { "frame": "00180", "timestamp_sec": 42.94, "offset_sec": 6, "rgb_path": "assets/cat_05/v_-3Ss9jSRbKk_1107_1602/0006s_00180_rgb.jpg", "mask_path": "assets/cat_05/v_-3Ss9jSRbKk_1107_1602/0006s_00180_mask.png", "mask_status": "available" }, { "frame": "00208", "timestamp_sec": 43.94, "offset_sec": 7, "rgb_path": "assets/cat_05/v_-3Ss9jSRbKk_1107_1602/0007s_00208_rgb.jpg", "mask_path": "assets/cat_05/v_-3Ss9jSRbKk_1107_1602/0007s_00208_mask.png", "mask_status": "available" }, { "frame": "00240", "timestamp_sec": 44.94, "offset_sec": 8, "rgb_path": "assets/cat_05/v_-3Ss9jSRbKk_1107_1602/0008s_00240_rgb.jpg", "mask_path": "assets/cat_05/v_-3Ss9jSRbKk_1107_1602/0008s_00240_mask.png", "mask_status": "available" }, { "frame": "00268", "timestamp_sec": 45.94, "offset_sec": 9, "rgb_path": "assets/cat_05/v_-3Ss9jSRbKk_1107_1602/0009s_00268_rgb.jpg", "mask_path": "assets/cat_05/v_-3Ss9jSRbKk_1107_1602/0009s_00268_mask.png", "mask_status": "available" }, { "frame": "00300", "timestamp_sec": 46.94, "offset_sec": 10, "rgb_path": "assets/cat_05/v_-3Ss9jSRbKk_1107_1602/0010s_00300_rgb.jpg", "mask_path": "assets/cat_05/v_-3Ss9jSRbKk_1107_1602/0010s_00300_mask.png", "mask_status": "available" }, { "frame": "00328", "timestamp_sec": 47.94, "offset_sec": 11, "rgb_path": "assets/cat_05/v_-3Ss9jSRbKk_1107_1602/0011s_00328_rgb.jpg", "mask_path": "assets/cat_05/v_-3Ss9jSRbKk_1107_1602/0011s_00328_mask.png", "mask_status": "available" }, { "frame": "00360", "timestamp_sec": 48.94, "offset_sec": 12, "rgb_path": "assets/cat_05/v_-3Ss9jSRbKk_1107_1602/0012s_00360_rgb.jpg", "mask_path": "assets/cat_05/v_-3Ss9jSRbKk_1107_1602/0012s_00360_mask.png", "mask_status": "available" }, { "frame": "00388", "timestamp_sec": 49.94, "offset_sec": 13, "rgb_path": "assets/cat_05/v_-3Ss9jSRbKk_1107_1602/0013s_00388_rgb.jpg", "mask_path": "assets/cat_05/v_-3Ss9jSRbKk_1107_1602/0013s_00388_mask.png", "mask_status": "available" }, { "frame": "00420", "timestamp_sec": 50.94, "offset_sec": 14, "rgb_path": "assets/cat_05/v_-3Ss9jSRbKk_1107_1602/0014s_00420_rgb.jpg", "mask_path": "assets/cat_05/v_-3Ss9jSRbKk_1107_1602/0014s_00420_mask.png", "mask_status": "available" }, { "frame": "00448", "timestamp_sec": 51.94, "offset_sec": 15, "rgb_path": "assets/cat_05/v_-3Ss9jSRbKk_1107_1602/0015s_00448_rgb.jpg", "mask_path": "assets/cat_05/v_-3Ss9jSRbKk_1107_1602/0015s_00448_mask.png", "mask_status": "available" }, { "frame": "00480", "timestamp_sec": 52.94, "offset_sec": 16, "rgb_path": "assets/cat_05/v_-3Ss9jSRbKk_1107_1602/0016s_00480_rgb.jpg", "mask_path": "assets/cat_05/v_-3Ss9jSRbKk_1107_1602/0016s_00480_mask.png", "mask_status": "available" } ], "obj_id": [ 0 ], "anno_id": [ 23 ], "fps_inferred_from_clip_interval": 30.0 }, "usage": { "input_tokens": 99503, "input_tokens_details": { "cache_write_tokens": 99500, "cached_tokens": 0 }, "output_tokens": 4828, "output_tokens_details": { "reasoning_tokens": 3106 }, "total_tokens": 104331 }, "model": "gpt-5.6-sol", "effort": "medium", "custom_id": "cat05_cap04_v_-3Ss9jSRbKk_1107_1602", "annotation": { "caption": "The person performs a second sit-up, lifting her head and upper back from the mat and holding the position. She then lowers her torso back down to the mat. She remains in the starting position with her hands behind her head.", "entity_groups": [ { "id": "person_1", "label": "person", "spans": [ { "text": "The person", "range": [ 0, 10 ] }, { "text": "her", "range": [ 45, 48 ] }, { "text": "She", "range": [ 108, 111 ] }, { "text": "her", "range": [ 124, 127 ] }, { "text": "She", "range": [ 156, 159 ] }, { "text": "her", "range": [ 198, 201 ] }, { "text": "her", "range": [ 215, 218 ] } ], "source_mask_link": "selected_person" }, { "id": "head_1", "label": "body_part", "spans": [ { "text": "her head", "range": [ 45, 53 ] }, { "text": "her head", "range": [ 215, 223 ] } ], "source_mask_link": null }, { "id": "upper_back_1", "label": "body_part", "spans": [ { "text": "upper back", "range": [ 58, 68 ] } ], "source_mask_link": null }, { "id": "mat_1", "label": "object", "spans": [ { "text": "the mat", "range": [ 74, 81 ] }, { "text": "the mat", "range": [ 147, 154 ] } ], "source_mask_link": null }, { "id": "torso_1", "label": "body_part", "spans": [ { "text": "her torso", "range": [ 124, 133 ] } ], "source_mask_link": null }, { "id": "hands_1", "label": "body_part", "spans": [ { "text": "her hands", "range": [ 198, 207 ] } ], "source_mask_link": null } ], "actions": [ { "id": "action_1", "mentions": [ { "text": "performs a second sit-up", "ranges": [ [ 11, 35 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "type": "light_verb", "subject_refs": [ "person_1" ], "object_refs": [], "confidence": "high", "temporal_status": "localized_from_sparse_samples", "intervals": [ { "start_s": 2.0, "end_s": 7.0, "source_start_s": 38.94, "source_end_s": 43.94 } ], "evidence": "The masked exerciser is flat at 2.0 s, raises her upper body by 3.0 s, holds it elevated through 4.0 s, and returns to the mat by 6.0 s.", "mask_relation": "subject_matches_supplied_selected-person_mask", "source_subject_refs": "person_1" }, { "id": "action_2", "mentions": [ { "text": "lifting", "ranges": [ [ 37, 44 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "type": "simple", "subject_refs": [ "person_1" ], "object_refs": [ "head_1", "upper_back_1" ], "confidence": "high", "temporal_status": "localized_from_sparse_samples", "intervals": [ { "start_s": 2.0, "end_s": 4.0, "source_start_s": 38.94, "source_end_s": 40.94 } ], "evidence": "Between the 2.0 s and 3.0 s clean frames, the masked person's head and upper back rise away from the mat.", "mask_relation": "subject_matches_supplied_selected-person_mask", "source_subject_refs": "person_1" }, { "id": "action_3", "mentions": [ { "text": "holding", "ranges": [ [ 86, 93 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "type": "posture", "subject_refs": [ "person_1" ], "object_refs": [], "confidence": "high", "temporal_status": "localized_from_sparse_samples", "intervals": [ { "start_s": 3.0, "end_s": 5.0, "source_start_s": 39.94, "source_end_s": 41.94 } ], "evidence": "At 3.0 s and 4.0 s, the masked person maintains the elevated crunch position.", "mask_relation": "subject_matches_supplied_selected-person_mask", "source_subject_refs": "person_1" }, { "id": "action_4", "mentions": [ { "text": "lowers", "ranges": [ [ 117, 123 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "type": "simple", "subject_refs": [ "person_1" ], "object_refs": [ "torso_1" ], "confidence": "high", "temporal_status": "localized_from_sparse_samples", "intervals": [ { "start_s": 4.0, "end_s": 7.0, "source_start_s": 40.94, "source_end_s": 43.94 } ], "evidence": "The masked person's torso is elevated at 4.0 s, lower at 5.0 s, and resting back on the mat by 6.0 s.", "mask_relation": "subject_matches_supplied_selected-person_mask", "source_subject_refs": "person_1" }, { "id": "action_5", "mentions": [ { "text": "remains", "ranges": [ [ 160, 167 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "type": "posture", "subject_refs": [ "person_1" ], "object_refs": [], "confidence": "high", "temporal_status": "localized_from_sparse_samples", "intervals": [ { "start_s": 6.0, "end_s": 14.0, "source_start_s": 42.94, "source_end_s": 50.94 } ], "evidence": "From 6.0 s through 13.0 s, the masked person stays supine with bent knees and hands behind her head; she begins rising again at 14.0 s.", "mask_relation": "subject_matches_supplied_selected-person_mask", "source_subject_refs": "person_1" } ], "notes": [ "The supplied masklet consistently identifies the exercising person lying on the mat, not the kneeling instructor.", "No separate masks were supplied for the head, upper back, torso, hands, or mat, so those entity groups have null source-mask links.", "The abstract references to 'the position' and 'the starting position' are not annotated as tangible entities.", "Temporal intervals are sparse clip-relative support intervals derived from the sampled clean RGB frames." ] }, "model_output_raw": { "caption": "The person performs a second sit-up, lifting her head and upper back from the mat and holding the position. She then lowers her torso back down to the mat. She remains in the starting position with her hands behind her head.", "entity_groups": [ { "id": "person_1", "entity_type": "person", "spans": [ { "text": "The person", "range": [ 0, 10 ] }, { "text": "her", "range": [ 45, 48 ] }, { "text": "She", "range": [ 108, 111 ] }, { "text": "her", "range": [ 124, 127 ] }, { "text": "She", "range": [ 156, 159 ] }, { "text": "her", "range": [ 198, 201 ] }, { "text": "her", "range": [ 215, 218 ] } ], "source_mask": "selected_person_masklet" }, { "id": "head_1", "entity_type": "body_part", "spans": [ { "text": "her head", "range": [ 45, 53 ] }, { "text": "her head", "range": [ 215, 223 ] } ], "source_mask": null }, { "id": "upper_back_1", "entity_type": "body_part", "spans": [ { "text": "upper back", "range": [ 58, 68 ] } ], "source_mask": null }, { "id": "mat_1", "entity_type": "object", "spans": [ { "text": "the mat", "range": [ 74, 81 ] }, { "text": "the mat", "range": [ 147, 154 ] } ], "source_mask": null }, { "id": "torso_1", "entity_type": "body_part", "spans": [ { "text": "her torso", "range": [ 124, 133 ] } ], "source_mask": null }, { "id": "hands_1", "entity_type": "body_part", "spans": [ { "text": "her hands", "range": [ 198, 207 ] } ], "source_mask": null } ], "actions": [ { "id": "action_1", "mentions": [ { "text": "performs a second sit-up", "ranges": [ [ 11, 35 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "type": "light_verb", "subject_ref": "person_1", "object_refs": [], "confidence": "high", "mask_relation": "subject_matches_supplied_selected-person_mask", "temporal_status": "localized_from_sparse_samples", "intervals": [ { "start_s": 2.0, "end_s": 7.0 } ], "evidence": "The masked exerciser is flat at 2.0 s, raises her upper body by 3.0 s, holds it elevated through 4.0 s, and returns to the mat by 6.0 s." }, { "id": "action_2", "mentions": [ { "text": "lifting", "ranges": [ [ 37, 44 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "type": "simple", "subject_ref": "person_1", "object_refs": [ "head_1", "upper_back_1" ], "confidence": "high", "mask_relation": "subject_matches_supplied_selected-person_mask", "temporal_status": "localized_from_sparse_samples", "intervals": [ { "start_s": 2.0, "end_s": 4.0 } ], "evidence": "Between the 2.0 s and 3.0 s clean frames, the masked person's head and upper back rise away from the mat." }, { "id": "action_3", "mentions": [ { "text": "holding", "ranges": [ [ 86, 93 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "type": "posture", "subject_ref": "person_1", "object_refs": [], "confidence": "high", "mask_relation": "subject_matches_supplied_selected-person_mask", "temporal_status": "localized_from_sparse_samples", "intervals": [ { "start_s": 3.0, "end_s": 5.0 } ], "evidence": "At 3.0 s and 4.0 s, the masked person maintains the elevated crunch position." }, { "id": "action_4", "mentions": [ { "text": "lowers", "ranges": [ [ 117, 123 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "type": "simple", "subject_ref": "person_1", "object_refs": [ "torso_1" ], "confidence": "high", "mask_relation": "subject_matches_supplied_selected-person_mask", "temporal_status": "localized_from_sparse_samples", "intervals": [ { "start_s": 4.0, "end_s": 7.0 } ], "evidence": "The masked person's torso is elevated at 4.0 s, lower at 5.0 s, and resting back on the mat by 6.0 s." }, { "id": "action_5", "mentions": [ { "text": "remains", "ranges": [ [ 160, 167 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "type": "posture", "subject_ref": "person_1", "object_refs": [], "confidence": "high", "mask_relation": "subject_matches_supplied_selected-person_mask", "temporal_status": "localized_from_sparse_samples", "intervals": [ { "start_s": 6.0, "end_s": 14.0 } ], "evidence": "From 6.0 s through 13.0 s, the masked person stays supine with bent knees and hands behind her head; she begins rising again at 14.0 s." } ], "notes": [ "The supplied masklet consistently identifies the exercising person lying on the mat, not the kneeling instructor.", "No separate masks were supplied for the head, upper back, torso, hands, or mat, so those entity groups have null source-mask links.", "The abstract references to 'the position' and 'the starting position' are not annotated as tangible entities.", "Temporal intervals are sparse clip-relative support intervals derived from the sampled clean RGB frames." ] }, "normalization_repairs": 0 }, { "source_video_id": "v_-3Ss9jSRbKk", "caption_count_category": 5, "clip": { "caption_index": 5, "caption": "The person changes her arm position, crossing them over her chest. She then performs another sit-up, lifting her torso up before lowering it back down. Finally, she uncrosses her arms and rests them by her sides on the mat.", "source_interval_sec": [ 53.44, 67.57 ], "loom_clip_id": "v_-3Ss9jSRbKk_1602_2025", "clip_frame_interval_4fps": [ 1602, 2025 ], "sample_count_1fps": 15, "sampled_frames": [ { "frame": "00000", "timestamp_sec": 53.44, "offset_sec": 0, "rgb_path": "assets/cat_05/v_-3Ss9jSRbKk_1602_2025/0000s_00000_rgb.jpg", "mask_path": "assets/cat_05/v_-3Ss9jSRbKk_1602_2025/0000s_00000_mask.png", "mask_status": "available" }, { "frame": "00028", "timestamp_sec": 54.44, "offset_sec": 1, "rgb_path": "assets/cat_05/v_-3Ss9jSRbKk_1602_2025/0001s_00028_rgb.jpg", "mask_path": "assets/cat_05/v_-3Ss9jSRbKk_1602_2025/0001s_00028_mask.png", "mask_status": "available" }, { "frame": "00060", "timestamp_sec": 55.44, "offset_sec": 2, "rgb_path": "assets/cat_05/v_-3Ss9jSRbKk_1602_2025/0002s_00060_rgb.jpg", "mask_path": "assets/cat_05/v_-3Ss9jSRbKk_1602_2025/0002s_00060_mask.png", "mask_status": "available" }, { "frame": "00088", "timestamp_sec": 56.44, "offset_sec": 3, "rgb_path": "assets/cat_05/v_-3Ss9jSRbKk_1602_2025/0003s_00088_rgb.jpg", "mask_path": "assets/cat_05/v_-3Ss9jSRbKk_1602_2025/0003s_00088_mask.png", "mask_status": "available" }, { "frame": "00120", "timestamp_sec": 57.44, "offset_sec": 4, "rgb_path": "assets/cat_05/v_-3Ss9jSRbKk_1602_2025/0004s_00120_rgb.jpg", "mask_path": "assets/cat_05/v_-3Ss9jSRbKk_1602_2025/0004s_00120_mask.png", "mask_status": "available" }, { "frame": "00148", "timestamp_sec": 58.44, "offset_sec": 5, "rgb_path": "assets/cat_05/v_-3Ss9jSRbKk_1602_2025/0005s_00148_rgb.jpg", "mask_path": "assets/cat_05/v_-3Ss9jSRbKk_1602_2025/0005s_00148_mask.png", "mask_status": "available" }, { "frame": "00180", "timestamp_sec": 59.44, "offset_sec": 6, "rgb_path": "assets/cat_05/v_-3Ss9jSRbKk_1602_2025/0006s_00180_rgb.jpg", "mask_path": "assets/cat_05/v_-3Ss9jSRbKk_1602_2025/0006s_00180_mask.png", "mask_status": "available" }, { "frame": "00208", "timestamp_sec": 60.44, "offset_sec": 7, "rgb_path": "assets/cat_05/v_-3Ss9jSRbKk_1602_2025/0007s_00208_rgb.jpg", "mask_path": "assets/cat_05/v_-3Ss9jSRbKk_1602_2025/0007s_00208_mask.png", "mask_status": "available" }, { "frame": "00240", "timestamp_sec": 61.44, "offset_sec": 8, "rgb_path": "assets/cat_05/v_-3Ss9jSRbKk_1602_2025/0008s_00240_rgb.jpg", "mask_path": "assets/cat_05/v_-3Ss9jSRbKk_1602_2025/0008s_00240_mask.png", "mask_status": "available" }, { "frame": "00268", "timestamp_sec": 62.44, "offset_sec": 9, "rgb_path": "assets/cat_05/v_-3Ss9jSRbKk_1602_2025/0009s_00268_rgb.jpg", "mask_path": "assets/cat_05/v_-3Ss9jSRbKk_1602_2025/0009s_00268_mask.png", "mask_status": "available" }, { "frame": "00300", "timestamp_sec": 63.44, "offset_sec": 10, "rgb_path": "assets/cat_05/v_-3Ss9jSRbKk_1602_2025/0010s_00300_rgb.jpg", "mask_path": "assets/cat_05/v_-3Ss9jSRbKk_1602_2025/0010s_00300_mask.png", "mask_status": "available" }, { "frame": "00328", "timestamp_sec": 64.44, "offset_sec": 11, "rgb_path": "assets/cat_05/v_-3Ss9jSRbKk_1602_2025/0011s_00328_rgb.jpg", "mask_path": "assets/cat_05/v_-3Ss9jSRbKk_1602_2025/0011s_00328_mask.png", "mask_status": "available" }, { "frame": "00360", "timestamp_sec": 65.44, "offset_sec": 12, "rgb_path": "assets/cat_05/v_-3Ss9jSRbKk_1602_2025/0012s_00360_rgb.jpg", "mask_path": "assets/cat_05/v_-3Ss9jSRbKk_1602_2025/0012s_00360_mask.png", "mask_status": "available" }, { "frame": "00388", "timestamp_sec": 66.44, "offset_sec": 13, "rgb_path": "assets/cat_05/v_-3Ss9jSRbKk_1602_2025/0013s_00388_rgb.jpg", "mask_path": "assets/cat_05/v_-3Ss9jSRbKk_1602_2025/0013s_00388_mask.png", "mask_status": "available" }, { "frame": "00420", "timestamp_sec": 67.44, "offset_sec": 14, "rgb_path": "assets/cat_05/v_-3Ss9jSRbKk_1602_2025/0014s_00420_rgb.jpg", "mask_path": "assets/cat_05/v_-3Ss9jSRbKk_1602_2025/0014s_00420_mask.png", "mask_status": "available" } ], "obj_id": [ 0 ], "anno_id": [ 24 ], "fps_inferred_from_clip_interval": 29.9363 }, "usage": { "input_tokens": 89603, "input_tokens_details": { "cache_write_tokens": 89600, "cached_tokens": 0 }, "output_tokens": 5148, "output_tokens_details": { "reasoning_tokens": 2992 }, "total_tokens": 94751 }, "model": "gpt-5.6-sol", "effort": "medium", "custom_id": "cat05_cap05_v_-3Ss9jSRbKk_1602_2025", "annotation": { "caption": "The person changes her arm position, crossing them over her chest. She then performs another sit-up, lifting her torso up before lowering it back down. Finally, she uncrosses her arms and rests them by her sides on the mat.", "entity_groups": [ { "id": "person_1", "label": "person", "spans": [ { "text": "The person", "range": [ 0, 10 ] }, { "text": "her", "range": [ 19, 22 ] }, { "text": "her", "range": [ 56, 59 ] }, { "text": "She", "range": [ 67, 70 ] }, { "text": "her", "range": [ 109, 112 ] }, { "text": "she", "range": [ 161, 164 ] }, { "text": "her", "range": [ 175, 178 ] }, { "text": "her", "range": [ 202, 205 ] } ], "source_mask_link": "selected_person" }, { "id": "arms_1", "label": "arms", "spans": [ { "text": "her arm", "range": [ 19, 26 ] }, { "text": "them", "range": [ 46, 50 ] }, { "text": "her arms", "range": [ 175, 183 ] }, { "text": "them", "range": [ 194, 198 ] } ], "source_mask_link": null }, { "id": "chest_1", "label": "chest", "spans": [ { "text": "her chest", "range": [ 56, 65 ] } ], "source_mask_link": null }, { "id": "torso_1", "label": "torso", "spans": [ { "text": "her torso", "range": [ 109, 118 ] }, { "text": "it", "range": [ 138, 140 ] } ], "source_mask_link": null }, { "id": "sides_1", "label": "sides", "spans": [ { "text": "her sides", "range": [ 202, 211 ] } ], "source_mask_link": null }, { "id": "mat_1", "label": "mat", "spans": [ { "text": "the mat", "range": [ 215, 222 ] } ], "source_mask_link": null } ], "actions": [ { "id": "action_1", "mentions": [ { "text": "changes", "ranges": [ [ 11, 18 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } }, { "text": "crossing", "ranges": [ [ 37, 45 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "type": "simple", "subject_refs": [ "person_1" ], "object_refs": [ "arms_1" ], "confidence": "high", "temporal_status": "localized", "intervals": [ { "start_s": 2.0, "end_s": 4.0, "source_start_s": 55.44, "source_end_s": 57.44 } ], "evidence": "At clip-relative 2.0 seconds the selected person is moving her arms toward her chest; by 3.0 seconds they are crossed over it.", "mask_relation": "subject_matches_selected_person_mask", "source_subject_refs": "person_1" }, { "id": "action_2", "mentions": [ { "text": "performs another sit-up", "ranges": [ [ 76, 99 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "type": "light_verb", "subject_refs": [ "person_1" ], "object_refs": [], "confidence": "high", "temporal_status": "localized", "intervals": [ { "start_s": 3.0, "end_s": 13.0, "source_start_s": 56.44, "source_end_s": 66.44 } ], "evidence": "Clean frames from 3.0 through 11.0 seconds show the selected person with her torso raised in a sit-up, followed by a return to the mat at 12.0 seconds.", "mask_relation": "subject_matches_selected_person_mask", "source_subject_refs": "person_1" }, { "id": "action_3", "mentions": [ { "text": "lifting", "ranges": [ [ 101, 108 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "type": "simple", "subject_refs": [ "person_1" ], "object_refs": [ "torso_1" ], "confidence": "high", "temporal_status": "localized", "intervals": [ { "start_s": 3.0, "end_s": 7.0, "source_start_s": 56.44, "source_end_s": 60.44 } ], "evidence": "From 3.0 through 6.0 seconds, the selected person's head, shoulders, and torso rise progressively away from the mat.", "mask_relation": "subject_matches_selected_person_mask", "source_subject_refs": "person_1" }, { "id": "action_4", "mentions": [ { "text": "lowering", "ranges": [ [ 129, 137 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "type": "simple", "subject_refs": [ "person_1" ], "object_refs": [ "torso_1" ], "confidence": "high", "temporal_status": "localized", "intervals": [ { "start_s": 7.0, "end_s": 13.0, "source_start_s": 60.44, "source_end_s": 66.44 } ], "evidence": "Across the later sampled frames, the selected person's raised torso returns toward the mat and is fully down by 12.0 seconds.", "mask_relation": "subject_matches_selected_person_mask", "source_subject_refs": "person_1" }, { "id": "action_5", "mentions": [ { "text": "uncrosses", "ranges": [ [ 165, 174 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "type": "simple", "subject_refs": [ "person_1" ], "object_refs": [ "arms_1" ], "confidence": "high", "temporal_status": "localized", "intervals": [ { "start_s": 11.0, "end_s": 13.0, "source_start_s": 64.44, "source_end_s": 66.44 } ], "evidence": "The selected person's arms are still crossed at 11.0 seconds and are uncrossed beside her body at 12.0 seconds.", "mask_relation": "subject_matches_selected_person_mask", "source_subject_refs": "person_1" }, { "id": "action_6", "mentions": [ { "text": "rests", "ranges": [ [ 188, 193 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "type": "posture", "subject_refs": [ "person_1" ], "object_refs": [ "arms_1" ], "confidence": "high", "temporal_status": "localized", "intervals": [ { "start_s": 12.0, "end_s": 14.129999999999995, "source_start_s": 65.44, "source_end_s": 67.57 } ], "evidence": "At 12.0, 13.0, and 14.0 seconds, the selected person lies on the mat with her arms resting alongside her body; the posture continues to the clip boundary.", "mask_relation": "subject_matches_selected_person_mask", "source_subject_refs": "person_1" } ], "notes": [ "The supplied masklet consistently identifies the person lying on the mat, who is the caption's action subject.", "The kneeling instructor is visible but is not selected by the supplied mask and is not mentioned as an action subject in the caption.", "No separate masks were supplied for the arms, chest, torso, sides, or mat; their source-mask links are therefore null.", "Temporal intervals are sparse support intervals based on the sampled clean RGB frames, not exact continuous-video boundaries." ] }, "model_output_raw": { "caption": "The person changes her arm position, crossing them over her chest. She then performs another sit-up, lifting her torso up before lowering it back down. Finally, she uncrosses her arms and rests them by her sides on the mat.", "entity_groups": [ { "id": "person_1", "entity": "person", "spans": [ { "text": "The person", "range": [ 0, 10 ] }, { "text": "her", "range": [ 19, 22 ] }, { "text": "her", "range": [ 56, 59 ] }, { "text": "She", "range": [ 67, 70 ] }, { "text": "her", "range": [ 109, 112 ] }, { "text": "she", "range": [ 161, 164 ] }, { "text": "her", "range": [ 175, 178 ] }, { "text": "her", "range": [ 202, 205 ] } ], "source_mask_link": "selected_person_masklet" }, { "id": "arms_1", "entity": "arms", "spans": [ { "text": "her arm", "range": [ 19, 26 ] }, { "text": "them", "range": [ 46, 50 ] }, { "text": "her arms", "range": [ 175, 183 ] }, { "text": "them", "range": [ 194, 198 ] } ], "source_mask_link": null }, { "id": "chest_1", "entity": "chest", "spans": [ { "text": "her chest", "range": [ 56, 65 ] } ], "source_mask_link": null }, { "id": "torso_1", "entity": "torso", "spans": [ { "text": "her torso", "range": [ 109, 118 ] }, { "text": "it", "range": [ 138, 140 ] } ], "source_mask_link": null }, { "id": "sides_1", "entity": "sides", "spans": [ { "text": "her sides", "range": [ 202, 211 ] } ], "source_mask_link": null }, { "id": "mat_1", "entity": "mat", "spans": [ { "text": "the mat", "range": [ 215, 222 ] } ], "source_mask_link": null } ], "actions": [ { "id": "action_1", "mentions": [ { "text": "changes", "ranges": [ [ 11, 18 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } }, { "text": "crossing", "ranges": [ [ 37, 45 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "type": "simple", "subject_ref": "person_1", "object_refs": [ "arms_1" ], "confidence": "high", "temporal_status": "localized", "intervals": [ { "start_s": 2.0, "end_s": 4.0, "source_start_s": 55.44, "source_end_s": 57.44 } ], "evidence": "At clip-relative 2.0 seconds the selected person is moving her arms toward her chest; by 3.0 seconds they are crossed over it.", "mask_relation": "subject_matches_selected_person_mask" }, { "id": "action_2", "mentions": [ { "text": "performs another sit-up", "ranges": [ [ 76, 99 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "type": "light_verb", "subject_ref": "person_1", "object_refs": [], "confidence": "high", "temporal_status": "localized", "intervals": [ { "start_s": 3.0, "end_s": 13.0, "source_start_s": 56.44, "source_end_s": 66.44 } ], "evidence": "Clean frames from 3.0 through 11.0 seconds show the selected person with her torso raised in a sit-up, followed by a return to the mat at 12.0 seconds.", "mask_relation": "subject_matches_selected_person_mask" }, { "id": "action_3", "mentions": [ { "text": "lifting", "ranges": [ [ 101, 108 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "type": "simple", "subject_ref": "person_1", "object_refs": [ "torso_1" ], "confidence": "high", "temporal_status": "localized", "intervals": [ { "start_s": 3.0, "end_s": 7.0, "source_start_s": 56.44, "source_end_s": 60.44 } ], "evidence": "From 3.0 through 6.0 seconds, the selected person's head, shoulders, and torso rise progressively away from the mat.", "mask_relation": "subject_matches_selected_person_mask" }, { "id": "action_4", "mentions": [ { "text": "lowering", "ranges": [ [ 129, 137 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "type": "simple", "subject_ref": "person_1", "object_refs": [ "torso_1" ], "confidence": "high", "temporal_status": "localized", "intervals": [ { "start_s": 7.0, "end_s": 13.0, "source_start_s": 60.44, "source_end_s": 66.44 } ], "evidence": "Across the later sampled frames, the selected person's raised torso returns toward the mat and is fully down by 12.0 seconds.", "mask_relation": "subject_matches_selected_person_mask" }, { "id": "action_5", "mentions": [ { "text": "uncrosses", "ranges": [ [ 165, 174 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "type": "simple", "subject_ref": "person_1", "object_refs": [ "arms_1" ], "confidence": "high", "temporal_status": "localized", "intervals": [ { "start_s": 11.0, "end_s": 13.0, "source_start_s": 64.44, "source_end_s": 66.44 } ], "evidence": "The selected person's arms are still crossed at 11.0 seconds and are uncrossed beside her body at 12.0 seconds.", "mask_relation": "subject_matches_selected_person_mask" }, { "id": "action_6", "mentions": [ { "text": "rests", "ranges": [ [ 188, 193 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "type": "posture", "subject_ref": "person_1", "object_refs": [ "arms_1" ], "location_refs": [ "sides_1", "mat_1" ], "confidence": "high", "temporal_status": "localized", "intervals": [ { "start_s": 12.0, "end_s": 14.13, "source_start_s": 65.44, "source_end_s": 67.57 } ], "evidence": "At 12.0, 13.0, and 14.0 seconds, the selected person lies on the mat with her arms resting alongside her body; the posture continues to the clip boundary.", "mask_relation": "subject_matches_selected_person_mask" } ], "notes": [ "The supplied masklet consistently identifies the person lying on the mat, who is the caption's action subject.", "The kneeling instructor is visible but is not selected by the supplied mask and is not mentioned as an action subject in the caption.", "No separate masks were supplied for the arms, chest, torso, sides, or mat; their source-mask links are therefore null.", "Temporal intervals are sparse support intervals based on the sampled clean RGB frames, not exact continuous-video boundaries." ] }, "normalization_repairs": 0 }, { "source_video_id": "v_--veKG73Di4", "caption_count_category": 6, "clip": { "caption_index": 1, "caption": "A woman in a green t-shirt, white mini-skirt, and sunglasses stands while holding a white bottle. Her hair is styled in long, green-tinted braids. She smiles, raises her left hand to wave, and then lowers it, remaining mostly stationary.", "source_interval_sec": [ 0.0, 10.0 ], "loom_clip_id": "v_--veKG73Di4_1_300", "clip_frame_interval_4fps": [ 1, 300 ], "sample_count_1fps": 10, "sampled_frames": [ { "frame": "00000", "timestamp_sec": 0.0, "offset_sec": 0, "rgb_path": "assets/cat_06/v_--veKG73Di4_1_300/0000s_00000_rgb.jpg", "mask_path": "assets/cat_06/v_--veKG73Di4_1_300/0000s_00000_mask.png", "mask_status": "available" }, { "frame": "00028", "timestamp_sec": 1.0, "offset_sec": 1, "rgb_path": "assets/cat_06/v_--veKG73Di4_1_300/0001s_00028_rgb.jpg", "mask_path": "assets/cat_06/v_--veKG73Di4_1_300/0001s_00028_mask.png", "mask_status": "available" }, { "frame": "00060", "timestamp_sec": 2.0, "offset_sec": 2, "rgb_path": "assets/cat_06/v_--veKG73Di4_1_300/0002s_00060_rgb.jpg", "mask_path": "assets/cat_06/v_--veKG73Di4_1_300/0002s_00060_mask.png", "mask_status": "available" }, { "frame": "00088", "timestamp_sec": 3.0, "offset_sec": 3, "rgb_path": "assets/cat_06/v_--veKG73Di4_1_300/0003s_00088_rgb.jpg", "mask_path": "assets/cat_06/v_--veKG73Di4_1_300/0003s_00088_mask.png", "mask_status": "available" }, { "frame": "00120", "timestamp_sec": 4.0, "offset_sec": 4, "rgb_path": "assets/cat_06/v_--veKG73Di4_1_300/0004s_00120_rgb.jpg", "mask_path": "assets/cat_06/v_--veKG73Di4_1_300/0004s_00120_mask.png", "mask_status": "available" }, { "frame": "00148", "timestamp_sec": 5.0, "offset_sec": 5, "rgb_path": "assets/cat_06/v_--veKG73Di4_1_300/0005s_00148_rgb.jpg", "mask_path": "assets/cat_06/v_--veKG73Di4_1_300/0005s_00148_mask.png", "mask_status": "available" }, { "frame": "00180", "timestamp_sec": 6.0, "offset_sec": 6, "rgb_path": "assets/cat_06/v_--veKG73Di4_1_300/0006s_00180_rgb.jpg", "mask_path": "assets/cat_06/v_--veKG73Di4_1_300/0006s_00180_mask.png", "mask_status": "available" }, { "frame": "00208", "timestamp_sec": 7.0, "offset_sec": 7, "rgb_path": "assets/cat_06/v_--veKG73Di4_1_300/0007s_00208_rgb.jpg", "mask_path": "assets/cat_06/v_--veKG73Di4_1_300/0007s_00208_mask.png", "mask_status": "available" }, { "frame": "00240", "timestamp_sec": 8.0, "offset_sec": 8, "rgb_path": "assets/cat_06/v_--veKG73Di4_1_300/0008s_00240_rgb.jpg", "mask_path": "assets/cat_06/v_--veKG73Di4_1_300/0008s_00240_mask.png", "mask_status": "available" }, { "frame": "00268", "timestamp_sec": 9.0, "offset_sec": 9, "rgb_path": "assets/cat_06/v_--veKG73Di4_1_300/0009s_00268_rgb.jpg", "mask_path": "assets/cat_06/v_--veKG73Di4_1_300/0009s_00268_mask.png", "mask_status": "available" } ], "obj_id": [ 0 ], "anno_id": [ 1 ], "fps_inferred_from_clip_interval": 29.9 }, "usage": { "input_tokens": 25614, "input_tokens_details": { "cache_write_tokens": 25611, "cached_tokens": 0 }, "output_tokens": 4116, "output_tokens_details": { "reasoning_tokens": 2285 }, "total_tokens": 29730 }, "model": "gpt-5.6-sol", "effort": "medium", "custom_id": "cat06_cap01_v_--veKG73Di4_1_300", "annotation": { "caption": "A woman in a green t-shirt, white mini-skirt, and sunglasses stands while holding a white bottle. Her hair is styled in long, green-tinted braids. She smiles, raises her left hand to wave, and then lowers it, remaining mostly stationary.", "entity_groups": [ { "id": "woman_1", "label": "woman", "spans": [ { "text": "A woman", "range": [ 0, 7 ] }, { "text": "Her", "range": [ 98, 101 ] }, { "text": "She", "range": [ 147, 150 ] }, { "text": "her", "range": [ 166, 169 ] } ], "source_mask_link": "selected_person" }, { "id": "tshirt_1", "label": "t-shirt", "spans": [ { "text": "a green t-shirt", "range": [ 11, 26 ] } ], "source_mask_link": null }, { "id": "miniskirt_1", "label": "mini-skirt", "spans": [ { "text": "white mini-skirt", "range": [ 28, 44 ] } ], "source_mask_link": null }, { "id": "sunglasses_1", "label": "sunglasses", "spans": [ { "text": "sunglasses", "range": [ 50, 60 ] } ], "source_mask_link": null }, { "id": "bottle_1", "label": "bottle", "spans": [ { "text": "a white bottle", "range": [ 82, 96 ] } ], "source_mask_link": null }, { "id": "hair_1", "label": "hair/braids", "spans": [ { "text": "Her hair", "range": [ 98, 106 ] }, { "text": "long, green-tinted braids", "range": [ 120, 145 ] } ], "source_mask_link": null }, { "id": "left_hand_1", "label": "left hand", "spans": [ { "text": "her left hand", "range": [ 166, 179 ] }, { "text": "it", "range": [ 205, 207 ] } ], "source_mask_link": null } ], "actions": [ { "id": "action_1", "mentions": [ { "text": "stands", "ranges": [ [ 61, 67 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "type": "posture", "subject_refs": [ "woman_1" ], "object_refs": [], "confidence": "high", "temporal_status": "supported", "intervals": [ { "start_s": 0.0, "end_s": 10.0, "source_start_s": 0.0, "source_end_s": 10.0 } ], "evidence": "The mask-linked woman remains upright in every sampled clean frame from 0 through 9 seconds.", "mask_relation": "subject matches the supplied selected-person mask", "source_subject_refs": "woman_1" }, { "id": "action_2", "mentions": [ { "text": "holding", "ranges": [ [ 74, 81 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "type": "simple", "subject_refs": [ "woman_1" ], "object_refs": [ "bottle_1" ], "confidence": "high", "temporal_status": "supported", "intervals": [ { "start_s": 0.0, "end_s": 10.0, "source_start_s": 0.0, "source_end_s": 10.0 } ], "evidence": "The selected woman visibly holds the white bottle throughout all sampled frames, including the final sample.", "mask_relation": "subject matches the supplied selected-person mask", "source_subject_refs": "woman_1" }, { "id": "action_3", "mentions": [ { "text": "smiles", "ranges": [ [ 151, 157 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "type": "simple", "subject_refs": [ "woman_1" ], "object_refs": [], "confidence": "high", "temporal_status": "supported", "intervals": [ { "start_s": 0.0, "end_s": 10.0, "source_start_s": 0.0, "source_end_s": 10.0 } ], "evidence": "The mask-linked woman displays a visible smile across the sampled sequence.", "mask_relation": "subject matches the supplied selected-person mask", "source_subject_refs": "woman_1" }, { "id": "action_4", "mentions": [ { "text": "raises", "ranges": [ [ 159, 165 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "type": "simple", "subject_refs": [ "woman_1" ], "object_refs": [ "left_hand_1" ], "confidence": "high", "temporal_status": "localized_sparse", "intervals": [ { "start_s": 2.0, "end_s": 3.0, "source_start_s": 2.0, "source_end_s": 3.0 } ], "evidence": "Her free hand changes from lowered in the earlier samples to raised above her head at 2 seconds.", "mask_relation": "subject matches the supplied selected-person mask", "source_subject_refs": "woman_1" }, { "id": "action_5", "mentions": [ { "text": "wave", "ranges": [ [ 183, 187 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "type": "simple", "subject_refs": [ "woman_1" ], "object_refs": [], "confidence": "high", "temporal_status": "localized_sparse", "intervals": [ { "start_s": 2.0, "end_s": 4.0, "source_start_s": 2.0, "source_end_s": 4.0 } ], "evidence": "At 2 and 3 seconds, the selected woman holds her open hand overhead in a visible waving gesture.", "mask_relation": "subject matches the supplied selected-person mask", "source_subject_refs": "woman_1" }, { "id": "action_6", "mentions": [ { "text": "lowers", "ranges": [ [ 198, 204 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "type": "simple", "subject_refs": [ "woman_1" ], "object_refs": [ "left_hand_1" ], "confidence": "high", "temporal_status": "localized_sparse", "intervals": [ { "start_s": 4.0, "end_s": 5.0, "source_start_s": 4.0, "source_end_s": 5.0 } ], "evidence": "The hand that is overhead at 3 seconds is visibly lowered beside her body at 4 seconds.", "mask_relation": "subject matches the supplied selected-person mask", "source_subject_refs": "woman_1" } ], "notes": [ "The supplied masklet consistently identifies the woman in the green t-shirt; the other visible person is not selected.", "No separate source masks were supplied for the clothing, sunglasses, bottle, hair/braids, or left hand, so those entity groups have null source-mask links.", "\"is styled\" describes the hair's resulting hairstyle and is not treated as an eventive action. Its subject also lacks a supplied mask.", "\"remaining mostly stationary\" is not separately extracted because it frames a static lack of motion rather than an additional physical action predicate." ] }, "model_output_raw": { "caption": "A woman in a green t-shirt, white mini-skirt, and sunglasses stands while holding a white bottle. Her hair is styled in long, green-tinted braids. She smiles, raises her left hand to wave, and then lowers it, remaining mostly stationary.", "entity_groups": [ { "id": "woman_1", "entity": "woman", "spans": [ { "text": "A woman", "range": [ 0, 7 ] }, { "text": "Her", "range": [ 98, 101 ] }, { "text": "She", "range": [ 147, 150 ] }, { "text": "her", "range": [ 166, 169 ] } ], "source_mask": "selected_person_masklet" }, { "id": "tshirt_1", "entity": "t-shirt", "spans": [ { "text": "a green t-shirt", "range": [ 11, 26 ] } ], "source_mask": null }, { "id": "miniskirt_1", "entity": "mini-skirt", "spans": [ { "text": "white mini-skirt", "range": [ 28, 44 ] } ], "source_mask": null }, { "id": "sunglasses_1", "entity": "sunglasses", "spans": [ { "text": "sunglasses", "range": [ 50, 60 ] } ], "source_mask": null }, { "id": "bottle_1", "entity": "bottle", "spans": [ { "text": "a white bottle", "range": [ 82, 96 ] } ], "source_mask": null }, { "id": "hair_1", "entity": "hair/braids", "spans": [ { "text": "Her hair", "range": [ 98, 106 ] }, { "text": "long, green-tinted braids", "range": [ 120, 145 ] } ], "source_mask": null }, { "id": "left_hand_1", "entity": "left hand", "spans": [ { "text": "her left hand", "range": [ 166, 179 ] }, { "text": "it", "range": [ 205, 207 ] } ], "source_mask": null } ], "actions": [ { "id": "action_1", "mentions": [ { "text": "stands", "ranges": [ [ 61, 67 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "type": "posture", "subject": "woman_1", "objects": [], "confidence": "high", "temporal_status": "supported", "intervals": [ { "start_s": 0.0, "end_s": 10.0 } ], "evidence": "The mask-linked woman remains upright in every sampled clean frame from 0 through 9 seconds.", "mask_relation": "subject matches the supplied selected-person mask" }, { "id": "action_2", "mentions": [ { "text": "holding", "ranges": [ [ 74, 81 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "type": "simple", "subject": "woman_1", "objects": [ "bottle_1" ], "confidence": "high", "temporal_status": "supported", "intervals": [ { "start_s": 0.0, "end_s": 10.0 } ], "evidence": "The selected woman visibly holds the white bottle throughout all sampled frames, including the final sample.", "mask_relation": "subject matches the supplied selected-person mask" }, { "id": "action_3", "mentions": [ { "text": "smiles", "ranges": [ [ 151, 157 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "type": "simple", "subject": "woman_1", "objects": [], "confidence": "high", "temporal_status": "supported", "intervals": [ { "start_s": 0.0, "end_s": 10.0 } ], "evidence": "The mask-linked woman displays a visible smile across the sampled sequence.", "mask_relation": "subject matches the supplied selected-person mask" }, { "id": "action_4", "mentions": [ { "text": "raises", "ranges": [ [ 159, 165 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "type": "simple", "subject": "woman_1", "objects": [ "left_hand_1" ], "confidence": "high", "temporal_status": "localized_sparse", "intervals": [ { "start_s": 2.0, "end_s": 3.0 } ], "evidence": "Her free hand changes from lowered in the earlier samples to raised above her head at 2 seconds.", "mask_relation": "subject matches the supplied selected-person mask" }, { "id": "action_5", "mentions": [ { "text": "wave", "ranges": [ [ 183, 187 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "type": "simple", "subject": "woman_1", "objects": [], "confidence": "high", "temporal_status": "localized_sparse", "intervals": [ { "start_s": 2.0, "end_s": 4.0 } ], "evidence": "At 2 and 3 seconds, the selected woman holds her open hand overhead in a visible waving gesture.", "mask_relation": "subject matches the supplied selected-person mask" }, { "id": "action_6", "mentions": [ { "text": "lowers", "ranges": [ [ 198, 204 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "type": "simple", "subject": "woman_1", "objects": [ "left_hand_1" ], "confidence": "high", "temporal_status": "localized_sparse", "intervals": [ { "start_s": 4.0, "end_s": 5.0 } ], "evidence": "The hand that is overhead at 3 seconds is visibly lowered beside her body at 4 seconds.", "mask_relation": "subject matches the supplied selected-person mask" } ], "notes": [ "The supplied masklet consistently identifies the woman in the green t-shirt; the other visible person is not selected.", "No separate source masks were supplied for the clothing, sunglasses, bottle, hair/braids, or left hand, so those entity groups have null source-mask links.", "\"is styled\" describes the hair's resulting hairstyle and is not treated as an eventive action. Its subject also lacks a supplied mask.", "\"remaining mostly stationary\" is not separately extracted because it frames a static lack of motion rather than an additional physical action predicate." ] }, "normalization_repairs": 0 }, { "source_video_id": "v_--veKG73Di4", "caption_count_category": 6, "clip": { "caption_index": 2, "caption": "She turns her body to the right to face the person next to her, gesturing with her right hand while speaking. She then turns back slightly and uses both hands to unscrew the cap from her bottle. She continues to hold the bottle and the cap separately.", "source_interval_sec": [ 10.0, 25.5 ], "loom_clip_id": "v_--veKG73Di4_300_764", "clip_frame_interval_4fps": [ 300, 764 ], "sample_count_1fps": 16, "sampled_frames": [ { "frame": "00000", "timestamp_sec": 10.0, "offset_sec": 0, "rgb_path": "assets/cat_06/v_--veKG73Di4_300_764/0000s_00000_rgb.jpg", "mask_path": "assets/cat_06/v_--veKG73Di4_300_764/0000s_00000_mask.png", "mask_status": "available" }, { "frame": "00028", "timestamp_sec": 11.0, "offset_sec": 1, "rgb_path": "assets/cat_06/v_--veKG73Di4_300_764/0001s_00028_rgb.jpg", "mask_path": "assets/cat_06/v_--veKG73Di4_300_764/0001s_00028_mask.png", "mask_status": "available" }, { "frame": "00060", "timestamp_sec": 12.0, "offset_sec": 2, "rgb_path": "assets/cat_06/v_--veKG73Di4_300_764/0002s_00060_rgb.jpg", "mask_path": "assets/cat_06/v_--veKG73Di4_300_764/0002s_00060_mask.png", "mask_status": "available" }, { "frame": "00088", "timestamp_sec": 13.0, "offset_sec": 3, "rgb_path": "assets/cat_06/v_--veKG73Di4_300_764/0003s_00088_rgb.jpg", "mask_path": "assets/cat_06/v_--veKG73Di4_300_764/0003s_00088_mask.png", "mask_status": "available" }, { "frame": "00120", "timestamp_sec": 14.0, "offset_sec": 4, "rgb_path": "assets/cat_06/v_--veKG73Di4_300_764/0004s_00120_rgb.jpg", "mask_path": "assets/cat_06/v_--veKG73Di4_300_764/0004s_00120_mask.png", "mask_status": "available" }, { "frame": "00148", "timestamp_sec": 15.0, "offset_sec": 5, "rgb_path": "assets/cat_06/v_--veKG73Di4_300_764/0005s_00148_rgb.jpg", "mask_path": "assets/cat_06/v_--veKG73Di4_300_764/0005s_00148_mask.png", "mask_status": "available" }, { "frame": "00180", "timestamp_sec": 16.0, "offset_sec": 6, "rgb_path": "assets/cat_06/v_--veKG73Di4_300_764/0006s_00180_rgb.jpg", "mask_path": "assets/cat_06/v_--veKG73Di4_300_764/0006s_00180_mask.png", "mask_status": "available" }, { "frame": "00208", "timestamp_sec": 17.0, "offset_sec": 7, "rgb_path": "assets/cat_06/v_--veKG73Di4_300_764/0007s_00208_rgb.jpg", "mask_path": "assets/cat_06/v_--veKG73Di4_300_764/0007s_00208_mask.png", "mask_status": "available" }, { "frame": "00240", "timestamp_sec": 18.0, "offset_sec": 8, "rgb_path": "assets/cat_06/v_--veKG73Di4_300_764/0008s_00240_rgb.jpg", "mask_path": "assets/cat_06/v_--veKG73Di4_300_764/0008s_00240_mask.png", "mask_status": "available" }, { "frame": "00268", "timestamp_sec": 19.0, "offset_sec": 9, "rgb_path": "assets/cat_06/v_--veKG73Di4_300_764/0009s_00268_rgb.jpg", "mask_path": "assets/cat_06/v_--veKG73Di4_300_764/0009s_00268_mask.png", "mask_status": "available" }, { "frame": "00300", "timestamp_sec": 20.0, "offset_sec": 10, "rgb_path": "assets/cat_06/v_--veKG73Di4_300_764/0010s_00300_rgb.jpg", "mask_path": "assets/cat_06/v_--veKG73Di4_300_764/0010s_00300_mask.png", "mask_status": "available" }, { "frame": "00328", "timestamp_sec": 21.0, "offset_sec": 11, "rgb_path": "assets/cat_06/v_--veKG73Di4_300_764/0011s_00328_rgb.jpg", "mask_path": "assets/cat_06/v_--veKG73Di4_300_764/0011s_00328_mask.png", "mask_status": "available" }, { "frame": "00360", "timestamp_sec": 22.0, "offset_sec": 12, "rgb_path": "assets/cat_06/v_--veKG73Di4_300_764/0012s_00360_rgb.jpg", "mask_path": "assets/cat_06/v_--veKG73Di4_300_764/0012s_00360_mask.png", "mask_status": "available" }, { "frame": "00388", "timestamp_sec": 23.0, "offset_sec": 13, "rgb_path": "assets/cat_06/v_--veKG73Di4_300_764/0013s_00388_rgb.jpg", "mask_path": "assets/cat_06/v_--veKG73Di4_300_764/0013s_00388_mask.png", "mask_status": "available" }, { "frame": "00420", "timestamp_sec": 24.0, "offset_sec": 14, "rgb_path": "assets/cat_06/v_--veKG73Di4_300_764/0014s_00420_rgb.jpg", "mask_path": "assets/cat_06/v_--veKG73Di4_300_764/0014s_00420_mask.png", "mask_status": "available" }, { "frame": "00448", "timestamp_sec": 25.0, "offset_sec": 15, "rgb_path": "assets/cat_06/v_--veKG73Di4_300_764/0015s_00448_rgb.jpg", "mask_path": "assets/cat_06/v_--veKG73Di4_300_764/0015s_00448_mask.png", "mask_status": "available" } ], "obj_id": [ 0 ], "anno_id": [ 3 ], "fps_inferred_from_clip_interval": 29.9355 }, "usage": { "input_tokens": 31775, "input_tokens_details": { "cache_write_tokens": 31772, "cached_tokens": 0 }, "output_tokens": 5187, "output_tokens_details": { "reasoning_tokens": 2900 }, "total_tokens": 36962 }, "model": "gpt-5.6-sol", "effort": "medium", "custom_id": "cat06_cap02_v_--veKG73Di4_300_764", "annotation": { "caption": "She turns her body to the right to face the person next to her, gesturing with her right hand while speaking. She then turns back slightly and uses both hands to unscrew the cap from her bottle. She continues to hold the bottle and the cap separately.", "entity_groups": [ { "id": "E1", "label": "woman", "spans": [ { "text": "She", "range": [ 0, 3 ] }, { "text": "her", "range": [ 10, 13 ] }, { "text": "her", "range": [ 59, 62 ] }, { "text": "her", "range": [ 79, 82 ] }, { "text": "She", "range": [ 110, 113 ] }, { "text": "her", "range": [ 183, 186 ] }, { "text": "She", "range": [ 195, 198 ] } ], "source_mask_link": "selected_person" }, { "id": "E2", "label": "body", "spans": [ { "text": "her body", "range": [ 10, 18 ] } ], "source_mask_link": null }, { "id": "E3", "label": "person", "spans": [ { "text": "the person", "range": [ 40, 50 ] } ], "source_mask_link": null }, { "id": "E4", "label": "right hand", "spans": [ { "text": "her right hand", "range": [ 79, 93 ] }, { "text": "both hands", "range": [ 148, 158 ] } ], "source_mask_link": null }, { "id": "E5", "label": "left hand", "spans": [ { "text": "both hands", "range": [ 148, 158 ] } ], "source_mask_link": null }, { "id": "E6", "label": "bottle", "spans": [ { "text": "her bottle", "range": [ 183, 193 ] }, { "text": "the bottle", "range": [ 217, 227 ] } ], "source_mask_link": null }, { "id": "E7", "label": "cap", "spans": [ { "text": "the cap", "range": [ 170, 177 ] }, { "text": "the cap", "range": [ 232, 239 ] } ], "source_mask_link": null } ], "actions": [ { "id": "A1", "mentions": [ { "text": "turns", "ranges": [ [ 4, 9 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "type": "simple", "subject_refs": [ "E1" ], "object_refs": [ "E2" ], "confidence": "high", "temporal_status": "localized", "intervals": [ { "start_s": 0.0, "end_s": 2.0, "source_start_s": 10.0, "source_end_s": 12.0 } ], "evidence": "From samples 0–2, the masked woman changes from a more forward-facing orientation to facing the adjacent person.", "mask_relation": "subject matches the supplied selected-person mask", "source_subject_refs": "E1" }, { "id": "A2", "mentions": [ { "text": "face", "ranges": [ [ 35, 39 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "type": "posture", "subject_refs": [ "E1" ], "object_refs": [ "E3" ], "confidence": "high", "temporal_status": "localized", "intervals": [ { "start_s": 1.0, "end_s": 11.0, "source_start_s": 11.0, "source_end_s": 21.0 } ], "evidence": "In samples 1–10, the masked woman's head and torso are oriented toward the person beside her.", "mask_relation": "subject matches the supplied selected-person mask", "source_subject_refs": "E1" }, { "id": "A3", "mentions": [ { "text": "gesturing", "ranges": [ [ 64, 73 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "type": "simple", "subject_refs": [ "E1" ], "object_refs": [ "E4" ], "confidence": "high", "temporal_status": "localized", "intervals": [ { "start_s": 3.0, "end_s": 11.0, "source_start_s": 13.0, "source_end_s": 21.0 } ], "evidence": "Samples 3–10 show the masked woman repeatedly extending or raising a hand while interacting with the adjacent person.", "mask_relation": "subject matches the supplied selected-person mask", "source_subject_refs": "E1" }, { "id": "A4", "mentions": [ { "text": "speaking", "ranges": [ [ 100, 108 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "type": "simple", "subject_refs": [ "E1" ], "object_refs": [], "confidence": "medium", "temporal_status": "uncertain", "intervals": [], "evidence": "The masked woman is engaged with the adjacent person, but speech and mouth movement cannot be reliably verified from the sparse sampled frames.", "mask_relation": "subject matches the supplied selected-person mask", "source_subject_refs": "E1" }, { "id": "A5", "mentions": [ { "text": "turns", "ranges": [ [ 119, 124 ] ], "occurrence": 2, "particle": { "text": null, "occurrence": 1 } } ], "type": "simple", "subject_refs": [ "E1" ], "object_refs": [], "confidence": "medium", "temporal_status": "localized", "intervals": [ { "start_s": 10.0, "end_s": 12.0, "source_start_s": 20.0, "source_end_s": 22.0 } ], "evidence": "Between samples 10 and 11, the masked woman shifts from facing the other person toward a slightly more forward and downward orientation.", "mask_relation": "subject matches the supplied selected-person mask", "source_subject_refs": "E1" }, { "id": "A6", "mentions": [ { "text": "uses", "ranges": [ [ 143, 147 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "type": "simple", "subject_refs": [ "E1" ], "object_refs": [ "E4", "E5" ], "confidence": "high", "temporal_status": "localized", "intervals": [ { "start_s": 11.0, "end_s": 15.0, "source_start_s": 21.0, "source_end_s": 25.0 } ], "evidence": "Samples 11–14 show the masked woman employing both hands around the bottle and its top.", "mask_relation": "subject matches the supplied selected-person mask", "source_subject_refs": "E1" }, { "id": "A7", "mentions": [ { "text": "unscrew", "ranges": [ [ 162, 169 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "type": "simple", "subject_refs": [ "E1" ], "object_refs": [ "E7", "E6" ], "confidence": "high", "temporal_status": "localized", "intervals": [ { "start_s": 11.0, "end_s": 15.0, "source_start_s": 21.0, "source_end_s": 25.0 } ], "evidence": "In samples 11–14, one hand stabilizes the bottle while the other manipulates the cap at its opening.", "mask_relation": "subject matches the supplied selected-person mask", "source_subject_refs": "E1" }, { "id": "A8", "mentions": [ { "text": "hold", "ranges": [ [ 212, 216 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "type": "simple", "subject_refs": [ "E1" ], "object_refs": [ "E6", "E7" ], "confidence": "high", "temporal_status": "localized", "intervals": [ { "start_s": 15.0, "end_s": 15.5, "source_start_s": 25.0, "source_end_s": 25.5 } ], "evidence": "At sample 15, the masked woman holds the bottle while her other hand is extended with the removed cap; support reaches the clip boundary.", "mask_relation": "subject matches the supplied selected-person mask", "source_subject_refs": "E1" } ], "notes": [ "The supplied masklet consistently identifies the woman in the green shirt, corresponding to the caption's three occurrences of “She.”", "The adjacent person is visible but has no supplied mask, so that person's actions are not inferred or annotated.", "No separate masks were supplied for the body, hands, bottle, or cap; their source-mask links are null.", "“both hands” is shared by the right-hand and left-hand entity groups, while the earlier specific phrase “her right hand” identifies the right-hand instance.", "The aspectual verb “continues” and directional or manner modifiers such as “to the right,” “back,” “slightly,” and “separately” are excluded from action spans." ] }, "model_output_raw": { "caption": "She turns her body to the right to face the person next to her, gesturing with her right hand while speaking. She then turns back slightly and uses both hands to unscrew the cap from her bottle. She continues to hold the bottle and the cap separately.", "entity_groups": [ { "id": "E1", "label": "woman", "spans": [ { "text": "She", "range": [ 0, 3 ] }, { "text": "her", "range": [ 10, 13 ] }, { "text": "her", "range": [ 59, 62 ] }, { "text": "her", "range": [ 79, 82 ] }, { "text": "She", "range": [ 110, 113 ] }, { "text": "her", "range": [ 183, 186 ] }, { "text": "She", "range": [ 195, 198 ] } ], "source_mask": "selected_person_masklet" }, { "id": "E2", "label": "body", "spans": [ { "text": "her body", "range": [ 10, 18 ] } ], "source_mask": null }, { "id": "E3", "label": "person", "spans": [ { "text": "the person", "range": [ 40, 50 ] } ], "source_mask": null }, { "id": "E4", "label": "right hand", "spans": [ { "text": "her right hand", "range": [ 79, 93 ] }, { "text": "both hands", "range": [ 148, 158 ] } ], "source_mask": null }, { "id": "E5", "label": "left hand", "spans": [ { "text": "both hands", "range": [ 148, 158 ] } ], "source_mask": null }, { "id": "E6", "label": "bottle", "spans": [ { "text": "her bottle", "range": [ 183, 193 ] }, { "text": "the bottle", "range": [ 217, 227 ] } ], "source_mask": null }, { "id": "E7", "label": "cap", "spans": [ { "text": "the cap", "range": [ 170, 177 ] }, { "text": "the cap", "range": [ 232, 239 ] } ], "source_mask": null } ], "actions": [ { "id": "A1", "mentions": [ { "text": "turns", "ranges": [ [ 4, 9 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "type": "simple", "subject": "E1", "objects": [ "E2" ], "confidence": "high", "mask_relation": "subject matches the supplied selected-person mask", "temporal_status": "localized", "intervals": [ { "start_s": 0.0, "end_s": 2.0 } ], "evidence": "From samples 0–2, the masked woman changes from a more forward-facing orientation to facing the adjacent person." }, { "id": "A2", "mentions": [ { "text": "face", "ranges": [ [ 35, 39 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "type": "posture", "subject": "E1", "objects": [ "E3" ], "confidence": "high", "mask_relation": "subject matches the supplied selected-person mask", "temporal_status": "localized", "intervals": [ { "start_s": 1.0, "end_s": 11.0 } ], "evidence": "In samples 1–10, the masked woman's head and torso are oriented toward the person beside her." }, { "id": "A3", "mentions": [ { "text": "gesturing", "ranges": [ [ 64, 73 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "type": "simple", "subject": "E1", "objects": [ "E4" ], "confidence": "high", "mask_relation": "subject matches the supplied selected-person mask", "temporal_status": "localized", "intervals": [ { "start_s": 3.0, "end_s": 11.0 } ], "evidence": "Samples 3–10 show the masked woman repeatedly extending or raising a hand while interacting with the adjacent person." }, { "id": "A4", "mentions": [ { "text": "speaking", "ranges": [ [ 100, 108 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "type": "simple", "subject": "E1", "objects": [], "confidence": "medium", "mask_relation": "subject matches the supplied selected-person mask", "temporal_status": "uncertain", "intervals": [], "evidence": "The masked woman is engaged with the adjacent person, but speech and mouth movement cannot be reliably verified from the sparse sampled frames." }, { "id": "A5", "mentions": [ { "text": "turns", "ranges": [ [ 119, 124 ] ], "occurrence": 2, "particle": { "text": null, "occurrence": 1 } } ], "type": "simple", "subject": "E1", "objects": [], "confidence": "medium", "mask_relation": "subject matches the supplied selected-person mask", "temporal_status": "localized", "intervals": [ { "start_s": 10.0, "end_s": 12.0 } ], "evidence": "Between samples 10 and 11, the masked woman shifts from facing the other person toward a slightly more forward and downward orientation." }, { "id": "A6", "mentions": [ { "text": "uses", "ranges": [ [ 143, 147 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "type": "simple", "subject": "E1", "objects": [ "E4", "E5" ], "confidence": "high", "mask_relation": "subject matches the supplied selected-person mask", "temporal_status": "localized", "intervals": [ { "start_s": 11.0, "end_s": 15.0 } ], "evidence": "Samples 11–14 show the masked woman employing both hands around the bottle and its top." }, { "id": "A7", "mentions": [ { "text": "unscrew", "ranges": [ [ 162, 169 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "type": "simple", "subject": "E1", "objects": [ "E7", "E6" ], "confidence": "high", "mask_relation": "subject matches the supplied selected-person mask", "temporal_status": "localized", "intervals": [ { "start_s": 11.0, "end_s": 15.0 } ], "evidence": "In samples 11–14, one hand stabilizes the bottle while the other manipulates the cap at its opening." }, { "id": "A8", "mentions": [ { "text": "hold", "ranges": [ [ 212, 216 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "type": "simple", "subject": "E1", "objects": [ "E6", "E7" ], "confidence": "high", "mask_relation": "subject matches the supplied selected-person mask", "temporal_status": "localized", "intervals": [ { "start_s": 15.0, "end_s": 15.5 } ], "evidence": "At sample 15, the masked woman holds the bottle while her other hand is extended with the removed cap; support reaches the clip boundary." } ], "notes": [ "The supplied masklet consistently identifies the woman in the green shirt, corresponding to the caption's three occurrences of “She.”", "The adjacent person is visible but has no supplied mask, so that person's actions are not inferred or annotated.", "No separate masks were supplied for the body, hands, bottle, or cap; their source-mask links are null.", "“both hands” is shared by the right-hand and left-hand entity groups, while the earlier specific phrase “her right hand” identifies the right-hand instance.", "The aspectual verb “continues” and directional or manner modifiers such as “to the right,” “back,” “slightly,” and “separately” are excluded from action spans." ] }, "normalization_repairs": 0 }, { "source_video_id": "v_--veKG73Di4", "caption_count_category": 6, "clip": { "caption_index": 3, "caption": "After a brief interaction with another person who enters from the right, she turns back and begins clapping her hands rhythmically. She sways slightly while clapping.", "source_interval_sec": [ 25.5, 32.0 ], "loom_clip_id": "v_--veKG73Di4_764_959", "clip_frame_interval_4fps": [ 764, 959 ], "sample_count_1fps": 7, "sampled_frames": [ { "frame": "00000", "timestamp_sec": 25.5, "offset_sec": 0, "rgb_path": "assets/cat_06/v_--veKG73Di4_764_959/0000s_00000_rgb.jpg", "mask_path": "assets/cat_06/v_--veKG73Di4_764_959/0000s_00000_mask.png", "mask_status": "available" }, { "frame": "00028", "timestamp_sec": 26.5, "offset_sec": 1, "rgb_path": "assets/cat_06/v_--veKG73Di4_764_959/0001s_00028_rgb.jpg", "mask_path": "assets/cat_06/v_--veKG73Di4_764_959/0001s_00028_mask.png", "mask_status": "available" }, { "frame": "00060", "timestamp_sec": 27.5, "offset_sec": 2, "rgb_path": "assets/cat_06/v_--veKG73Di4_764_959/0002s_00060_rgb.jpg", "mask_path": "assets/cat_06/v_--veKG73Di4_764_959/0002s_00060_mask.png", "mask_status": "available" }, { "frame": "00088", "timestamp_sec": 28.5, "offset_sec": 3, "rgb_path": "assets/cat_06/v_--veKG73Di4_764_959/0003s_00088_rgb.jpg", "mask_path": "assets/cat_06/v_--veKG73Di4_764_959/0003s_00088_mask.png", "mask_status": "available" }, { "frame": "00120", "timestamp_sec": 29.5, "offset_sec": 4, "rgb_path": "assets/cat_06/v_--veKG73Di4_764_959/0004s_00120_rgb.jpg", "mask_path": "assets/cat_06/v_--veKG73Di4_764_959/0004s_00120_mask.png", "mask_status": "available" }, { "frame": "00148", "timestamp_sec": 30.5, "offset_sec": 5, "rgb_path": "assets/cat_06/v_--veKG73Di4_764_959/0005s_00148_rgb.jpg", "mask_path": "assets/cat_06/v_--veKG73Di4_764_959/0005s_00148_mask.png", "mask_status": "available" }, { "frame": "00180", "timestamp_sec": 31.5, "offset_sec": 6, "rgb_path": "assets/cat_06/v_--veKG73Di4_764_959/0006s_00180_rgb.jpg", "mask_path": "assets/cat_06/v_--veKG73Di4_764_959/0006s_00180_mask.png", "mask_status": "available" } ], "obj_id": [ 0 ], "anno_id": [ 4 ], "fps_inferred_from_clip_interval": 30.0 }, "usage": { "input_tokens": 22512, "input_tokens_details": { "cache_write_tokens": 22509, "cached_tokens": 0 }, "output_tokens": 3721, "output_tokens_details": { "reasoning_tokens": 2588 }, "total_tokens": 26233 }, "model": "gpt-5.6-sol", "effort": "medium", "custom_id": "cat06_cap03_v_--veKG73Di4_764_959", "annotation": { "caption": "After a brief interaction with another person who enters from the right, she turns back and begins clapping her hands rhythmically. She sways slightly while clapping.", "entity_groups": [ { "id": "entity_1", "label": "another person", "spans": [ { "text": "another person", "range": [ 31, 45 ] } ], "source_mask_link": null }, { "id": "entity_2", "label": "selected person", "spans": [ { "text": "she", "range": [ 73, 76 ] }, { "text": "her", "range": [ 108, 111 ] }, { "text": "She", "range": [ 132, 135 ] } ], "source_mask_link": "selected_person" }, { "id": "entity_3", "label": "hands", "spans": [ { "text": "her hands", "range": [ 108, 117 ] } ], "source_mask_link": null } ], "actions": [ { "id": "action_1", "mentions": [ { "text": "turns back", "ranges": [ [ 77, 87 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "type": "phrasal", "subject_refs": [ "entity_2" ], "object_refs": [], "confidence": "medium", "temporal_status": "localized_sparse", "intervals": [ { "start_s": 2.0, "end_s": 3.0, "source_start_s": 27.5, "source_end_s": 28.5 } ], "evidence": "Between the 2.0 s and 3.0 s samples, the masked woman reorients after the person entering from the right passes beside her.", "mask_relation": "subject_matches_supplied_selected_person_mask", "source_subject_refs": [ "entity_2" ] }, { "id": "action_2", "mentions": [ { "text": "clapping", "ranges": [ [ 99, 107 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } }, { "text": "clapping", "ranges": [ [ 157, 165 ] ], "occurrence": 2, "particle": { "text": null, "occurrence": 1 } } ], "type": "simple", "subject_refs": [ "entity_2" ], "object_refs": [ "entity_3" ], "confidence": "medium", "temporal_status": "localized_sparse", "intervals": [ { "start_s": 3.0, "end_s": 6.5, "source_start_s": 28.5, "source_end_s": 32.0 } ], "evidence": "From 3.0 s through the final 6.0 s sample, the masked woman repeatedly changes the positions of her raised hands; sparse one-second sampling does not capture every hand contact.", "mask_relation": "subject_matches_supplied_selected_person_mask", "source_subject_refs": [ "entity_2" ] }, { "id": "action_3", "mentions": [ { "text": "sways", "ranges": [ [ 136, 141 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "type": "simple", "subject_refs": [ "entity_2" ], "object_refs": [], "confidence": "medium", "temporal_status": "localized_sparse", "intervals": [ { "start_s": 3.0, "end_s": 6.5, "source_start_s": 28.5, "source_end_s": 32.0 } ], "evidence": "The masked woman's torso and stance shift slightly across the 3.0–6.0 s samples while her hands remain active.", "mask_relation": "subject_matches_supplied_selected_person_mask", "source_subject_refs": [ "entity_2" ] } ], "notes": [ "The supplied mask consistently identifies the woman in the green shirt, corresponding to “she,” “her,” and “She.”", "The action “enters” has the unmasked “another person” as its subject and is therefore omitted from actions.", "No separate source masks are supplied for the other person or the selected woman's hands.", "The source-video interval is [25.5, 32.0) seconds, corresponding to clip-relative [0.0, 6.5) seconds.", "Clapping and swaying localization is approximate because the frames are sampled sparsely at nominal one-second intervals." ] }, "model_output_raw": { "caption": "After a brief interaction with another person who enters from the right, she turns back and begins clapping her hands rhythmically. She sways slightly while clapping.", "entity_groups": [ { "id": "entity_1", "label": "another person", "spans": [ { "text": "another person", "start": 31, "end": 45 } ], "source_mask": null }, { "id": "entity_2", "label": "selected person", "spans": [ { "text": "she", "start": 73, "end": 76 }, { "text": "her", "start": 108, "end": 111 }, { "text": "She", "start": 132, "end": 135 } ], "source_mask": "supplied_selected_person_masklet" }, { "id": "entity_3", "label": "hands", "spans": [ { "text": "her hands", "start": 108, "end": 117 } ], "source_mask": null } ], "actions": [ { "mentions": [ { "text": "turns back", "occurrence": 1, "ranges": [ { "start": 77, "end": 87 } ], "particle": { "text": null, "occurrence": 1 } } ], "type": "phrasal", "subject_refs": [ "entity_2" ], "object_refs": [], "confidence": "medium", "temporal_status": "localized_sparse", "intervals": [ { "start_s": 2.0, "end_s": 3.0 } ], "evidence": "Between the 2.0 s and 3.0 s samples, the masked woman reorients after the person entering from the right passes beside her.", "mask_relation": "subject_matches_supplied_selected_person_mask" }, { "mentions": [ { "text": "clapping", "occurrence": 1, "ranges": [ { "start": 99, "end": 107 } ], "particle": { "text": null, "occurrence": 1 } }, { "text": "clapping", "occurrence": 2, "ranges": [ { "start": 157, "end": 165 } ], "particle": { "text": null, "occurrence": 1 } } ], "type": "simple", "subject_refs": [ "entity_2" ], "object_refs": [ "entity_3" ], "confidence": "medium", "temporal_status": "localized_sparse", "intervals": [ { "start_s": 3.0, "end_s": 6.5 } ], "evidence": "From 3.0 s through the final 6.0 s sample, the masked woman repeatedly changes the positions of her raised hands; sparse one-second sampling does not capture every hand contact.", "mask_relation": "subject_matches_supplied_selected_person_mask" }, { "mentions": [ { "text": "sways", "occurrence": 1, "ranges": [ { "start": 136, "end": 141 } ], "particle": { "text": null, "occurrence": 1 } } ], "type": "simple", "subject_refs": [ "entity_2" ], "object_refs": [], "confidence": "medium", "temporal_status": "localized_sparse", "intervals": [ { "start_s": 3.0, "end_s": 6.5 } ], "evidence": "The masked woman's torso and stance shift slightly across the 3.0–6.0 s samples while her hands remain active.", "mask_relation": "subject_matches_supplied_selected_person_mask" } ], "notes": [ "The supplied mask consistently identifies the woman in the green shirt, corresponding to “she,” “her,” and “She.”", "The action “enters” has the unmasked “another person” as its subject and is therefore omitted from actions.", "No separate source masks are supplied for the other person or the selected woman's hands.", "The source-video interval is [25.5, 32.0) seconds, corresponding to clip-relative [0.0, 6.5) seconds.", "Clapping and swaying localization is approximate because the frames are sampled sparsely at nominal one-second intervals." ] }, "normalization_repairs": 0 }, { "source_video_id": "v_--veKG73Di4", "caption_count_category": 6, "clip": { "caption_index": 4, "caption": "She begins dancing energetically, moving her hips and arms while still holding the bottle and cap. Her movements include raising her arms, turning in a circle, and swaying from side to side. Her braids swing as she dances.", "source_interval_sec": [ 32.0, 58.5 ], "loom_clip_id": "v_--veKG73Di4_959_1753", "clip_frame_interval_4fps": [ 959, 1753 ], "sample_count_1fps": 27, "sampled_frames": [ { "frame": "00000", "timestamp_sec": 32.0, "offset_sec": 0, "rgb_path": "assets/cat_06/v_--veKG73Di4_959_1753/0000s_00000_rgb.jpg", "mask_path": "assets/cat_06/v_--veKG73Di4_959_1753/0000s_00000_mask.png", "mask_status": "available" }, { "frame": "00028", "timestamp_sec": 33.0, "offset_sec": 1, "rgb_path": "assets/cat_06/v_--veKG73Di4_959_1753/0001s_00028_rgb.jpg", "mask_path": "assets/cat_06/v_--veKG73Di4_959_1753/0001s_00028_mask.png", "mask_status": "available" }, { "frame": "00060", "timestamp_sec": 34.0, "offset_sec": 2, "rgb_path": "assets/cat_06/v_--veKG73Di4_959_1753/0002s_00060_rgb.jpg", "mask_path": "assets/cat_06/v_--veKG73Di4_959_1753/0002s_00060_mask.png", "mask_status": "available" }, { "frame": "00088", "timestamp_sec": 35.0, "offset_sec": 3, "rgb_path": "assets/cat_06/v_--veKG73Di4_959_1753/0003s_00088_rgb.jpg", "mask_path": "assets/cat_06/v_--veKG73Di4_959_1753/0003s_00088_mask.png", "mask_status": "available" }, { "frame": "00120", "timestamp_sec": 36.0, "offset_sec": 4, "rgb_path": "assets/cat_06/v_--veKG73Di4_959_1753/0004s_00120_rgb.jpg", "mask_path": "assets/cat_06/v_--veKG73Di4_959_1753/0004s_00120_mask.png", "mask_status": "available" }, { "frame": "00148", "timestamp_sec": 37.0, "offset_sec": 5, "rgb_path": "assets/cat_06/v_--veKG73Di4_959_1753/0005s_00148_rgb.jpg", "mask_path": "assets/cat_06/v_--veKG73Di4_959_1753/0005s_00148_mask.png", "mask_status": "available" }, { "frame": "00180", "timestamp_sec": 38.0, "offset_sec": 6, "rgb_path": "assets/cat_06/v_--veKG73Di4_959_1753/0006s_00180_rgb.jpg", "mask_path": "assets/cat_06/v_--veKG73Di4_959_1753/0006s_00180_mask.png", "mask_status": "available" }, { "frame": "00208", "timestamp_sec": 39.0, "offset_sec": 7, "rgb_path": "assets/cat_06/v_--veKG73Di4_959_1753/0007s_00208_rgb.jpg", "mask_path": "assets/cat_06/v_--veKG73Di4_959_1753/0007s_00208_mask.png", "mask_status": "available" }, { "frame": "00240", "timestamp_sec": 40.0, "offset_sec": 8, "rgb_path": "assets/cat_06/v_--veKG73Di4_959_1753/0008s_00240_rgb.jpg", "mask_path": "assets/cat_06/v_--veKG73Di4_959_1753/0008s_00240_mask.png", "mask_status": "available" }, { "frame": "00268", "timestamp_sec": 41.0, "offset_sec": 9, "rgb_path": "assets/cat_06/v_--veKG73Di4_959_1753/0009s_00268_rgb.jpg", "mask_path": "assets/cat_06/v_--veKG73Di4_959_1753/0009s_00268_mask.png", "mask_status": "available" }, { "frame": "00300", "timestamp_sec": 42.0, "offset_sec": 10, "rgb_path": "assets/cat_06/v_--veKG73Di4_959_1753/0010s_00300_rgb.jpg", "mask_path": "assets/cat_06/v_--veKG73Di4_959_1753/0010s_00300_mask.png", "mask_status": "available" }, { "frame": "00328", "timestamp_sec": 43.0, "offset_sec": 11, "rgb_path": "assets/cat_06/v_--veKG73Di4_959_1753/0011s_00328_rgb.jpg", "mask_path": "assets/cat_06/v_--veKG73Di4_959_1753/0011s_00328_mask.png", "mask_status": "available" }, { "frame": "00360", "timestamp_sec": 44.0, "offset_sec": 12, "rgb_path": "assets/cat_06/v_--veKG73Di4_959_1753/0012s_00360_rgb.jpg", "mask_path": "assets/cat_06/v_--veKG73Di4_959_1753/0012s_00360_mask.png", "mask_status": "available" }, { "frame": "00388", "timestamp_sec": 45.0, "offset_sec": 13, "rgb_path": "assets/cat_06/v_--veKG73Di4_959_1753/0013s_00388_rgb.jpg", "mask_path": "assets/cat_06/v_--veKG73Di4_959_1753/0013s_00388_mask.png", "mask_status": "available" }, { "frame": "00420", "timestamp_sec": 46.0, "offset_sec": 14, "rgb_path": "assets/cat_06/v_--veKG73Di4_959_1753/0014s_00420_rgb.jpg", "mask_path": "assets/cat_06/v_--veKG73Di4_959_1753/0014s_00420_mask.png", "mask_status": "available" }, { "frame": "00448", "timestamp_sec": 47.0, "offset_sec": 15, "rgb_path": "assets/cat_06/v_--veKG73Di4_959_1753/0015s_00448_rgb.jpg", "mask_path": "assets/cat_06/v_--veKG73Di4_959_1753/0015s_00448_mask.png", "mask_status": "available" }, { "frame": "00480", "timestamp_sec": 48.0, "offset_sec": 16, "rgb_path": "assets/cat_06/v_--veKG73Di4_959_1753/0016s_00480_rgb.jpg", "mask_path": "assets/cat_06/v_--veKG73Di4_959_1753/0016s_00480_mask.png", "mask_status": "available" }, { "frame": "00508", "timestamp_sec": 49.0, "offset_sec": 17, "rgb_path": "assets/cat_06/v_--veKG73Di4_959_1753/0017s_00508_rgb.jpg", "mask_path": "assets/cat_06/v_--veKG73Di4_959_1753/0017s_00508_mask.png", "mask_status": "available" }, { "frame": "00540", "timestamp_sec": 50.0, "offset_sec": 18, "rgb_path": "assets/cat_06/v_--veKG73Di4_959_1753/0018s_00540_rgb.jpg", "mask_path": "assets/cat_06/v_--veKG73Di4_959_1753/0018s_00540_mask.png", "mask_status": "available" }, { "frame": "00568", "timestamp_sec": 51.0, "offset_sec": 19, "rgb_path": "assets/cat_06/v_--veKG73Di4_959_1753/0019s_00568_rgb.jpg", "mask_path": "assets/cat_06/v_--veKG73Di4_959_1753/0019s_00568_mask.png", "mask_status": "available" }, { "frame": "00600", "timestamp_sec": 52.0, "offset_sec": 20, "rgb_path": "assets/cat_06/v_--veKG73Di4_959_1753/0020s_00600_rgb.jpg", "mask_path": "assets/cat_06/v_--veKG73Di4_959_1753/0020s_00600_mask.png", "mask_status": "available" }, { "frame": "00628", "timestamp_sec": 53.0, "offset_sec": 21, "rgb_path": "assets/cat_06/v_--veKG73Di4_959_1753/0021s_00628_rgb.jpg", "mask_path": "assets/cat_06/v_--veKG73Di4_959_1753/0021s_00628_mask.png", "mask_status": "available" }, { "frame": "00660", "timestamp_sec": 54.0, "offset_sec": 22, "rgb_path": "assets/cat_06/v_--veKG73Di4_959_1753/0022s_00660_rgb.jpg", "mask_path": "assets/cat_06/v_--veKG73Di4_959_1753/0022s_00660_mask.png", "mask_status": "available" }, { "frame": "00688", "timestamp_sec": 55.0, "offset_sec": 23, "rgb_path": "assets/cat_06/v_--veKG73Di4_959_1753/0023s_00688_rgb.jpg", "mask_path": "assets/cat_06/v_--veKG73Di4_959_1753/0023s_00688_mask.png", "mask_status": "available" }, { "frame": "00720", "timestamp_sec": 56.0, "offset_sec": 24, "rgb_path": "assets/cat_06/v_--veKG73Di4_959_1753/0024s_00720_rgb.jpg", "mask_path": "assets/cat_06/v_--veKG73Di4_959_1753/0024s_00720_mask.png", "mask_status": "available" }, { "frame": "00748", "timestamp_sec": 57.0, "offset_sec": 25, "rgb_path": "assets/cat_06/v_--veKG73Di4_959_1753/0025s_00748_rgb.jpg", "mask_path": "assets/cat_06/v_--veKG73Di4_959_1753/0025s_00748_mask.png", "mask_status": "available" }, { "frame": "00780", "timestamp_sec": 58.0, "offset_sec": 26, "rgb_path": "assets/cat_06/v_--veKG73Di4_959_1753/0026s_00780_rgb.jpg", "mask_path": "assets/cat_06/v_--veKG73Di4_959_1753/0026s_00780_mask.png", "mask_status": "available" } ], "obj_id": [ 0 ], "anno_id": [ 5 ], "fps_inferred_from_clip_interval": 29.9623 }, "usage": { "input_tokens": 43066, "input_tokens_details": { "cache_write_tokens": 43063, "cached_tokens": 0 }, "output_tokens": 4469, "output_tokens_details": { "reasoning_tokens": 2537 }, "total_tokens": 47535 }, "model": "gpt-5.6-sol", "effort": "medium", "custom_id": "cat06_cap04_v_--veKG73Di4_959_1753", "annotation": { "caption": "She begins dancing energetically, moving her hips and arms while still holding the bottle and cap. Her movements include raising her arms, turning in a circle, and swaying from side to side. Her braids swing as she dances.", "entity_groups": [ { "id": "entity_1", "label": "woman", "spans": [ { "text": "She", "range": [ 0, 3 ] }, { "text": "her", "range": [ 41, 44 ] }, { "text": "Her", "range": [ 99, 102 ] }, { "text": "her", "range": [ 129, 132 ] }, { "text": "Her", "range": [ 191, 194 ] }, { "text": "she", "range": [ 211, 214 ] } ], "source_mask_link": "selected_person" }, { "id": "entity_2", "label": "hips", "spans": [ { "text": "her hips", "range": [ 41, 49 ] } ], "source_mask_link": null }, { "id": "entity_3", "label": "arms", "spans": [ { "text": "arms", "range": [ 54, 58 ] }, { "text": "her arms", "range": [ 129, 137 ] } ], "source_mask_link": null }, { "id": "entity_4", "label": "bottle", "spans": [ { "text": "the bottle", "range": [ 79, 89 ] } ], "source_mask_link": null }, { "id": "entity_5", "label": "cap", "spans": [ { "text": "cap", "range": [ 94, 97 ] } ], "source_mask_link": null }, { "id": "entity_6", "label": "braids", "spans": [ { "text": "Her braids", "range": [ 191, 201 ] } ], "source_mask_link": null } ], "actions": [ { "id": "action_1", "mentions": [ { "text": "dancing", "ranges": [ [ 11, 18 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } }, { "text": "dances", "ranges": [ [ 215, 221 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "type": "simple", "subject_refs": [ "entity_1" ], "object_refs": [], "confidence": "high", "temporal_status": "localized", "intervals": [ { "start_s": 0.0, "end_s": 26.5, "source_start_s": 32.0, "source_end_s": 58.5 } ], "evidence": "Across the sampled clean frames, the masked woman repeatedly changes her stance, weight distribution, hip position, and arm pose in a continuous dance through the final frame.", "mask_relation": "subject_matches_supplied_selected_person_mask", "source_subject_refs": "entity_1" }, { "id": "action_2", "mentions": [ { "text": "moving", "ranges": [ [ 34, 40 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "type": "simple", "subject_refs": [ "entity_1" ], "object_refs": [ "entity_2", "entity_3" ], "confidence": "high", "temporal_status": "localized", "intervals": [ { "start_s": 0.0, "end_s": 26.5, "source_start_s": 32.0, "source_end_s": 58.5 } ], "evidence": "The masked woman's hip alignment and arm positions vary continuously across the frame sequence.", "mask_relation": "subject_matches_supplied_selected_person_mask", "source_subject_refs": "entity_1" }, { "id": "action_3", "mentions": [ { "text": "holding", "ranges": [ [ 71, 78 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "type": "simple", "subject_refs": [ "entity_1" ], "object_refs": [ "entity_4", "entity_5" ], "confidence": "medium", "temporal_status": "localized", "intervals": [ { "start_s": 0.0, "end_s": 2.0, "source_start_s": 32.0, "source_end_s": 34.0 } ], "evidence": "At 0 and 1 seconds, the masked woman holds a small elongated item between her hands; the bottle and cap are not clearly distinguishable afterward.", "mask_relation": "subject_matches_supplied_selected_person_mask", "source_subject_refs": "entity_1" }, { "id": "action_4", "mentions": [ { "text": "raising", "ranges": [ [ 121, 128 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "type": "simple", "subject_refs": [ "entity_1" ], "object_refs": [ "entity_3" ], "confidence": "high", "temporal_status": "localized", "intervals": [ { "start_s": 1.0, "end_s": 5.0, "source_start_s": 33.0, "source_end_s": 37.0 }, { "start_s": 8.0, "end_s": 11.0, "source_start_s": 40.0, "source_end_s": 43.0 }, { "start_s": 17.0, "end_s": 22.0, "source_start_s": 49.0, "source_end_s": 54.0 }, { "start_s": 23.0, "end_s": 25.0, "source_start_s": 55.0, "source_end_s": 57.0 } ], "evidence": "The sampled sequence shows the masked woman repeatedly lifting one or both arms from lower positions to chest, shoulder, or head height.", "mask_relation": "subject_matches_supplied_selected_person_mask", "source_subject_refs": "entity_1" }, { "id": "action_5", "mentions": [ { "text": "turning", "ranges": [ [ 139, 146 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "type": "simple", "subject_refs": [ "entity_1" ], "object_refs": [], "confidence": "high", "temporal_status": "localized", "intervals": [ { "start_s": 9.0, "end_s": 18.0, "source_start_s": 41.0, "source_end_s": 50.0 } ], "evidence": "From 9 through 17 seconds, the masked woman changes from front-facing to side/back-facing orientations and then returns toward the front.", "mask_relation": "subject_matches_supplied_selected_person_mask", "source_subject_refs": "entity_1" }, { "id": "action_6", "mentions": [ { "text": "swaying", "ranges": [ [ 164, 171 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "type": "simple", "subject_refs": [ "entity_1" ], "object_refs": [], "confidence": "high", "temporal_status": "localized", "intervals": [ { "start_s": 5.0, "end_s": 26.5, "source_start_s": 37.0, "source_end_s": 58.5 } ], "evidence": "Successive clean frames show repeated lateral weight and hip shifts by the masked woman, continuing through the final sample.", "mask_relation": "subject_matches_supplied_selected_person_mask", "source_subject_refs": "entity_1" } ], "notes": [ "The supplied masklet consistently identifies the dancing woman; no separate masks are supplied for her hips, arms, bottle, cap, or braids.", "The action 'swing' at [202,207) is omitted because its grammatical subject is the separately annotated braids entity, which has no corresponding supplied mask.", "The aspectual predicate 'begins' and the predicate 'include' are not action spans under the RDCap V5 rules.", "Temporal intervals are sparse support intervals based on the sampled clean RGB frames, not exact continuous-video onset and offset estimates." ] }, "model_output_raw": { "caption": "She begins dancing energetically, moving her hips and arms while still holding the bottle and cap. Her movements include raising her arms, turning in a circle, and swaying from side to side. Her braids swing as she dances.", "entity_groups": [ { "id": "entity_1", "label": "woman", "spans": [ { "text": "She", "range": [ 0, 3 ] }, { "text": "her", "range": [ 41, 44 ] }, { "text": "Her", "range": [ 99, 102 ] }, { "text": "her", "range": [ 129, 132 ] }, { "text": "Her", "range": [ 191, 194 ] }, { "text": "she", "range": [ 211, 214 ] } ], "source_mask": "selected_person_masklet" }, { "id": "entity_2", "label": "hips", "spans": [ { "text": "her hips", "range": [ 41, 49 ] } ], "source_mask": null }, { "id": "entity_3", "label": "arms", "spans": [ { "text": "arms", "range": [ 54, 58 ] }, { "text": "her arms", "range": [ 129, 137 ] } ], "source_mask": null }, { "id": "entity_4", "label": "bottle", "spans": [ { "text": "the bottle", "range": [ 79, 89 ] } ], "source_mask": null }, { "id": "entity_5", "label": "cap", "spans": [ { "text": "cap", "range": [ 94, 97 ] } ], "source_mask": null }, { "id": "entity_6", "label": "braids", "spans": [ { "text": "Her braids", "range": [ 191, 201 ] } ], "source_mask": null } ], "actions": [ { "id": "action_1", "mentions": [ { "text": "dancing", "ranges": [ [ 11, 18 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } }, { "text": "dances", "ranges": [ [ 215, 221 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "type": "simple", "subject": "entity_1", "objects": [], "confidence": "high", "mask_relation": "subject_matches_supplied_selected_person_mask", "temporal_status": "localized", "intervals": [ { "start_s": 0.0, "end_s": 26.5 } ], "evidence": "Across the sampled clean frames, the masked woman repeatedly changes her stance, weight distribution, hip position, and arm pose in a continuous dance through the final frame." }, { "id": "action_2", "mentions": [ { "text": "moving", "ranges": [ [ 34, 40 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "type": "simple", "subject": "entity_1", "objects": [ "entity_2", "entity_3" ], "confidence": "high", "mask_relation": "subject_matches_supplied_selected_person_mask", "temporal_status": "localized", "intervals": [ { "start_s": 0.0, "end_s": 26.5 } ], "evidence": "The masked woman's hip alignment and arm positions vary continuously across the frame sequence." }, { "id": "action_3", "mentions": [ { "text": "holding", "ranges": [ [ 71, 78 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "type": "simple", "subject": "entity_1", "objects": [ "entity_4", "entity_5" ], "confidence": "medium", "mask_relation": "subject_matches_supplied_selected_person_mask", "temporal_status": "localized", "intervals": [ { "start_s": 0.0, "end_s": 2.0 } ], "evidence": "At 0 and 1 seconds, the masked woman holds a small elongated item between her hands; the bottle and cap are not clearly distinguishable afterward." }, { "id": "action_4", "mentions": [ { "text": "raising", "ranges": [ [ 121, 128 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "type": "simple", "subject": "entity_1", "objects": [ "entity_3" ], "confidence": "high", "mask_relation": "subject_matches_supplied_selected_person_mask", "temporal_status": "localized", "intervals": [ { "start_s": 1.0, "end_s": 5.0 }, { "start_s": 8.0, "end_s": 11.0 }, { "start_s": 17.0, "end_s": 22.0 }, { "start_s": 23.0, "end_s": 25.0 } ], "evidence": "The sampled sequence shows the masked woman repeatedly lifting one or both arms from lower positions to chest, shoulder, or head height." }, { "id": "action_5", "mentions": [ { "text": "turning", "ranges": [ [ 139, 146 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "type": "simple", "subject": "entity_1", "objects": [], "confidence": "high", "mask_relation": "subject_matches_supplied_selected_person_mask", "temporal_status": "localized", "intervals": [ { "start_s": 9.0, "end_s": 18.0 } ], "evidence": "From 9 through 17 seconds, the masked woman changes from front-facing to side/back-facing orientations and then returns toward the front." }, { "id": "action_6", "mentions": [ { "text": "swaying", "ranges": [ [ 164, 171 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "type": "simple", "subject": "entity_1", "objects": [], "confidence": "high", "mask_relation": "subject_matches_supplied_selected_person_mask", "temporal_status": "localized", "intervals": [ { "start_s": 5.0, "end_s": 26.5 } ], "evidence": "Successive clean frames show repeated lateral weight and hip shifts by the masked woman, continuing through the final sample." } ], "notes": [ "The supplied masklet consistently identifies the dancing woman; no separate masks are supplied for her hips, arms, bottle, cap, or braids.", "The action 'swing' at [202,207) is omitted because its grammatical subject is the separately annotated braids entity, which has no corresponding supplied mask.", "The aspectual predicate 'begins' and the predicate 'include' are not action spans under the RDCap V5 rules.", "Temporal intervals are sparse support intervals based on the sampled clean RGB frames, not exact continuous-video onset and offset estimates." ] }, "normalization_repairs": 0 }, { "source_video_id": "v_--veKG73Di4", "caption_count_category": 6, "clip": { "caption_index": 5, "caption": "Continuing to dance, she moves closer to the conga drums and mimes playing them with her hands. She then steps back, turns her body from side to side, and continues dancing with fluid arm movements.", "source_interval_sec": [ 58.5, 94.5 ], "loom_clip_id": "v_--veKG73Di4_1753_2832", "clip_frame_interval_4fps": [ 1753, 2832 ], "sample_count_1fps": 36, "sampled_frames": [ { "frame": "00000", "timestamp_sec": 58.5, "offset_sec": 0, "rgb_path": "assets/cat_06/v_--veKG73Di4_1753_2832/0000s_00000_rgb.jpg", "mask_path": "assets/cat_06/v_--veKG73Di4_1753_2832/0000s_00000_mask.png", "mask_status": "available" }, { "frame": "00028", "timestamp_sec": 59.5, "offset_sec": 1, "rgb_path": "assets/cat_06/v_--veKG73Di4_1753_2832/0001s_00028_rgb.jpg", "mask_path": "assets/cat_06/v_--veKG73Di4_1753_2832/0001s_00028_mask.png", "mask_status": "available" }, { "frame": "00060", "timestamp_sec": 60.5, "offset_sec": 2, "rgb_path": "assets/cat_06/v_--veKG73Di4_1753_2832/0002s_00060_rgb.jpg", "mask_path": "assets/cat_06/v_--veKG73Di4_1753_2832/0002s_00060_mask.png", "mask_status": "available" }, { "frame": "00088", "timestamp_sec": 61.5, "offset_sec": 3, "rgb_path": "assets/cat_06/v_--veKG73Di4_1753_2832/0003s_00088_rgb.jpg", "mask_path": "assets/cat_06/v_--veKG73Di4_1753_2832/0003s_00088_mask.png", "mask_status": "available" }, { "frame": "00120", "timestamp_sec": 62.5, "offset_sec": 4, "rgb_path": "assets/cat_06/v_--veKG73Di4_1753_2832/0004s_00120_rgb.jpg", "mask_path": "assets/cat_06/v_--veKG73Di4_1753_2832/0004s_00120_mask.png", "mask_status": "available" }, { "frame": "00148", "timestamp_sec": 63.5, "offset_sec": 5, "rgb_path": "assets/cat_06/v_--veKG73Di4_1753_2832/0005s_00148_rgb.jpg", "mask_path": "assets/cat_06/v_--veKG73Di4_1753_2832/0005s_00148_mask.png", "mask_status": "available" }, { "frame": "00180", "timestamp_sec": 64.5, "offset_sec": 6, "rgb_path": "assets/cat_06/v_--veKG73Di4_1753_2832/0006s_00180_rgb.jpg", "mask_path": "assets/cat_06/v_--veKG73Di4_1753_2832/0006s_00180_mask.png", "mask_status": "available" }, { "frame": "00208", "timestamp_sec": 65.5, "offset_sec": 7, "rgb_path": "assets/cat_06/v_--veKG73Di4_1753_2832/0007s_00208_rgb.jpg", "mask_path": "assets/cat_06/v_--veKG73Di4_1753_2832/0007s_00208_mask.png", "mask_status": "available" }, { "frame": "00240", "timestamp_sec": 66.5, "offset_sec": 8, "rgb_path": "assets/cat_06/v_--veKG73Di4_1753_2832/0008s_00240_rgb.jpg", "mask_path": "assets/cat_06/v_--veKG73Di4_1753_2832/0008s_00240_mask.png", "mask_status": "available" }, { "frame": "00268", "timestamp_sec": 67.5, "offset_sec": 9, "rgb_path": "assets/cat_06/v_--veKG73Di4_1753_2832/0009s_00268_rgb.jpg", "mask_path": "assets/cat_06/v_--veKG73Di4_1753_2832/0009s_00268_mask.png", "mask_status": "available" }, { "frame": "00300", "timestamp_sec": 68.5, "offset_sec": 10, "rgb_path": "assets/cat_06/v_--veKG73Di4_1753_2832/0010s_00300_rgb.jpg", "mask_path": "assets/cat_06/v_--veKG73Di4_1753_2832/0010s_00300_mask.png", "mask_status": "available" }, { "frame": "00328", "timestamp_sec": 69.5, "offset_sec": 11, "rgb_path": "assets/cat_06/v_--veKG73Di4_1753_2832/0011s_00328_rgb.jpg", "mask_path": "assets/cat_06/v_--veKG73Di4_1753_2832/0011s_00328_mask.png", "mask_status": "available" }, { "frame": "00360", "timestamp_sec": 70.5, "offset_sec": 12, "rgb_path": "assets/cat_06/v_--veKG73Di4_1753_2832/0012s_00360_rgb.jpg", "mask_path": "assets/cat_06/v_--veKG73Di4_1753_2832/0012s_00360_mask.png", "mask_status": "available" }, { "frame": "00388", "timestamp_sec": 71.5, "offset_sec": 13, "rgb_path": "assets/cat_06/v_--veKG73Di4_1753_2832/0013s_00388_rgb.jpg", "mask_path": "assets/cat_06/v_--veKG73Di4_1753_2832/0013s_00388_mask.png", "mask_status": "available" }, { "frame": "00420", "timestamp_sec": 72.5, "offset_sec": 14, "rgb_path": "assets/cat_06/v_--veKG73Di4_1753_2832/0014s_00420_rgb.jpg", "mask_path": "assets/cat_06/v_--veKG73Di4_1753_2832/0014s_00420_mask.png", "mask_status": "available" }, { "frame": "00448", "timestamp_sec": 73.5, "offset_sec": 15, "rgb_path": "assets/cat_06/v_--veKG73Di4_1753_2832/0015s_00448_rgb.jpg", "mask_path": "assets/cat_06/v_--veKG73Di4_1753_2832/0015s_00448_mask.png", "mask_status": "available" }, { "frame": "00480", "timestamp_sec": 74.5, "offset_sec": 16, "rgb_path": "assets/cat_06/v_--veKG73Di4_1753_2832/0016s_00480_rgb.jpg", "mask_path": "assets/cat_06/v_--veKG73Di4_1753_2832/0016s_00480_mask.png", "mask_status": "available" }, { "frame": "00508", "timestamp_sec": 75.5, "offset_sec": 17, "rgb_path": "assets/cat_06/v_--veKG73Di4_1753_2832/0017s_00508_rgb.jpg", "mask_path": "assets/cat_06/v_--veKG73Di4_1753_2832/0017s_00508_mask.png", "mask_status": "available" }, { "frame": "00540", "timestamp_sec": 76.5, "offset_sec": 18, "rgb_path": "assets/cat_06/v_--veKG73Di4_1753_2832/0018s_00540_rgb.jpg", "mask_path": "assets/cat_06/v_--veKG73Di4_1753_2832/0018s_00540_mask.png", "mask_status": "available" }, { "frame": "00568", "timestamp_sec": 77.5, "offset_sec": 19, "rgb_path": "assets/cat_06/v_--veKG73Di4_1753_2832/0019s_00568_rgb.jpg", "mask_path": "assets/cat_06/v_--veKG73Di4_1753_2832/0019s_00568_mask.png", "mask_status": "available" }, { "frame": "00600", "timestamp_sec": 78.5, "offset_sec": 20, "rgb_path": "assets/cat_06/v_--veKG73Di4_1753_2832/0020s_00600_rgb.jpg", "mask_path": "assets/cat_06/v_--veKG73Di4_1753_2832/0020s_00600_mask.png", "mask_status": "available" }, { "frame": "00628", "timestamp_sec": 79.5, "offset_sec": 21, "rgb_path": "assets/cat_06/v_--veKG73Di4_1753_2832/0021s_00628_rgb.jpg", "mask_path": "assets/cat_06/v_--veKG73Di4_1753_2832/0021s_00628_mask.png", "mask_status": "available" }, { "frame": "00660", "timestamp_sec": 80.5, "offset_sec": 22, "rgb_path": "assets/cat_06/v_--veKG73Di4_1753_2832/0022s_00660_rgb.jpg", "mask_path": "assets/cat_06/v_--veKG73Di4_1753_2832/0022s_00660_mask.png", "mask_status": "available" }, { "frame": "00688", "timestamp_sec": 81.5, "offset_sec": 23, "rgb_path": "assets/cat_06/v_--veKG73Di4_1753_2832/0023s_00688_rgb.jpg", "mask_path": "assets/cat_06/v_--veKG73Di4_1753_2832/0023s_00688_mask.png", "mask_status": "available" }, { "frame": "00720", "timestamp_sec": 82.5, "offset_sec": 24, "rgb_path": "assets/cat_06/v_--veKG73Di4_1753_2832/0024s_00720_rgb.jpg", "mask_path": "assets/cat_06/v_--veKG73Di4_1753_2832/0024s_00720_mask.png", "mask_status": "available" }, { "frame": "00748", "timestamp_sec": 83.5, "offset_sec": 25, "rgb_path": "assets/cat_06/v_--veKG73Di4_1753_2832/0025s_00748_rgb.jpg", "mask_path": "assets/cat_06/v_--veKG73Di4_1753_2832/0025s_00748_mask.png", "mask_status": "available" }, { "frame": "00780", "timestamp_sec": 84.5, "offset_sec": 26, "rgb_path": "assets/cat_06/v_--veKG73Di4_1753_2832/0026s_00780_rgb.jpg", "mask_path": "assets/cat_06/v_--veKG73Di4_1753_2832/0026s_00780_mask.png", "mask_status": "available" }, { "frame": "00808", "timestamp_sec": 85.5, "offset_sec": 27, "rgb_path": "assets/cat_06/v_--veKG73Di4_1753_2832/0027s_00808_rgb.jpg", "mask_path": "assets/cat_06/v_--veKG73Di4_1753_2832/0027s_00808_mask.png", "mask_status": "available" }, { "frame": "00840", "timestamp_sec": 86.5, "offset_sec": 28, "rgb_path": "assets/cat_06/v_--veKG73Di4_1753_2832/0028s_00840_rgb.jpg", "mask_path": "assets/cat_06/v_--veKG73Di4_1753_2832/0028s_00840_mask.png", "mask_status": "available" }, { "frame": "00868", "timestamp_sec": 87.5, "offset_sec": 29, "rgb_path": "assets/cat_06/v_--veKG73Di4_1753_2832/0029s_00868_rgb.jpg", "mask_path": "assets/cat_06/v_--veKG73Di4_1753_2832/0029s_00868_mask.png", "mask_status": "available" }, { "frame": "00900", "timestamp_sec": 88.5, "offset_sec": 30, "rgb_path": "assets/cat_06/v_--veKG73Di4_1753_2832/0030s_00900_rgb.jpg", "mask_path": "assets/cat_06/v_--veKG73Di4_1753_2832/0030s_00900_mask.png", "mask_status": "available" }, { "frame": "00928", "timestamp_sec": 89.5, "offset_sec": 31, "rgb_path": "assets/cat_06/v_--veKG73Di4_1753_2832/0031s_00928_rgb.jpg", "mask_path": "assets/cat_06/v_--veKG73Di4_1753_2832/0031s_00928_mask.png", "mask_status": "available" }, { "frame": "00960", "timestamp_sec": 90.5, "offset_sec": 32, "rgb_path": "assets/cat_06/v_--veKG73Di4_1753_2832/0032s_00960_rgb.jpg", "mask_path": "assets/cat_06/v_--veKG73Di4_1753_2832/0032s_00960_mask.png", "mask_status": "available" }, { "frame": "00988", "timestamp_sec": 91.5, "offset_sec": 33, "rgb_path": "assets/cat_06/v_--veKG73Di4_1753_2832/0033s_00988_rgb.jpg", "mask_path": "assets/cat_06/v_--veKG73Di4_1753_2832/0033s_00988_mask.png", "mask_status": "available" }, { "frame": "01020", "timestamp_sec": 92.5, "offset_sec": 34, "rgb_path": "assets/cat_06/v_--veKG73Di4_1753_2832/0034s_01020_rgb.jpg", "mask_path": "assets/cat_06/v_--veKG73Di4_1753_2832/0034s_01020_mask.png", "mask_status": "available" }, { "frame": "01048", "timestamp_sec": 93.5, "offset_sec": 35, "rgb_path": "assets/cat_06/v_--veKG73Di4_1753_2832/0035s_01048_rgb.jpg", "mask_path": "assets/cat_06/v_--veKG73Di4_1753_2832/0035s_01048_mask.png", "mask_status": "available" } ], "obj_id": [ 0 ], "anno_id": [ 0 ], "fps_inferred_from_clip_interval": 29.9722 }, "usage": { "input_tokens": 52305, "input_tokens_details": { "cache_write_tokens": 52302, "cached_tokens": 0 }, "output_tokens": 3889, "output_tokens_details": { "reasoning_tokens": 2070 }, "total_tokens": 56194 }, "model": "gpt-5.6-sol", "effort": "medium", "custom_id": "cat06_cap05_v_--veKG73Di4_1753_2832", "annotation": { "caption": "Continuing to dance, she moves closer to the conga drums and mimes playing them with her hands. She then steps back, turns her body from side to side, and continues dancing with fluid arm movements.", "entity_groups": [ { "id": "person_1", "label": "woman", "spans": [ { "text": "she", "range": [ 21, 24 ] }, { "text": "her", "range": [ 85, 88 ] }, { "text": "She", "range": [ 96, 99 ] }, { "text": "her", "range": [ 123, 126 ] } ], "source_mask_link": "selected_person" }, { "id": "conga_drums_1", "label": "conga drums", "spans": [ { "text": "the conga drums", "range": [ 41, 56 ] }, { "text": "them", "range": [ 75, 79 ] } ], "source_mask_link": null }, { "id": "hands_1", "label": "hands", "spans": [ { "text": "her hands", "range": [ 85, 94 ] } ], "source_mask_link": null }, { "id": "body_1", "label": "body", "spans": [ { "text": "her body", "range": [ 123, 131 ] } ], "source_mask_link": null }, { "id": "arms_1", "label": "arms", "spans": [ { "text": "arm", "range": [ 184, 187 ] } ], "source_mask_link": null } ], "actions": [ { "id": "action_1", "mentions": [ { "text": "dance", "ranges": [ [ 14, 19 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } }, { "text": "dancing", "ranges": [ [ 165, 172 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "type": "simple", "subject_refs": [ "person_1" ], "object_refs": [], "confidence": "high", "temporal_status": "localized", "intervals": [ { "start_s": 0.0, "end_s": 36.0, "source_start_s": 58.5, "source_end_s": 94.5 } ], "evidence": "Across the sampled clean frames, the masked woman repeatedly shifts her weight, bends her knees, changes pose, and moves her arms rhythmically; this continues through the final sample at 35.0 seconds.", "mask_relation": "The action subject matches the supplied selected-person mask.", "source_subject_refs": [ "person_1" ] }, { "id": "action_2", "mentions": [ { "text": "moves", "ranges": [ [ 25, 30 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "type": "simple", "subject_refs": [ "person_1" ], "object_refs": [], "confidence": "high", "temporal_status": "localized", "intervals": [ { "start_s": 0.0, "end_s": 2.0, "source_start_s": 58.5, "source_end_s": 60.5 } ], "evidence": "At 0.0–2.0 seconds, the masked woman is immediately beside the conga drums and shifts toward their playing surfaces.", "mask_relation": "The action subject matches the supplied selected-person mask.", "source_subject_refs": [ "person_1" ] }, { "id": "action_3", "mentions": [ { "text": "mimes", "ranges": [ [ 61, 66 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "type": "simple", "subject_refs": [ "person_1" ], "object_refs": [ "conga_drums_1" ], "confidence": "high", "temporal_status": "localized", "intervals": [ { "start_s": 0.0, "end_s": 3.0, "source_start_s": 58.5, "source_end_s": 61.5 } ], "evidence": "In the first three sampled frames, the masked woman holds and moves her hands over the conga playing area in imitation of the adjacent drummer.", "mask_relation": "The action subject matches the supplied selected-person mask.", "source_subject_refs": [ "person_1" ] }, { "id": "action_4", "mentions": [ { "text": "playing", "ranges": [ [ 67, 74 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "type": "simple", "subject_refs": [ "person_1" ], "object_refs": [ "conga_drums_1" ], "confidence": "medium", "temporal_status": "localized", "intervals": [ { "start_s": 0.0, "end_s": 3.0, "source_start_s": 58.5, "source_end_s": 61.5 } ], "evidence": "At 0.0–2.0 seconds, her hand positions and gestures visually represent playing the congas, although the caption characterizes the action as mimed rather than actual performance.", "mask_relation": "The action subject matches the supplied selected-person mask.", "source_subject_refs": [ "person_1" ] }, { "id": "action_5", "mentions": [ { "text": "steps", "ranges": [ [ 105, 110 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "type": "simple", "subject_refs": [ "person_1" ], "object_refs": [], "confidence": "high", "temporal_status": "localized", "intervals": [ { "start_s": 2.0, "end_s": 5.0, "source_start_s": 60.5, "source_end_s": 63.5 } ], "evidence": "Between the 2.0- and 4.0-second samples, the masked woman withdraws from the drum fronts and shifts back toward the right.", "mask_relation": "The action subject matches the supplied selected-person mask.", "source_subject_refs": [ "person_1" ] }, { "id": "action_6", "mentions": [ { "text": "turns", "ranges": [ [ 117, 122 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "type": "simple", "subject_refs": [ "person_1" ], "object_refs": [ "body_1" ], "confidence": "high", "temporal_status": "localized", "intervals": [ { "start_s": 4.0, "end_s": 36.0, "source_start_s": 62.5, "source_end_s": 94.5 } ], "evidence": "From about 4.0 seconds onward, successive clean frames show the masked woman's torso repeatedly changing orientation while she dances, including side-facing and back-facing poses.", "mask_relation": "The action subject matches the supplied selected-person mask.", "source_subject_refs": [ "person_1" ] } ], "notes": [ "The supplied mask consistently identifies the dancing woman in green; it does not identify the male conga player.", "The conga drums and the woman's hands, body, and arms are text-groundable entities but have no separate supplied instance masks.", "The source-video interval is [58.5, 94.5) seconds, corresponding to the clip-relative interval [0.0, 36.0) seconds.", "Temporal intervals are sparse support intervals inferred from the sampled clean RGB frames, not exact continuous-video boundaries." ] }, "model_output_raw": { "caption": "Continuing to dance, she moves closer to the conga drums and mimes playing them with her hands. She then steps back, turns her body from side to side, and continues dancing with fluid arm movements.", "entity_groups": [ { "id": "person_1", "entity": "woman", "spans": [ { "range": [ 21, 24 ], "text": "she" }, { "range": [ 85, 88 ], "text": "her" }, { "range": [ 96, 99 ], "text": "She" }, { "range": [ 123, 126 ], "text": "her" } ], "source_mask_link": "selected_person" }, { "id": "conga_drums_1", "entity": "conga drums", "spans": [ { "range": [ 41, 56 ], "text": "the conga drums" }, { "range": [ 75, 79 ], "text": "them" } ], "source_mask_link": null }, { "id": "hands_1", "entity": "hands", "spans": [ { "range": [ 85, 94 ], "text": "her hands" } ], "source_mask_link": null }, { "id": "body_1", "entity": "body", "spans": [ { "range": [ 123, 131 ], "text": "her body" } ], "source_mask_link": null }, { "id": "arms_1", "entity": "arms", "spans": [ { "range": [ 184, 187 ], "text": "arm" } ], "source_mask_link": null } ], "actions": [ { "id": "action_1", "mentions": [ { "text": "dance", "ranges": [ [ 14, 19 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } }, { "text": "dancing", "ranges": [ [ 165, 172 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "type": "simple", "subject_refs": [ "person_1" ], "object_refs": [], "confidence": "high", "temporal_status": "localized", "intervals": [ { "start_s": 0.0, "end_s": 36.0 } ], "evidence": "Across the sampled clean frames, the masked woman repeatedly shifts her weight, bends her knees, changes pose, and moves her arms rhythmically; this continues through the final sample at 35.0 seconds.", "mask_relation": "The action subject matches the supplied selected-person mask." }, { "id": "action_2", "mentions": [ { "text": "moves", "ranges": [ [ 25, 30 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "type": "simple", "subject_refs": [ "person_1" ], "object_refs": [], "confidence": "high", "temporal_status": "localized", "intervals": [ { "start_s": 0.0, "end_s": 2.0 } ], "evidence": "At 0.0–2.0 seconds, the masked woman is immediately beside the conga drums and shifts toward their playing surfaces.", "mask_relation": "The action subject matches the supplied selected-person mask." }, { "id": "action_3", "mentions": [ { "text": "mimes", "ranges": [ [ 61, 66 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "type": "simple", "subject_refs": [ "person_1" ], "object_refs": [ "conga_drums_1" ], "confidence": "high", "temporal_status": "localized", "intervals": [ { "start_s": 0.0, "end_s": 3.0 } ], "evidence": "In the first three sampled frames, the masked woman holds and moves her hands over the conga playing area in imitation of the adjacent drummer.", "mask_relation": "The action subject matches the supplied selected-person mask." }, { "id": "action_4", "mentions": [ { "text": "playing", "ranges": [ [ 67, 74 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "type": "simple", "subject_refs": [ "person_1" ], "object_refs": [ "conga_drums_1" ], "confidence": "medium", "temporal_status": "localized", "intervals": [ { "start_s": 0.0, "end_s": 3.0 } ], "evidence": "At 0.0–2.0 seconds, her hand positions and gestures visually represent playing the congas, although the caption characterizes the action as mimed rather than actual performance.", "mask_relation": "The action subject matches the supplied selected-person mask." }, { "id": "action_5", "mentions": [ { "text": "steps", "ranges": [ [ 105, 110 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "type": "simple", "subject_refs": [ "person_1" ], "object_refs": [], "confidence": "high", "temporal_status": "localized", "intervals": [ { "start_s": 2.0, "end_s": 5.0 } ], "evidence": "Between the 2.0- and 4.0-second samples, the masked woman withdraws from the drum fronts and shifts back toward the right.", "mask_relation": "The action subject matches the supplied selected-person mask." }, { "id": "action_6", "mentions": [ { "text": "turns", "ranges": [ [ 117, 122 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "type": "simple", "subject_refs": [ "person_1" ], "object_refs": [ "body_1" ], "confidence": "high", "temporal_status": "localized", "intervals": [ { "start_s": 4.0, "end_s": 36.0 } ], "evidence": "From about 4.0 seconds onward, successive clean frames show the masked woman's torso repeatedly changing orientation while she dances, including side-facing and back-facing poses.", "mask_relation": "The action subject matches the supplied selected-person mask." } ], "notes": [ "The supplied mask consistently identifies the dancing woman in green; it does not identify the male conga player.", "The conga drums and the woman's hands, body, and arms are text-groundable entities but have no separate supplied instance masks.", "The source-video interval is [58.5, 94.5) seconds, corresponding to the clip-relative interval [0.0, 36.0) seconds.", "Temporal intervals are sparse support intervals inferred from the sampled clean RGB frames, not exact continuous-video boundaries." ] }, "normalization_repairs": 0 }, { "source_video_id": "v_--veKG73Di4", "caption_count_category": 6, "clip": { "caption_index": 6, "caption": "She briefly stops her dance to point her right index finger at the drummer. Afterward, she resumes dancing more gently before turning to her right and walking out of the frame.", "source_interval_sec": [ 94.5, 109.81 ], "loom_clip_id": "v_--veKG73Di4_2832_3291", "clip_frame_interval_4fps": [ 2832, 3291 ], "sample_count_1fps": 16, "sampled_frames": [ { "frame": "00000", "timestamp_sec": 94.5, "offset_sec": 0, "rgb_path": "assets/cat_06/v_--veKG73Di4_2832_3291/0000s_00000_rgb.jpg", "mask_path": "assets/cat_06/v_--veKG73Di4_2832_3291/0000s_00000_mask.png", "mask_status": "available" }, { "frame": "00028", "timestamp_sec": 95.5, "offset_sec": 1, "rgb_path": "assets/cat_06/v_--veKG73Di4_2832_3291/0001s_00028_rgb.jpg", "mask_path": "assets/cat_06/v_--veKG73Di4_2832_3291/0001s_00028_mask.png", "mask_status": "available" }, { "frame": "00060", "timestamp_sec": 96.5, "offset_sec": 2, "rgb_path": "assets/cat_06/v_--veKG73Di4_2832_3291/0002s_00060_rgb.jpg", "mask_path": "assets/cat_06/v_--veKG73Di4_2832_3291/0002s_00060_mask.png", "mask_status": "available" }, { "frame": "00088", "timestamp_sec": 97.5, "offset_sec": 3, "rgb_path": "assets/cat_06/v_--veKG73Di4_2832_3291/0003s_00088_rgb.jpg", "mask_path": "assets/cat_06/v_--veKG73Di4_2832_3291/0003s_00088_mask.png", "mask_status": "available" }, { "frame": "00120", "timestamp_sec": 98.5, "offset_sec": 4, "rgb_path": "assets/cat_06/v_--veKG73Di4_2832_3291/0004s_00120_rgb.jpg", "mask_path": "assets/cat_06/v_--veKG73Di4_2832_3291/0004s_00120_mask.png", "mask_status": "available" }, { "frame": "00148", "timestamp_sec": 99.5, "offset_sec": 5, "rgb_path": "assets/cat_06/v_--veKG73Di4_2832_3291/0005s_00148_rgb.jpg", "mask_path": "assets/cat_06/v_--veKG73Di4_2832_3291/0005s_00148_mask.png", "mask_status": "available" }, { "frame": "00180", "timestamp_sec": 100.5, "offset_sec": 6, "rgb_path": "assets/cat_06/v_--veKG73Di4_2832_3291/0006s_00180_rgb.jpg", "mask_path": "assets/cat_06/v_--veKG73Di4_2832_3291/0006s_00180_mask.png", "mask_status": "available" }, { "frame": "00208", "timestamp_sec": 101.5, "offset_sec": 7, "rgb_path": "assets/cat_06/v_--veKG73Di4_2832_3291/0007s_00208_rgb.jpg", "mask_path": "assets/cat_06/v_--veKG73Di4_2832_3291/0007s_00208_mask.png", "mask_status": "available" }, { "frame": "00240", "timestamp_sec": 102.5, "offset_sec": 8, "rgb_path": "assets/cat_06/v_--veKG73Di4_2832_3291/0008s_00240_rgb.jpg", "mask_path": "assets/cat_06/v_--veKG73Di4_2832_3291/0008s_00240_mask.png", "mask_status": "available" }, { "frame": "00268", "timestamp_sec": 103.5, "offset_sec": 9, "rgb_path": "assets/cat_06/v_--veKG73Di4_2832_3291/0009s_00268_rgb.jpg", "mask_path": "assets/cat_06/v_--veKG73Di4_2832_3291/0009s_00268_mask.png", "mask_status": "available" }, { "frame": "00300", "timestamp_sec": 104.5, "offset_sec": 10, "rgb_path": "assets/cat_06/v_--veKG73Di4_2832_3291/0010s_00300_rgb.jpg", "mask_path": "assets/cat_06/v_--veKG73Di4_2832_3291/0010s_00300_mask.png", "mask_status": "available" }, { "frame": "00328", "timestamp_sec": 105.5, "offset_sec": 11, "rgb_path": "assets/cat_06/v_--veKG73Di4_2832_3291/0011s_00328_rgb.jpg", "mask_path": "assets/cat_06/v_--veKG73Di4_2832_3291/0011s_00328_mask.png", "mask_status": "available" }, { "frame": "00360", "timestamp_sec": 106.5, "offset_sec": 12, "rgb_path": "assets/cat_06/v_--veKG73Di4_2832_3291/0012s_00360_rgb.jpg", "mask_path": "assets/cat_06/v_--veKG73Di4_2832_3291/0012s_00360_mask.png", "mask_status": "available" }, { "frame": "00388", "timestamp_sec": 107.5, "offset_sec": 13, "rgb_path": "assets/cat_06/v_--veKG73Di4_2832_3291/0013s_00388_rgb.jpg", "mask_path": "assets/cat_06/v_--veKG73Di4_2832_3291/0013s_00388_mask.png", "mask_status": "available" }, { "frame": "00420", "timestamp_sec": 108.5, "offset_sec": 14, "rgb_path": "assets/cat_06/v_--veKG73Di4_2832_3291/0014s_00420_rgb.jpg", "mask_path": "assets/cat_06/v_--veKG73Di4_2832_3291/0014s_00420_mask.png", "mask_status": "available" }, { "frame": "00448", "timestamp_sec": 109.5, "offset_sec": 15, "rgb_path": "assets/cat_06/v_--veKG73Di4_2832_3291/0015s_00448_rgb.jpg", "mask_path": "assets/cat_06/v_--veKG73Di4_2832_3291/0015s_00448_mask.png", "mask_status": "available" } ], "obj_id": [ 0 ], "anno_id": [ 2 ], "fps_inferred_from_clip_interval": 29.9804 }, "usage": { "input_tokens": 31757, "input_tokens_details": { "cache_write_tokens": 31754, "cached_tokens": 0 }, "output_tokens": 4692, "output_tokens_details": { "reasoning_tokens": 3106 }, "total_tokens": 36449 }, "model": "gpt-5.6-sol", "effort": "medium", "custom_id": "cat06_cap06_v_--veKG73Di4_2832_3291", "annotation": { "caption": "She briefly stops her dance to point her right index finger at the drummer. Afterward, she resumes dancing more gently before turning to her right and walking out of the frame.", "entity_groups": [ { "id": "entity_1", "label": "woman", "spans": [ { "text": "She", "range": [ 0, 3 ] }, { "text": "her", "range": [ 18, 21 ] }, { "text": "her", "range": [ 37, 40 ] }, { "text": "she", "range": [ 87, 90 ] }, { "text": "her", "range": [ 137, 140 ] } ], "source_mask_link": "selected_person" }, { "id": "entity_2", "label": "right index finger", "spans": [ { "text": "her right index finger", "range": [ 37, 59 ] } ], "source_mask_link": null }, { "id": "entity_3", "label": "drummer", "spans": [ { "text": "the drummer", "range": [ 63, 74 ] } ], "source_mask_link": null } ], "actions": [ { "id": "action_1", "mentions": [ { "text": "stops", "ranges": [ [ 12, 17 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "type": "simple", "subject_refs": [ "entity_1" ], "object_refs": [], "confidence": "medium", "temporal_status": "supported", "intervals": [ { "start_s": 2.0, "end_s": 3.0, "source_start_s": 96.5, "source_end_s": 97.5 } ], "evidence": "At 2.0 seconds the masked woman pauses her broader dance movement while holding a pointing gesture.", "mask_relation": "The action subject matches the supplied selected-person mask.", "source_subject_refs": "entity_1" }, { "id": "action_2", "mentions": [ { "text": "point", "ranges": [ [ 31, 36 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "type": "prepositional", "subject_refs": [ "entity_1" ], "object_refs": [ "entity_3" ], "confidence": "high", "temporal_status": "supported", "intervals": [ { "start_s": 2.0, "end_s": 3.0, "source_start_s": 96.5, "source_end_s": 97.5 } ], "evidence": "The 2.0-second frame clearly shows the masked woman extending her index finger toward the drummer.", "mask_relation": "The action subject matches the supplied selected-person mask.", "source_subject_refs": "entity_1" }, { "id": "action_3", "mentions": [ { "text": "dancing", "ranges": [ [ 99, 106 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "type": "simple", "subject_refs": [ "entity_1" ], "object_refs": [], "confidence": "high", "temporal_status": "supported", "intervals": [ { "start_s": 4.0, "end_s": 14.0, "source_start_s": 98.5, "source_end_s": 108.5 } ], "evidence": "Across the 4.0–13.0-second samples, the masked woman repeatedly changes her arm, torso, and leg poses in gentle rhythmic movements.", "mask_relation": "The action subject matches the supplied selected-person mask.", "source_subject_refs": "entity_1" }, { "id": "action_4", "mentions": [ { "text": "turning", "ranges": [ [ 126, 133 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "type": "simple", "subject_refs": [ "entity_1" ], "object_refs": [], "confidence": "medium", "temporal_status": "supported", "intervals": [ { "start_s": 13.0, "end_s": 14.0, "source_start_s": 107.5, "source_end_s": 108.5 } ], "evidence": "Between the 13.0- and 14.0-second samples, the masked woman reorients toward the right side of the frame.", "mask_relation": "The action subject matches the supplied selected-person mask.", "source_subject_refs": "entity_1" }, { "id": "action_5", "mentions": [ { "text": "walking", "ranges": [ [ 151, 158 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "type": "simple", "subject_refs": [ "entity_1" ], "object_refs": [], "confidence": "high", "temporal_status": "supported", "intervals": [ { "start_s": 14.0, "end_s": 15.310000000000002, "source_start_s": 108.5, "source_end_s": 109.81 } ], "evidence": "At 14.0 seconds the masked woman is partly beyond the right edge, and by 15.0 seconds only a small masked portion remains; the exit continues to the clip boundary.", "mask_relation": "The action subject matches the supplied selected-person mask.", "source_subject_refs": "entity_1" } ], "notes": [ "The supplied masklet tracks only the woman in green; the drummer and the woman's index finger have no separate source masks.", "The possessive span \"her\" overlaps with \"her right index finger\" across the woman and body-part groups, as permitted by the entity rules.", "\"resumes\" is excluded as an aspectual verb; \"more gently\" is excluded as a modifier.", "\"out of the frame\" is excluded from the walking span as directional/prepositional material, and \"the frame\" is not annotated as an entity because it is a meta/spatial image reference." ] }, "model_output_raw": { "caption": "She briefly stops her dance to point her right index finger at the drummer. Afterward, she resumes dancing more gently before turning to her right and walking out of the frame.", "entity_groups": [ { "id": "entity_1", "label": "woman", "spans": [ { "text": "She", "range": [ 0, 3 ] }, { "text": "her", "range": [ 18, 21 ] }, { "text": "her", "range": [ 37, 40 ] }, { "text": "she", "range": [ 87, 90 ] }, { "text": "her", "range": [ 137, 140 ] } ], "source_mask": { "track": "selected_person", "description": "The supplied masks identify the woman in green." } }, { "id": "entity_2", "label": "right index finger", "spans": [ { "text": "her right index finger", "range": [ 37, 59 ] } ], "source_mask": null }, { "id": "entity_3", "label": "drummer", "spans": [ { "text": "the drummer", "range": [ 63, 74 ] } ], "source_mask": null } ], "actions": [ { "id": "action_1", "mentions": [ { "text": "stops", "ranges": [ [ 12, 17 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "type": "simple", "subject_entity_group": "entity_1", "object_entity_groups": [], "confidence": "medium", "mask_relation": "The action subject matches the supplied selected-person mask.", "temporal_status": "supported", "intervals": [ { "start_s": 2.0, "end_s": 3.0, "source_start_s": 96.5, "source_end_s": 97.5 } ], "evidence": "At 2.0 seconds the masked woman pauses her broader dance movement while holding a pointing gesture." }, { "id": "action_2", "mentions": [ { "text": "point", "ranges": [ [ 31, 36 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "type": "prepositional", "subject_entity_group": "entity_1", "object_entity_groups": [ "entity_3" ], "related_entity_groups": { "instrument": [ "entity_2" ], "target": [ "entity_3" ] }, "confidence": "high", "mask_relation": "The action subject matches the supplied selected-person mask.", "temporal_status": "supported", "intervals": [ { "start_s": 2.0, "end_s": 3.0, "source_start_s": 96.5, "source_end_s": 97.5 } ], "evidence": "The 2.0-second frame clearly shows the masked woman extending her index finger toward the drummer." }, { "id": "action_3", "mentions": [ { "text": "dancing", "ranges": [ [ 99, 106 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "type": "simple", "subject_entity_group": "entity_1", "object_entity_groups": [], "confidence": "high", "mask_relation": "The action subject matches the supplied selected-person mask.", "temporal_status": "supported", "intervals": [ { "start_s": 4.0, "end_s": 14.0, "source_start_s": 98.5, "source_end_s": 108.5 } ], "evidence": "Across the 4.0–13.0-second samples, the masked woman repeatedly changes her arm, torso, and leg poses in gentle rhythmic movements." }, { "id": "action_4", "mentions": [ { "text": "turning", "ranges": [ [ 126, 133 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "type": "simple", "subject_entity_group": "entity_1", "object_entity_groups": [], "confidence": "medium", "mask_relation": "The action subject matches the supplied selected-person mask.", "temporal_status": "supported", "intervals": [ { "start_s": 13.0, "end_s": 14.0, "source_start_s": 107.5, "source_end_s": 108.5 } ], "evidence": "Between the 13.0- and 14.0-second samples, the masked woman reorients toward the right side of the frame." }, { "id": "action_5", "mentions": [ { "text": "walking", "ranges": [ [ 151, 158 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "type": "simple", "subject_entity_group": "entity_1", "object_entity_groups": [], "confidence": "high", "mask_relation": "The action subject matches the supplied selected-person mask.", "temporal_status": "supported", "intervals": [ { "start_s": 14.0, "end_s": 15.31, "source_start_s": 108.5, "source_end_s": 109.81 } ], "evidence": "At 14.0 seconds the masked woman is partly beyond the right edge, and by 15.0 seconds only a small masked portion remains; the exit continues to the clip boundary." } ], "notes": [ "The supplied masklet tracks only the woman in green; the drummer and the woman's index finger have no separate source masks.", "The possessive span \"her\" overlaps with \"her right index finger\" across the woman and body-part groups, as permitted by the entity rules.", "\"resumes\" is excluded as an aspectual verb; \"more gently\" is excluded as a modifier.", "\"out of the frame\" is excluded from the walking span as directional/prepositional material, and \"the frame\" is not annotated as an entity because it is a meta/spatial image reference." ] }, "normalization_repairs": 0 }, { "source_video_id": "v_00Dk03Jr70M", "caption_count_category": 7, "clip": { "caption_index": 1, "caption": "A man with a beard, wearing white overalls and a cap, bends over by a window. He uses a trowel to scoop white plaster from a container onto a hawk tool held in his left hand before straightening up.", "source_interval_sec": [ 4.2, 11.7 ], "loom_clip_id": "v_00Dk03Jr70M_105_292", "clip_frame_interval_4fps": [ 105, 292 ], "sample_count_1fps": 8, "sampled_frames": [ { "frame": "00000", "timestamp_sec": 4.2, "offset_sec": 0, "rgb_path": "assets/cat_07/v_00Dk03Jr70M_105_292/0000s_00000_rgb.jpg", "mask_path": "assets/cat_07/v_00Dk03Jr70M_105_292/0000s_00000_mask.png", "mask_status": "available" }, { "frame": "00024", "timestamp_sec": 5.2, "offset_sec": 1, "rgb_path": "assets/cat_07/v_00Dk03Jr70M_105_292/0001s_00024_rgb.jpg", "mask_path": "assets/cat_07/v_00Dk03Jr70M_105_292/0001s_00024_mask.png", "mask_status": "available" }, { "frame": "00048", "timestamp_sec": 6.2, "offset_sec": 2, "rgb_path": "assets/cat_07/v_00Dk03Jr70M_105_292/0002s_00048_rgb.jpg", "mask_path": "assets/cat_07/v_00Dk03Jr70M_105_292/0002s_00048_mask.png", "mask_status": "available" }, { "frame": "00076", "timestamp_sec": 7.2, "offset_sec": 3, "rgb_path": "assets/cat_07/v_00Dk03Jr70M_105_292/0003s_00076_rgb.jpg", "mask_path": "assets/cat_07/v_00Dk03Jr70M_105_292/0003s_00076_mask.png", "mask_status": "available" }, { "frame": "00100", "timestamp_sec": 8.2, "offset_sec": 4, "rgb_path": "assets/cat_07/v_00Dk03Jr70M_105_292/0004s_00100_rgb.jpg", "mask_path": "assets/cat_07/v_00Dk03Jr70M_105_292/0004s_00100_mask.png", "mask_status": "available" }, { "frame": "00124", "timestamp_sec": 9.2, "offset_sec": 5, "rgb_path": "assets/cat_07/v_00Dk03Jr70M_105_292/0005s_00124_rgb.jpg", "mask_path": "assets/cat_07/v_00Dk03Jr70M_105_292/0005s_00124_mask.png", "mask_status": "available" }, { "frame": "00148", "timestamp_sec": 10.2, "offset_sec": 6, "rgb_path": "assets/cat_07/v_00Dk03Jr70M_105_292/0006s_00148_rgb.jpg", "mask_path": "assets/cat_07/v_00Dk03Jr70M_105_292/0006s_00148_mask.png", "mask_status": "available" }, { "frame": "00176", "timestamp_sec": 11.2, "offset_sec": 7, "rgb_path": "assets/cat_07/v_00Dk03Jr70M_105_292/0007s_00176_rgb.jpg", "mask_path": "assets/cat_07/v_00Dk03Jr70M_105_292/0007s_00176_mask.png", "mask_status": "available" } ], "obj_id": [ 0 ], "anno_id": [ 153 ], "fps_inferred_from_clip_interval": 24.9333 }, "usage": { "input_tokens": 21538, "input_tokens_details": { "cache_write_tokens": 21535, "cached_tokens": 0 }, "output_tokens": 4665, "output_tokens_details": { "reasoning_tokens": 2944 }, "total_tokens": 26203 }, "model": "gpt-5.6-sol", "effort": "medium", "custom_id": "cat07_cap01_v_00Dk03Jr70M_105_292", "annotation": { "caption": "A man with a beard, wearing white overalls and a cap, bends over by a window. He uses a trowel to scoop white plaster from a container onto a hawk tool held in his left hand before straightening up.", "entity_groups": [ { "id": "person_1", "label": "man", "spans": [ { "text": "A man", "range": [ 0, 5 ] }, { "text": "He", "range": [ 78, 80 ] }, { "text": "his", "range": [ 160, 163 ] } ], "source_mask_link": "selected_person" }, { "id": "beard_1", "label": "beard", "spans": [ { "text": "a beard", "range": [ 11, 18 ] } ], "source_mask_link": null }, { "id": "overalls_1", "label": "overalls", "spans": [ { "text": "white overalls", "range": [ 28, 42 ] } ], "source_mask_link": null }, { "id": "cap_1", "label": "cap", "spans": [ { "text": "a cap", "range": [ 47, 52 ] } ], "source_mask_link": null }, { "id": "window_1", "label": "window", "spans": [ { "text": "a window", "range": [ 68, 76 ] } ], "source_mask_link": null }, { "id": "trowel_1", "label": "trowel", "spans": [ { "text": "a trowel", "range": [ 86, 94 ] } ], "source_mask_link": null }, { "id": "plaster_1", "label": "plaster", "spans": [ { "text": "white plaster", "range": [ 104, 117 ] } ], "source_mask_link": null }, { "id": "container_1", "label": "container", "spans": [ { "text": "a container", "range": [ 123, 134 ] } ], "source_mask_link": null }, { "id": "hawk_tool_1", "label": "hawk tool", "spans": [ { "text": "a hawk tool", "range": [ 140, 151 ] } ], "source_mask_link": null }, { "id": "hand_1", "label": "left hand", "spans": [ { "text": "his left hand", "range": [ 160, 173 ] } ], "source_mask_link": null } ], "actions": [ { "id": "action_1", "mentions": [ { "text": "bends over", "ranges": [ [ 54, 64 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "type": "phrasal", "subject_refs": [ "person_1" ], "object_refs": [], "confidence": "high", "temporal_status": "localized_sparse_support", "intervals": [ { "start_s": 1.0, "end_s": 3.0, "source_start_s": 5.2, "source_end_s": 7.2 } ], "evidence": "At clip-relative 1.0–2.0 seconds, the masked man enters beside the window and lowers his torso into a pronounced bent posture.", "mask_relation": "The action subject matches the supplied selected-person mask.", "source_subject_refs": "person_1" }, { "id": "action_2", "mentions": [ { "text": "uses", "ranges": [ [ 81, 85 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "type": "simple", "subject_refs": [ "person_1" ], "object_refs": [ "trowel_1" ], "confidence": "medium", "temporal_status": "localized_sparse_support", "intervals": [ { "start_s": 1.0, "end_s": 6.0, "source_start_s": 5.2, "source_end_s": 10.2 } ], "evidence": "Across clip-relative 1.0–5.0 seconds, the masked man repeatedly works with his hands and a small tool over the work surface.", "mask_relation": "The action subject matches the supplied selected-person mask.", "source_subject_refs": "person_1" }, { "id": "action_3", "mentions": [ { "text": "scoop", "ranges": [ [ 98, 103 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "type": "simple", "subject_refs": [ "person_1" ], "object_refs": [ "plaster_1" ], "confidence": "medium", "temporal_status": "localized_sparse_support", "intervals": [ { "start_s": 1.0, "end_s": 6.0, "source_start_s": 5.2, "source_end_s": 10.2 } ], "evidence": "The clean frames at clip-relative 1.0–5.0 seconds show the masked man bent over the materials and making repeated hand-and-tool transfer motions.", "mask_relation": "The action subject matches the supplied selected-person mask.", "source_subject_refs": "person_1" }, { "id": "action_4", "mentions": [ { "text": "straightening up", "ranges": [ [ 181, 197 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "type": "phrasal", "subject_refs": [ "person_1" ], "object_refs": [], "confidence": "medium", "temporal_status": "localized_sparse_support", "intervals": [ { "start_s": 5.0, "end_s": 7.0, "source_start_s": 9.2, "source_end_s": 11.2 } ], "evidence": "Between clip-relative 5.0 and 6.0 seconds, the masked man's head and upper torso rise substantially from the lower bent position.", "mask_relation": "The action subject matches the supplied selected-person mask.", "source_subject_refs": "person_1" } ], "notes": [ "The supplied masklet identifies only the man; it is empty at clip-relative 0.0 seconds and tracks him in the subsequent samples.", "No separate source masks were supplied for the beard, clothing, cap, window, tools, plaster, container, or hand.", "The modifier \"wearing\" describes apparel rather than an eligible action or bodily posture and was not annotated.", "The bare passive \"held\" has the hawk tool as its stated subject; because that entity has no supplied mask, it was omitted under the masked-subject requirement." ] }, "model_output_raw": { "caption": "A man with a beard, wearing white overalls and a cap, bends over by a window. He uses a trowel to scoop white plaster from a container onto a hawk tool held in his left hand before straightening up.", "entity_groups": [ { "id": "person_1", "label": "man", "spans": [ { "text": "A man", "range": [ 0, 5 ] }, { "text": "He", "range": [ 78, 80 ] }, { "text": "his", "range": [ 160, 163 ] } ], "source_mask": { "id": "selected_person_masklet" } }, { "id": "beard_1", "label": "beard", "spans": [ { "text": "a beard", "range": [ 11, 18 ] } ], "source_mask": null }, { "id": "overalls_1", "label": "overalls", "spans": [ { "text": "white overalls", "range": [ 28, 42 ] } ], "source_mask": null }, { "id": "cap_1", "label": "cap", "spans": [ { "text": "a cap", "range": [ 47, 52 ] } ], "source_mask": null }, { "id": "window_1", "label": "window", "spans": [ { "text": "a window", "range": [ 68, 76 ] } ], "source_mask": null }, { "id": "trowel_1", "label": "trowel", "spans": [ { "text": "a trowel", "range": [ 86, 94 ] } ], "source_mask": null }, { "id": "plaster_1", "label": "plaster", "spans": [ { "text": "white plaster", "range": [ 104, 117 ] } ], "source_mask": null }, { "id": "container_1", "label": "container", "spans": [ { "text": "a container", "range": [ 123, 134 ] } ], "source_mask": null }, { "id": "hawk_tool_1", "label": "hawk tool", "spans": [ { "text": "a hawk tool", "range": [ 140, 151 ] } ], "source_mask": null }, { "id": "hand_1", "label": "left hand", "spans": [ { "text": "his left hand", "range": [ 160, 173 ] } ], "source_mask": null } ], "actions": [ { "mentions": [ { "text": "bends over", "ranges": [ [ 54, 64 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "type": "phrasal", "subject_ref": "person_1", "object_refs": [], "confidence": "high", "temporal_status": "localized_sparse_support", "intervals": [ { "start_s": 1.0, "end_s": 3.0, "source_start_s": 5.2, "source_end_s": 7.2 } ], "evidence": "At clip-relative 1.0–2.0 seconds, the masked man enters beside the window and lowers his torso into a pronounced bent posture.", "mask_relation": "The action subject matches the supplied selected-person mask." }, { "mentions": [ { "text": "uses", "ranges": [ [ 81, 85 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "type": "simple", "subject_ref": "person_1", "object_refs": [ "trowel_1" ], "confidence": "medium", "temporal_status": "localized_sparse_support", "intervals": [ { "start_s": 1.0, "end_s": 6.0, "source_start_s": 5.2, "source_end_s": 10.2 } ], "evidence": "Across clip-relative 1.0–5.0 seconds, the masked man repeatedly works with his hands and a small tool over the work surface.", "mask_relation": "The action subject matches the supplied selected-person mask." }, { "mentions": [ { "text": "scoop", "ranges": [ [ 98, 103 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "type": "simple", "subject_ref": "person_1", "object_refs": [ "plaster_1" ], "confidence": "medium", "temporal_status": "localized_sparse_support", "intervals": [ { "start_s": 1.0, "end_s": 6.0, "source_start_s": 5.2, "source_end_s": 10.2 } ], "evidence": "The clean frames at clip-relative 1.0–5.0 seconds show the masked man bent over the materials and making repeated hand-and-tool transfer motions.", "mask_relation": "The action subject matches the supplied selected-person mask." }, { "mentions": [ { "text": "straightening up", "ranges": [ [ 181, 197 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "type": "phrasal", "subject_ref": "person_1", "object_refs": [], "confidence": "medium", "temporal_status": "localized_sparse_support", "intervals": [ { "start_s": 5.0, "end_s": 7.0, "source_start_s": 9.2, "source_end_s": 11.2 } ], "evidence": "Between clip-relative 5.0 and 6.0 seconds, the masked man's head and upper torso rise substantially from the lower bent position.", "mask_relation": "The action subject matches the supplied selected-person mask." } ], "notes": [ "The supplied masklet identifies only the man; it is empty at clip-relative 0.0 seconds and tracks him in the subsequent samples.", "No separate source masks were supplied for the beard, clothing, cap, window, tools, plaster, container, or hand.", "The modifier \"wearing\" describes apparel rather than an eligible action or bodily posture and was not annotated.", "The bare passive \"held\" has the hawk tool as its stated subject; because that entity has no supplied mask, it was omitted under the masked-subject requirement." ] }, "normalization_repairs": 0 }, { "source_video_id": "v_00Dk03Jr70M", "caption_count_category": 7, "clip": { "caption_index": 2, "caption": "He turns to the wall and begins applying plaster with the trowel. Working from right to left, he spreads the plaster in a series of vertical strokes across the lower and middle sections of the wall. He creates several new patches as he moves along.", "source_interval_sec": [ 11.7, 44.7 ], "loom_clip_id": "v_00Dk03Jr70M_292_1118", "clip_frame_interval_4fps": [ 292, 1118 ], "sample_count_1fps": 33, "sampled_frames": [ { "frame": "00000", "timestamp_sec": 11.7, "offset_sec": 0, "rgb_path": "assets/cat_07/v_00Dk03Jr70M_292_1118/0000s_00000_rgb.jpg", "mask_path": "assets/cat_07/v_00Dk03Jr70M_292_1118/0000s_00000_mask.png", "mask_status": "available" }, { "frame": "00024", "timestamp_sec": 12.7, "offset_sec": 1, "rgb_path": "assets/cat_07/v_00Dk03Jr70M_292_1118/0001s_00024_rgb.jpg", "mask_path": "assets/cat_07/v_00Dk03Jr70M_292_1118/0001s_00024_mask.png", "mask_status": "available" }, { "frame": "00052", "timestamp_sec": 13.7, "offset_sec": 2, "rgb_path": "assets/cat_07/v_00Dk03Jr70M_292_1118/0002s_00052_rgb.jpg", "mask_path": "assets/cat_07/v_00Dk03Jr70M_292_1118/0002s_00052_mask.png", "mask_status": "available" }, { "frame": "00076", "timestamp_sec": 14.7, "offset_sec": 3, "rgb_path": "assets/cat_07/v_00Dk03Jr70M_292_1118/0003s_00076_rgb.jpg", "mask_path": "assets/cat_07/v_00Dk03Jr70M_292_1118/0003s_00076_mask.png", "mask_status": "available" }, { "frame": "00100", "timestamp_sec": 15.7, "offset_sec": 4, "rgb_path": "assets/cat_07/v_00Dk03Jr70M_292_1118/0004s_00100_rgb.jpg", "mask_path": "assets/cat_07/v_00Dk03Jr70M_292_1118/0004s_00100_mask.png", "mask_status": "available" }, { "frame": "00124", "timestamp_sec": 16.7, "offset_sec": 5, "rgb_path": "assets/cat_07/v_00Dk03Jr70M_292_1118/0005s_00124_rgb.jpg", "mask_path": "assets/cat_07/v_00Dk03Jr70M_292_1118/0005s_00124_mask.png", "mask_status": "available" }, { "frame": "00152", "timestamp_sec": 17.7, "offset_sec": 6, "rgb_path": "assets/cat_07/v_00Dk03Jr70M_292_1118/0006s_00152_rgb.jpg", "mask_path": "assets/cat_07/v_00Dk03Jr70M_292_1118/0006s_00152_mask.png", "mask_status": "available" }, { "frame": "00176", "timestamp_sec": 18.7, "offset_sec": 7, "rgb_path": "assets/cat_07/v_00Dk03Jr70M_292_1118/0007s_00176_rgb.jpg", "mask_path": "assets/cat_07/v_00Dk03Jr70M_292_1118/0007s_00176_mask.png", "mask_status": "available" }, { "frame": "00200", "timestamp_sec": 19.7, "offset_sec": 8, "rgb_path": "assets/cat_07/v_00Dk03Jr70M_292_1118/0008s_00200_rgb.jpg", "mask_path": "assets/cat_07/v_00Dk03Jr70M_292_1118/0008s_00200_mask.png", "mask_status": "available" }, { "frame": "00224", "timestamp_sec": 20.7, "offset_sec": 9, "rgb_path": "assets/cat_07/v_00Dk03Jr70M_292_1118/0009s_00224_rgb.jpg", "mask_path": "assets/cat_07/v_00Dk03Jr70M_292_1118/0009s_00224_mask.png", "mask_status": "available" }, { "frame": "00252", "timestamp_sec": 21.7, "offset_sec": 10, "rgb_path": "assets/cat_07/v_00Dk03Jr70M_292_1118/0010s_00252_rgb.jpg", "mask_path": "assets/cat_07/v_00Dk03Jr70M_292_1118/0010s_00252_mask.png", "mask_status": "available" }, { "frame": "00276", "timestamp_sec": 22.7, "offset_sec": 11, "rgb_path": "assets/cat_07/v_00Dk03Jr70M_292_1118/0011s_00276_rgb.jpg", "mask_path": "assets/cat_07/v_00Dk03Jr70M_292_1118/0011s_00276_mask.png", "mask_status": "available" }, { "frame": "00300", "timestamp_sec": 23.7, "offset_sec": 12, "rgb_path": "assets/cat_07/v_00Dk03Jr70M_292_1118/0012s_00300_rgb.jpg", "mask_path": "assets/cat_07/v_00Dk03Jr70M_292_1118/0012s_00300_mask.png", "mask_status": "available" }, { "frame": "00324", "timestamp_sec": 24.7, "offset_sec": 13, "rgb_path": "assets/cat_07/v_00Dk03Jr70M_292_1118/0013s_00324_rgb.jpg", "mask_path": "assets/cat_07/v_00Dk03Jr70M_292_1118/0013s_00324_mask.png", "mask_status": "available" }, { "frame": "00352", "timestamp_sec": 25.7, "offset_sec": 14, "rgb_path": "assets/cat_07/v_00Dk03Jr70M_292_1118/0014s_00352_rgb.jpg", "mask_path": "assets/cat_07/v_00Dk03Jr70M_292_1118/0014s_00352_mask.png", "mask_status": "available" }, { "frame": "00376", "timestamp_sec": 26.7, "offset_sec": 15, "rgb_path": "assets/cat_07/v_00Dk03Jr70M_292_1118/0015s_00376_rgb.jpg", "mask_path": "assets/cat_07/v_00Dk03Jr70M_292_1118/0015s_00376_mask.png", "mask_status": "available" }, { "frame": "00400", "timestamp_sec": 27.7, "offset_sec": 16, "rgb_path": "assets/cat_07/v_00Dk03Jr70M_292_1118/0016s_00400_rgb.jpg", "mask_path": "assets/cat_07/v_00Dk03Jr70M_292_1118/0016s_00400_mask.png", "mask_status": "available" }, { "frame": "00424", "timestamp_sec": 28.7, "offset_sec": 17, "rgb_path": "assets/cat_07/v_00Dk03Jr70M_292_1118/0017s_00424_rgb.jpg", "mask_path": "assets/cat_07/v_00Dk03Jr70M_292_1118/0017s_00424_mask.png", "mask_status": "available" }, { "frame": "00452", "timestamp_sec": 29.7, "offset_sec": 18, "rgb_path": "assets/cat_07/v_00Dk03Jr70M_292_1118/0018s_00452_rgb.jpg", "mask_path": "assets/cat_07/v_00Dk03Jr70M_292_1118/0018s_00452_mask.png", "mask_status": "available" }, { "frame": "00476", "timestamp_sec": 30.7, "offset_sec": 19, "rgb_path": "assets/cat_07/v_00Dk03Jr70M_292_1118/0019s_00476_rgb.jpg", "mask_path": "assets/cat_07/v_00Dk03Jr70M_292_1118/0019s_00476_mask.png", "mask_status": "available" }, { "frame": "00500", "timestamp_sec": 31.7, "offset_sec": 20, "rgb_path": "assets/cat_07/v_00Dk03Jr70M_292_1118/0020s_00500_rgb.jpg", "mask_path": "assets/cat_07/v_00Dk03Jr70M_292_1118/0020s_00500_mask.png", "mask_status": "available" }, { "frame": "00524", "timestamp_sec": 32.7, "offset_sec": 21, "rgb_path": "assets/cat_07/v_00Dk03Jr70M_292_1118/0021s_00524_rgb.jpg", "mask_path": "assets/cat_07/v_00Dk03Jr70M_292_1118/0021s_00524_mask.png", "mask_status": "available" }, { "frame": "00552", "timestamp_sec": 33.7, "offset_sec": 22, "rgb_path": "assets/cat_07/v_00Dk03Jr70M_292_1118/0022s_00552_rgb.jpg", "mask_path": "assets/cat_07/v_00Dk03Jr70M_292_1118/0022s_00552_mask.png", "mask_status": "available" }, { "frame": "00576", "timestamp_sec": 34.7, "offset_sec": 23, "rgb_path": "assets/cat_07/v_00Dk03Jr70M_292_1118/0023s_00576_rgb.jpg", "mask_path": "assets/cat_07/v_00Dk03Jr70M_292_1118/0023s_00576_mask.png", "mask_status": "available" }, { "frame": "00600", "timestamp_sec": 35.7, "offset_sec": 24, "rgb_path": "assets/cat_07/v_00Dk03Jr70M_292_1118/0024s_00600_rgb.jpg", "mask_path": "assets/cat_07/v_00Dk03Jr70M_292_1118/0024s_00600_mask.png", "mask_status": "available" }, { "frame": "00624", "timestamp_sec": 36.7, "offset_sec": 25, "rgb_path": "assets/cat_07/v_00Dk03Jr70M_292_1118/0025s_00624_rgb.jpg", "mask_path": "assets/cat_07/v_00Dk03Jr70M_292_1118/0025s_00624_mask.png", "mask_status": "available" }, { "frame": "00652", "timestamp_sec": 37.7, "offset_sec": 26, "rgb_path": "assets/cat_07/v_00Dk03Jr70M_292_1118/0026s_00652_rgb.jpg", "mask_path": "assets/cat_07/v_00Dk03Jr70M_292_1118/0026s_00652_mask.png", "mask_status": "available" }, { "frame": "00676", "timestamp_sec": 38.7, "offset_sec": 27, "rgb_path": "assets/cat_07/v_00Dk03Jr70M_292_1118/0027s_00676_rgb.jpg", "mask_path": "assets/cat_07/v_00Dk03Jr70M_292_1118/0027s_00676_mask.png", "mask_status": "available" }, { "frame": "00700", "timestamp_sec": 39.7, "offset_sec": 28, "rgb_path": "assets/cat_07/v_00Dk03Jr70M_292_1118/0028s_00700_rgb.jpg", "mask_path": "assets/cat_07/v_00Dk03Jr70M_292_1118/0028s_00700_mask.png", "mask_status": "available" }, { "frame": "00724", "timestamp_sec": 40.7, "offset_sec": 29, "rgb_path": "assets/cat_07/v_00Dk03Jr70M_292_1118/0029s_00724_rgb.jpg", "mask_path": "assets/cat_07/v_00Dk03Jr70M_292_1118/0029s_00724_mask.png", "mask_status": "available" }, { "frame": "00752", "timestamp_sec": 41.7, "offset_sec": 30, "rgb_path": "assets/cat_07/v_00Dk03Jr70M_292_1118/0030s_00752_rgb.jpg", "mask_path": "assets/cat_07/v_00Dk03Jr70M_292_1118/0030s_00752_mask.png", "mask_status": "available" }, { "frame": "00776", "timestamp_sec": 42.7, "offset_sec": 31, "rgb_path": "assets/cat_07/v_00Dk03Jr70M_292_1118/0031s_00776_rgb.jpg", "mask_path": "assets/cat_07/v_00Dk03Jr70M_292_1118/0031s_00776_mask.png", "mask_status": "available" }, { "frame": "00800", "timestamp_sec": 43.7, "offset_sec": 32, "rgb_path": "assets/cat_07/v_00Dk03Jr70M_292_1118/0032s_00800_rgb.jpg", "mask_path": "assets/cat_07/v_00Dk03Jr70M_292_1118/0032s_00800_mask.png", "mask_status": "available" } ], "obj_id": [ 0 ], "anno_id": [ 158 ], "fps_inferred_from_clip_interval": 25.0303 }, "usage": { "input_tokens": 40916, "input_tokens_details": { "cache_write_tokens": 40913, "cached_tokens": 0 }, "output_tokens": 4633, "output_tokens_details": { "reasoning_tokens": 3032 }, "total_tokens": 45549 }, "model": "gpt-5.6-sol", "effort": "medium", "custom_id": "cat07_cap02_v_00Dk03Jr70M_292_1118", "annotation": { "caption": "He turns to the wall and begins applying plaster with the trowel. Working from right to left, he spreads the plaster in a series of vertical strokes across the lower and middle sections of the wall. He creates several new patches as he moves along.", "entity_groups": [ { "id": "person_1", "label": "person", "spans": [ { "text": "He", "range": [ 0, 2 ] }, { "text": "he", "range": [ 94, 96 ] }, { "text": "He", "range": [ 199, 201 ] }, { "text": "he", "range": [ 233, 235 ] } ], "source_mask_link": "selected_person" }, { "id": "wall_1", "label": "wall", "spans": [ { "text": "the wall", "range": [ 12, 20 ] }, { "text": "the wall", "range": [ 189, 197 ] } ], "source_mask_link": null }, { "id": "plaster_1", "label": "plaster", "spans": [ { "text": "plaster", "range": [ 41, 48 ] }, { "text": "the plaster", "range": [ 105, 116 ] } ], "source_mask_link": null }, { "id": "trowel_1", "label": "trowel", "spans": [ { "text": "the trowel", "range": [ 54, 64 ] } ], "source_mask_link": null }, { "id": "patches_1", "label": "plaster patches", "spans": [ { "text": "several new patches", "range": [ 210, 229 ] } ], "source_mask_link": null } ], "actions": [ { "id": "action_1", "mentions": [ { "text": "turns", "ranges": [ [ 3, 8 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "type": "simple", "subject_refs": [ "person_1" ], "object_refs": [ "wall_1" ], "confidence": "medium", "temporal_status": "localized", "intervals": [ { "start_s": 0.0, "end_s": 2.0, "source_start_s": 11.7, "source_end_s": 13.7 } ], "evidence": "Across samples 0–2, the masked person changes from bending beside the work surface to orienting toward the wall.", "mask_relation": "subject matches the supplied selected-person mask", "source_subject_refs": "person_1" }, { "id": "action_2", "mentions": [ { "text": "applying", "ranges": [ [ 32, 40 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } }, { "text": "Working", "ranges": [ [ 66, 73 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } }, { "text": "spreads", "ranges": [ [ 97, 104 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "type": "simple", "subject_refs": [ "person_1" ], "object_refs": [ "plaster_1", "wall_1" ], "confidence": "high", "temporal_status": "localized", "intervals": [ { "start_s": 2.0, "end_s": 33.0, "source_start_s": 13.7, "source_end_s": 44.7 } ], "evidence": "From sample 2 through the final sample, the masked person repeatedly presses and draws the trowel over the wall while the white plastered area expands.", "mask_relation": "subject matches the supplied selected-person mask", "source_subject_refs": "person_1" }, { "id": "action_3", "mentions": [ { "text": "creates", "ranges": [ [ 202, 209 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "type": "simple", "subject_refs": [ "person_1" ], "object_refs": [ "patches_1" ], "confidence": "high", "temporal_status": "localized", "intervals": [ { "start_s": 2.0, "end_s": 33.0, "source_start_s": 13.7, "source_end_s": 44.7 } ], "evidence": "Successive clean frames show new white plaster regions appearing over previously yellow portions of the wall as the masked person works.", "mask_relation": "subject matches the supplied selected-person mask", "source_subject_refs": "person_1" }, { "id": "action_4", "mentions": [ { "text": "moves", "ranges": [ [ 236, 241 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "type": "simple", "subject_refs": [ "person_1" ], "object_refs": [], "confidence": "medium", "temporal_status": "localized", "intervals": [ { "start_s": 2.0, "end_s": 31.0, "source_start_s": 13.7, "source_end_s": 42.7 } ], "evidence": "The masked person changes lateral position and working height along the wall between repeated trowel strokes.", "mask_relation": "subject matches the supplied selected-person mask", "source_subject_refs": "person_1" } ], "notes": [ "The supplied mask consistently identifies the sole person described by all four pronoun mentions.", "No source masks were supplied for the wall, plaster, trowel, or plaster patches; their source-mask links are therefore null.", "The aspectual verb \"begins\" and the directional word \"along\" are excluded from action spans.", "The plaster-application descriptions \"applying\", \"Working\", and \"spreads\" are grouped as mentions of the same ongoing plastering event." ] }, "model_output_raw": { "caption": "He turns to the wall and begins applying plaster with the trowel. Working from right to left, he spreads the plaster in a series of vertical strokes across the lower and middle sections of the wall. He creates several new patches as he moves along.", "entity_groups": [ { "id": "person_1", "label": "person", "spans": [ { "text": "He", "range": [ 0, 2 ] }, { "text": "he", "range": [ 94, 96 ] }, { "text": "He", "range": [ 199, 201 ] }, { "text": "he", "range": [ 233, 235 ] } ], "source_mask": "selected_person_mask" }, { "id": "wall_1", "label": "wall", "spans": [ { "text": "the wall", "range": [ 12, 20 ] }, { "text": "the wall", "range": [ 189, 197 ] } ], "source_mask": null }, { "id": "plaster_1", "label": "plaster", "spans": [ { "text": "plaster", "range": [ 41, 48 ] }, { "text": "the plaster", "range": [ 105, 116 ] } ], "source_mask": null }, { "id": "trowel_1", "label": "trowel", "spans": [ { "text": "the trowel", "range": [ 54, 64 ] } ], "source_mask": null }, { "id": "patches_1", "label": "plaster patches", "spans": [ { "text": "several new patches", "range": [ 210, 229 ] } ], "source_mask": null } ], "actions": [ { "id": "action_1", "mentions": [ { "text": "turns", "ranges": [ [ 3, 8 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "type": "simple", "subject_ref": "person_1", "object_refs": [ "wall_1" ], "confidence": "medium", "mask_relation": "subject matches the supplied selected-person mask", "temporal_status": "localized", "intervals": [ { "start_s": 0.0, "end_s": 2.0, "source_start_s": 11.7, "source_end_s": 13.7 } ], "evidence": "Across samples 0–2, the masked person changes from bending beside the work surface to orienting toward the wall." }, { "id": "action_2", "mentions": [ { "text": "applying", "ranges": [ [ 32, 40 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } }, { "text": "Working", "ranges": [ [ 66, 73 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } }, { "text": "spreads", "ranges": [ [ 97, 104 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "type": "simple", "subject_ref": "person_1", "object_refs": [ "plaster_1", "wall_1" ], "instrument_refs": [ "trowel_1" ], "confidence": "high", "mask_relation": "subject matches the supplied selected-person mask", "temporal_status": "localized", "intervals": [ { "start_s": 2.0, "end_s": 33.0, "source_start_s": 13.7, "source_end_s": 44.7 } ], "evidence": "From sample 2 through the final sample, the masked person repeatedly presses and draws the trowel over the wall while the white plastered area expands." }, { "id": "action_3", "mentions": [ { "text": "creates", "ranges": [ [ 202, 209 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "type": "simple", "subject_ref": "person_1", "object_refs": [ "patches_1" ], "confidence": "high", "mask_relation": "subject matches the supplied selected-person mask", "temporal_status": "localized", "intervals": [ { "start_s": 2.0, "end_s": 33.0, "source_start_s": 13.7, "source_end_s": 44.7 } ], "evidence": "Successive clean frames show new white plaster regions appearing over previously yellow portions of the wall as the masked person works." }, { "id": "action_4", "mentions": [ { "text": "moves", "ranges": [ [ 236, 241 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "type": "simple", "subject_ref": "person_1", "object_refs": [], "confidence": "medium", "mask_relation": "subject matches the supplied selected-person mask", "temporal_status": "localized", "intervals": [ { "start_s": 2.0, "end_s": 31.0, "source_start_s": 13.7, "source_end_s": 42.7 } ], "evidence": "The masked person changes lateral position and working height along the wall between repeated trowel strokes." } ], "notes": [ "The supplied mask consistently identifies the sole person described by all four pronoun mentions.", "No source masks were supplied for the wall, plaster, trowel, or plaster patches; their source-mask links are therefore null.", "The aspectual verb \"begins\" and the directional word \"along\" are excluded from action spans.", "The plaster-application descriptions \"applying\", \"Working\", and \"spreads\" are grouped as mentions of the same ongoing plastering event." ] }, "normalization_repairs": 0 }, { "source_video_id": "v_00Dk03Jr70M", "caption_count_category": 7, "clip": { "caption_index": 3, "caption": "After using the plaster on his hawk, he turns back towards the window. He bends over, using his trowel to reload the hawk with more plaster from the container.", "source_interval_sec": [ 44.7, 54.7 ], "loom_clip_id": "v_00Dk03Jr70M_1118_1368", "clip_frame_interval_4fps": [ 1118, 1368 ], "sample_count_1fps": 10, "sampled_frames": [ { "frame": "00000", "timestamp_sec": 44.7, "offset_sec": 0, "rgb_path": "assets/cat_07/v_00Dk03Jr70M_1118_1368/0000s_00000_rgb.jpg", "mask_path": "assets/cat_07/v_00Dk03Jr70M_1118_1368/0000s_00000_mask.png", "mask_status": "available" }, { "frame": "00024", "timestamp_sec": 45.7, "offset_sec": 1, "rgb_path": "assets/cat_07/v_00Dk03Jr70M_1118_1368/0001s_00024_rgb.jpg", "mask_path": "assets/cat_07/v_00Dk03Jr70M_1118_1368/0001s_00024_mask.png", "mask_status": "available" }, { "frame": "00048", "timestamp_sec": 46.7, "offset_sec": 2, "rgb_path": "assets/cat_07/v_00Dk03Jr70M_1118_1368/0002s_00048_rgb.jpg", "mask_path": "assets/cat_07/v_00Dk03Jr70M_1118_1368/0002s_00048_mask.png", "mask_status": "available" }, { "frame": "00076", "timestamp_sec": 47.7, "offset_sec": 3, "rgb_path": "assets/cat_07/v_00Dk03Jr70M_1118_1368/0003s_00076_rgb.jpg", "mask_path": "assets/cat_07/v_00Dk03Jr70M_1118_1368/0003s_00076_mask.png", "mask_status": "available" }, { "frame": "00100", "timestamp_sec": 48.7, "offset_sec": 4, "rgb_path": "assets/cat_07/v_00Dk03Jr70M_1118_1368/0004s_00100_rgb.jpg", "mask_path": "assets/cat_07/v_00Dk03Jr70M_1118_1368/0004s_00100_mask.png", "mask_status": "available" }, { "frame": "00124", "timestamp_sec": 49.7, "offset_sec": 5, "rgb_path": "assets/cat_07/v_00Dk03Jr70M_1118_1368/0005s_00124_rgb.jpg", "mask_path": "assets/cat_07/v_00Dk03Jr70M_1118_1368/0005s_00124_mask.png", "mask_status": "available" }, { "frame": "00148", "timestamp_sec": 50.7, "offset_sec": 6, "rgb_path": "assets/cat_07/v_00Dk03Jr70M_1118_1368/0006s_00148_rgb.jpg", "mask_path": "assets/cat_07/v_00Dk03Jr70M_1118_1368/0006s_00148_mask.png", "mask_status": "available" }, { "frame": "00176", "timestamp_sec": 51.7, "offset_sec": 7, "rgb_path": "assets/cat_07/v_00Dk03Jr70M_1118_1368/0007s_00176_rgb.jpg", "mask_path": "assets/cat_07/v_00Dk03Jr70M_1118_1368/0007s_00176_mask.png", "mask_status": "available" }, { "frame": "00200", "timestamp_sec": 52.7, "offset_sec": 8, "rgb_path": "assets/cat_07/v_00Dk03Jr70M_1118_1368/0008s_00200_rgb.jpg", "mask_path": "assets/cat_07/v_00Dk03Jr70M_1118_1368/0008s_00200_mask.png", "mask_status": "available" }, { "frame": "00224", "timestamp_sec": 53.7, "offset_sec": 9, "rgb_path": "assets/cat_07/v_00Dk03Jr70M_1118_1368/0009s_00224_rgb.jpg", "mask_path": "assets/cat_07/v_00Dk03Jr70M_1118_1368/0009s_00224_mask.png", "mask_status": "available" } ], "obj_id": [ 0 ], "anno_id": [ 154 ], "fps_inferred_from_clip_interval": 25.0 }, "usage": { "input_tokens": 23077, "input_tokens_details": { "cache_write_tokens": 23074, "cached_tokens": 0 }, "output_tokens": 3784, "output_tokens_details": { "reasoning_tokens": 2070 }, "total_tokens": 26861 }, "model": "gpt-5.6-sol", "effort": "medium", "custom_id": "cat07_cap03_v_00Dk03Jr70M_1118_1368", "annotation": { "caption": "After using the plaster on his hawk, he turns back towards the window. He bends over, using his trowel to reload the hawk with more plaster from the container.", "entity_groups": [ { "id": "person_1", "label": "person", "spans": [ { "text": "his", "range": [ 27, 30 ] }, { "text": "he", "range": [ 37, 39 ] }, { "text": "He", "range": [ 71, 73 ] }, { "text": "his", "range": [ 92, 95 ] } ], "source_mask_link": "selected_person" }, { "id": "plaster_1", "label": "plaster", "spans": [ { "text": "the plaster", "range": [ 12, 23 ] }, { "text": "more plaster", "range": [ 127, 139 ] } ], "source_mask_link": null }, { "id": "hawk_1", "label": "hawk", "spans": [ { "text": "his hawk", "range": [ 27, 35 ] }, { "text": "the hawk", "range": [ 113, 121 ] } ], "source_mask_link": null }, { "id": "window_1", "label": "window", "spans": [ { "text": "the window", "range": [ 59, 69 ] } ], "source_mask_link": null }, { "id": "trowel_1", "label": "trowel", "spans": [ { "text": "his trowel", "range": [ 92, 102 ] } ], "source_mask_link": null }, { "id": "container_1", "label": "container", "spans": [ { "text": "the container", "range": [ 145, 158 ] } ], "source_mask_link": null } ], "actions": [ { "id": "action_1", "mentions": [ { "text": "using", "ranges": [ [ 6, 11 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "type": "simple", "subject_refs": [ "person_1" ], "object_refs": [ "plaster_1", "hawk_1" ], "confidence": "medium", "temporal_status": "localized", "intervals": [ { "start_s": 0.0, "end_s": 3.0, "source_start_s": 44.7, "source_end_s": 47.7 } ], "evidence": "At samples 0–2, the masked person works beside the wall while holding the hawk and manipulating material with his other hand.", "mask_relation": "subject matches the supplied selected-person mask", "source_subject_refs": "person_1" }, { "id": "action_2", "mentions": [ { "text": "turns", "ranges": [ [ 40, 45 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "type": "simple", "subject_refs": [ "person_1" ], "object_refs": [ "window_1" ], "confidence": "high", "temporal_status": "localized", "intervals": [ { "start_s": 2.0, "end_s": 4.0, "source_start_s": 46.7, "source_end_s": 48.7 } ], "evidence": "Across samples 2–3, the masked person changes orientation from the wall toward the window/work area.", "mask_relation": "subject matches the supplied selected-person mask", "source_subject_refs": "person_1" }, { "id": "action_3", "mentions": [ { "text": "bends over", "ranges": [ [ 74, 84 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "type": "phrasal", "subject_refs": [ "person_1" ], "object_refs": [], "confidence": "high", "temporal_status": "localized", "intervals": [ { "start_s": 4.0, "end_s": 10.0, "source_start_s": 48.7, "source_end_s": 54.7 } ], "evidence": "The masked person is visibly bent forward over the work surface in samples 4–9, including the final sampled frame.", "mask_relation": "subject matches the supplied selected-person mask", "source_subject_refs": "person_1" }, { "id": "action_4", "mentions": [ { "text": "using", "ranges": [ [ 86, 91 ] ], "occurrence": 2, "particle": { "text": null, "occurrence": 1 } } ], "type": "simple", "subject_refs": [ "person_1" ], "object_refs": [ "trowel_1" ], "confidence": "medium", "temporal_status": "localized", "intervals": [ { "start_s": 3.0, "end_s": 10.0, "source_start_s": 47.7, "source_end_s": 54.7 } ], "evidence": "In samples 3–9, the masked person repeatedly moves a hand/tool at the work surface; the small trowel is only partly discernible.", "mask_relation": "subject matches the supplied selected-person mask", "source_subject_refs": "person_1" }, { "id": "action_5", "mentions": [ { "text": "reload", "ranges": [ [ 106, 112 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "type": "simple", "subject_refs": [ "person_1" ], "object_refs": [ "hawk_1", "plaster_1", "container_1" ], "confidence": "medium", "temporal_status": "localized", "intervals": [ { "start_s": 3.0, "end_s": 10.0, "source_start_s": 47.7, "source_end_s": 54.7 } ], "evidence": "Samples 3–9 show the masked person bent over and manipulating the hawk/tool near the material source, consistent with loading plaster onto the hawk.", "mask_relation": "subject matches the supplied selected-person mask", "source_subject_refs": "person_1" } ], "notes": [ "The supplied mask consistently identifies the single person mentioned by “his,” “he,” and “He.”", "No separate source masks were supplied for the plaster, hawk, window, trowel, or container.", "Temporal intervals are sparse support intervals based on the sampled clean RGB frames; the final bending and loading-related intervals end at the clip boundary because support remains visible in the last sample." ] }, "model_output_raw": { "caption": "After using the plaster on his hawk, he turns back towards the window. He bends over, using his trowel to reload the hawk with more plaster from the container.", "entity_groups": [ { "id": "person_1", "entity_type": "person", "spans": [ { "text": "his", "range": [ 27, 30 ] }, { "text": "he", "range": [ 37, 39 ] }, { "text": "He", "range": [ 71, 73 ] }, { "text": "his", "range": [ 92, 95 ] } ], "source_mask": "selected_person_mask" }, { "id": "plaster_1", "entity_type": "plaster", "spans": [ { "text": "the plaster", "range": [ 12, 23 ] }, { "text": "more plaster", "range": [ 127, 139 ] } ], "source_mask": null }, { "id": "hawk_1", "entity_type": "hawk", "spans": [ { "text": "his hawk", "range": [ 27, 35 ] }, { "text": "the hawk", "range": [ 113, 121 ] } ], "source_mask": null }, { "id": "window_1", "entity_type": "window", "spans": [ { "text": "the window", "range": [ 59, 69 ] } ], "source_mask": null }, { "id": "trowel_1", "entity_type": "trowel", "spans": [ { "text": "his trowel", "range": [ 92, 102 ] } ], "source_mask": null }, { "id": "container_1", "entity_type": "container", "spans": [ { "text": "the container", "range": [ 145, 158 ] } ], "source_mask": null } ], "actions": [ { "id": "action_1", "mentions": [ { "text": "using", "ranges": [ [ 6, 11 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "type": "simple", "subject_entity_group": "person_1", "object_entity_groups": [ "plaster_1", "hawk_1" ], "confidence": "medium", "mask_relation": "subject matches the supplied selected-person mask", "temporal_status": "localized", "intervals": [ { "clip_relative_s": [ 0.0, 3.0 ], "source_video_s": [ 44.7, 47.7 ] } ], "evidence": "At samples 0–2, the masked person works beside the wall while holding the hawk and manipulating material with his other hand." }, { "id": "action_2", "mentions": [ { "text": "turns", "ranges": [ [ 40, 45 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "type": "simple", "subject_entity_group": "person_1", "object_entity_groups": [ "window_1" ], "confidence": "high", "mask_relation": "subject matches the supplied selected-person mask", "temporal_status": "localized", "intervals": [ { "clip_relative_s": [ 2.0, 4.0 ], "source_video_s": [ 46.7, 48.7 ] } ], "evidence": "Across samples 2–3, the masked person changes orientation from the wall toward the window/work area." }, { "id": "action_3", "mentions": [ { "text": "bends over", "ranges": [ [ 74, 84 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "type": "phrasal", "subject_entity_group": "person_1", "object_entity_groups": [], "confidence": "high", "mask_relation": "subject matches the supplied selected-person mask", "temporal_status": "localized", "intervals": [ { "clip_relative_s": [ 4.0, 10.0 ], "source_video_s": [ 48.7, 54.7 ] } ], "evidence": "The masked person is visibly bent forward over the work surface in samples 4–9, including the final sampled frame." }, { "id": "action_4", "mentions": [ { "text": "using", "ranges": [ [ 86, 91 ] ], "occurrence": 2, "particle": { "text": null, "occurrence": 1 } } ], "type": "simple", "subject_entity_group": "person_1", "object_entity_groups": [ "trowel_1" ], "confidence": "medium", "mask_relation": "subject matches the supplied selected-person mask", "temporal_status": "localized", "intervals": [ { "clip_relative_s": [ 3.0, 10.0 ], "source_video_s": [ 47.7, 54.7 ] } ], "evidence": "In samples 3–9, the masked person repeatedly moves a hand/tool at the work surface; the small trowel is only partly discernible." }, { "id": "action_5", "mentions": [ { "text": "reload", "ranges": [ [ 106, 112 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "type": "simple", "subject_entity_group": "person_1", "object_entity_groups": [ "hawk_1", "plaster_1", "container_1" ], "confidence": "medium", "mask_relation": "subject matches the supplied selected-person mask", "temporal_status": "localized", "intervals": [ { "clip_relative_s": [ 3.0, 10.0 ], "source_video_s": [ 47.7, 54.7 ] } ], "evidence": "Samples 3–9 show the masked person bent over and manipulating the hawk/tool near the material source, consistent with loading plaster onto the hawk." } ], "notes": [ "The supplied mask consistently identifies the single person mentioned by “his,” “he,” and “He.”", "No separate source masks were supplied for the plaster, hawk, window, trowel, or container.", "Temporal intervals are sparse support intervals based on the sampled clean RGB frames; the final bending and loading-related intervals end at the clip boundary because support remains visible in the last sample." ] }, "normalization_repairs": 0 }, { "source_video_id": "v_00Dk03Jr70M", "caption_count_category": 7, "clip": { "caption_index": 4, "caption": "He turns back to the wall and continues plastering, moving from the corner outwards. He applies the material in long vertical strips, using the trowel to smooth the surface with broad upward and downward motions.", "source_interval_sec": [ 54.7, 100.2 ], "loom_clip_id": "v_00Dk03Jr70M_1368_2505", "clip_frame_interval_4fps": [ 1368, 2505 ], "sample_count_1fps": 46, "sampled_frames": [ { "frame": "00000", "timestamp_sec": 54.7, "offset_sec": 0, "rgb_path": "assets/cat_07/v_00Dk03Jr70M_1368_2505/0000s_00000_rgb.jpg", "mask_path": "assets/cat_07/v_00Dk03Jr70M_1368_2505/0000s_00000_mask.png", "mask_status": "available" }, { "frame": "00024", "timestamp_sec": 55.7, "offset_sec": 1, "rgb_path": "assets/cat_07/v_00Dk03Jr70M_1368_2505/0001s_00024_rgb.jpg", "mask_path": "assets/cat_07/v_00Dk03Jr70M_1368_2505/0001s_00024_mask.png", "mask_status": "available" }, { "frame": "00048", "timestamp_sec": 56.7, "offset_sec": 2, "rgb_path": "assets/cat_07/v_00Dk03Jr70M_1368_2505/0002s_00048_rgb.jpg", "mask_path": "assets/cat_07/v_00Dk03Jr70M_1368_2505/0002s_00048_mask.png", "mask_status": "available" }, { "frame": "00076", "timestamp_sec": 57.7, "offset_sec": 3, "rgb_path": "assets/cat_07/v_00Dk03Jr70M_1368_2505/0003s_00076_rgb.jpg", "mask_path": "assets/cat_07/v_00Dk03Jr70M_1368_2505/0003s_00076_mask.png", "mask_status": "available" }, { "frame": "00100", "timestamp_sec": 58.7, "offset_sec": 4, "rgb_path": "assets/cat_07/v_00Dk03Jr70M_1368_2505/0004s_00100_rgb.jpg", "mask_path": "assets/cat_07/v_00Dk03Jr70M_1368_2505/0004s_00100_mask.png", "mask_status": "available" }, { "frame": "00124", "timestamp_sec": 59.7, "offset_sec": 5, "rgb_path": "assets/cat_07/v_00Dk03Jr70M_1368_2505/0005s_00124_rgb.jpg", "mask_path": "assets/cat_07/v_00Dk03Jr70M_1368_2505/0005s_00124_mask.png", "mask_status": "available" }, { "frame": "00148", "timestamp_sec": 60.7, "offset_sec": 6, "rgb_path": "assets/cat_07/v_00Dk03Jr70M_1368_2505/0006s_00148_rgb.jpg", "mask_path": "assets/cat_07/v_00Dk03Jr70M_1368_2505/0006s_00148_mask.png", "mask_status": "available" }, { "frame": "00176", "timestamp_sec": 61.7, "offset_sec": 7, "rgb_path": "assets/cat_07/v_00Dk03Jr70M_1368_2505/0007s_00176_rgb.jpg", "mask_path": "assets/cat_07/v_00Dk03Jr70M_1368_2505/0007s_00176_mask.png", "mask_status": "available" }, { "frame": "00200", "timestamp_sec": 62.7, "offset_sec": 8, "rgb_path": "assets/cat_07/v_00Dk03Jr70M_1368_2505/0008s_00200_rgb.jpg", "mask_path": "assets/cat_07/v_00Dk03Jr70M_1368_2505/0008s_00200_mask.png", "mask_status": "available" }, { "frame": "00224", "timestamp_sec": 63.7, "offset_sec": 9, "rgb_path": "assets/cat_07/v_00Dk03Jr70M_1368_2505/0009s_00224_rgb.jpg", "mask_path": "assets/cat_07/v_00Dk03Jr70M_1368_2505/0009s_00224_mask.png", "mask_status": "available" }, { "frame": "00248", "timestamp_sec": 64.7, "offset_sec": 10, "rgb_path": "assets/cat_07/v_00Dk03Jr70M_1368_2505/0010s_00248_rgb.jpg", "mask_path": "assets/cat_07/v_00Dk03Jr70M_1368_2505/0010s_00248_mask.png", "mask_status": "available" }, { "frame": "00276", "timestamp_sec": 65.7, "offset_sec": 11, "rgb_path": "assets/cat_07/v_00Dk03Jr70M_1368_2505/0011s_00276_rgb.jpg", "mask_path": "assets/cat_07/v_00Dk03Jr70M_1368_2505/0011s_00276_mask.png", "mask_status": "available" }, { "frame": "00300", "timestamp_sec": 66.7, "offset_sec": 12, "rgb_path": "assets/cat_07/v_00Dk03Jr70M_1368_2505/0012s_00300_rgb.jpg", "mask_path": "assets/cat_07/v_00Dk03Jr70M_1368_2505/0012s_00300_mask.png", "mask_status": "available" }, { "frame": "00324", "timestamp_sec": 67.7, "offset_sec": 13, "rgb_path": "assets/cat_07/v_00Dk03Jr70M_1368_2505/0013s_00324_rgb.jpg", "mask_path": "assets/cat_07/v_00Dk03Jr70M_1368_2505/0013s_00324_mask.png", "mask_status": "available" }, { "frame": "00348", "timestamp_sec": 68.7, "offset_sec": 14, "rgb_path": "assets/cat_07/v_00Dk03Jr70M_1368_2505/0014s_00348_rgb.jpg", "mask_path": "assets/cat_07/v_00Dk03Jr70M_1368_2505/0014s_00348_mask.png", "mask_status": "available" }, { "frame": "00376", "timestamp_sec": 69.7, "offset_sec": 15, "rgb_path": "assets/cat_07/v_00Dk03Jr70M_1368_2505/0015s_00376_rgb.jpg", "mask_path": "assets/cat_07/v_00Dk03Jr70M_1368_2505/0015s_00376_mask.png", "mask_status": "available" }, { "frame": "00400", "timestamp_sec": 70.7, "offset_sec": 16, "rgb_path": "assets/cat_07/v_00Dk03Jr70M_1368_2505/0016s_00400_rgb.jpg", "mask_path": "assets/cat_07/v_00Dk03Jr70M_1368_2505/0016s_00400_mask.png", "mask_status": "available" }, { "frame": "00424", "timestamp_sec": 71.7, "offset_sec": 17, "rgb_path": "assets/cat_07/v_00Dk03Jr70M_1368_2505/0017s_00424_rgb.jpg", "mask_path": "assets/cat_07/v_00Dk03Jr70M_1368_2505/0017s_00424_mask.png", "mask_status": "available" }, { "frame": "00448", "timestamp_sec": 72.7, "offset_sec": 18, "rgb_path": "assets/cat_07/v_00Dk03Jr70M_1368_2505/0018s_00448_rgb.jpg", "mask_path": "assets/cat_07/v_00Dk03Jr70M_1368_2505/0018s_00448_mask.png", "mask_status": "available" }, { "frame": "00476", "timestamp_sec": 73.7, "offset_sec": 19, "rgb_path": "assets/cat_07/v_00Dk03Jr70M_1368_2505/0019s_00476_rgb.jpg", "mask_path": "assets/cat_07/v_00Dk03Jr70M_1368_2505/0019s_00476_mask.png", "mask_status": "available" }, { "frame": "00500", "timestamp_sec": 74.7, "offset_sec": 20, "rgb_path": "assets/cat_07/v_00Dk03Jr70M_1368_2505/0020s_00500_rgb.jpg", "mask_path": "assets/cat_07/v_00Dk03Jr70M_1368_2505/0020s_00500_mask.png", "mask_status": "available" }, { "frame": "00524", "timestamp_sec": 75.7, "offset_sec": 21, "rgb_path": "assets/cat_07/v_00Dk03Jr70M_1368_2505/0021s_00524_rgb.jpg", "mask_path": "assets/cat_07/v_00Dk03Jr70M_1368_2505/0021s_00524_mask.png", "mask_status": "available" }, { "frame": "00548", "timestamp_sec": 76.7, "offset_sec": 22, "rgb_path": "assets/cat_07/v_00Dk03Jr70M_1368_2505/0022s_00548_rgb.jpg", "mask_path": "assets/cat_07/v_00Dk03Jr70M_1368_2505/0022s_00548_mask.png", "mask_status": "available" }, { "frame": "00576", "timestamp_sec": 77.7, "offset_sec": 23, "rgb_path": "assets/cat_07/v_00Dk03Jr70M_1368_2505/0023s_00576_rgb.jpg", "mask_path": "assets/cat_07/v_00Dk03Jr70M_1368_2505/0023s_00576_mask.png", "mask_status": "available" }, { "frame": "00600", "timestamp_sec": 78.7, "offset_sec": 24, "rgb_path": "assets/cat_07/v_00Dk03Jr70M_1368_2505/0024s_00600_rgb.jpg", "mask_path": "assets/cat_07/v_00Dk03Jr70M_1368_2505/0024s_00600_mask.png", "mask_status": "available" }, { "frame": "00624", "timestamp_sec": 79.7, "offset_sec": 25, "rgb_path": "assets/cat_07/v_00Dk03Jr70M_1368_2505/0025s_00624_rgb.jpg", "mask_path": "assets/cat_07/v_00Dk03Jr70M_1368_2505/0025s_00624_mask.png", "mask_status": "available" }, { "frame": "00648", "timestamp_sec": 80.7, "offset_sec": 26, "rgb_path": "assets/cat_07/v_00Dk03Jr70M_1368_2505/0026s_00648_rgb.jpg", "mask_path": "assets/cat_07/v_00Dk03Jr70M_1368_2505/0026s_00648_mask.png", "mask_status": "available" }, { "frame": "00676", "timestamp_sec": 81.7, "offset_sec": 27, "rgb_path": "assets/cat_07/v_00Dk03Jr70M_1368_2505/0027s_00676_rgb.jpg", "mask_path": "assets/cat_07/v_00Dk03Jr70M_1368_2505/0027s_00676_mask.png", "mask_status": "available" }, { "frame": "00700", "timestamp_sec": 82.7, "offset_sec": 28, "rgb_path": "assets/cat_07/v_00Dk03Jr70M_1368_2505/0028s_00700_rgb.jpg", "mask_path": "assets/cat_07/v_00Dk03Jr70M_1368_2505/0028s_00700_mask.png", "mask_status": "available" }, { "frame": "00724", "timestamp_sec": 83.7, "offset_sec": 29, "rgb_path": "assets/cat_07/v_00Dk03Jr70M_1368_2505/0029s_00724_rgb.jpg", "mask_path": "assets/cat_07/v_00Dk03Jr70M_1368_2505/0029s_00724_mask.png", "mask_status": "available" }, { "frame": "00748", "timestamp_sec": 84.7, "offset_sec": 30, "rgb_path": "assets/cat_07/v_00Dk03Jr70M_1368_2505/0030s_00748_rgb.jpg", "mask_path": "assets/cat_07/v_00Dk03Jr70M_1368_2505/0030s_00748_mask.png", "mask_status": "available" }, { "frame": "00776", "timestamp_sec": 85.7, "offset_sec": 31, "rgb_path": "assets/cat_07/v_00Dk03Jr70M_1368_2505/0031s_00776_rgb.jpg", "mask_path": "assets/cat_07/v_00Dk03Jr70M_1368_2505/0031s_00776_mask.png", "mask_status": "available" }, { "frame": "00800", "timestamp_sec": 86.7, "offset_sec": 32, "rgb_path": "assets/cat_07/v_00Dk03Jr70M_1368_2505/0032s_00800_rgb.jpg", "mask_path": "assets/cat_07/v_00Dk03Jr70M_1368_2505/0032s_00800_mask.png", "mask_status": "available" }, { "frame": "00824", "timestamp_sec": 87.7, "offset_sec": 33, "rgb_path": "assets/cat_07/v_00Dk03Jr70M_1368_2505/0033s_00824_rgb.jpg", "mask_path": "assets/cat_07/v_00Dk03Jr70M_1368_2505/0033s_00824_mask.png", "mask_status": "available" }, { "frame": "00848", "timestamp_sec": 88.7, "offset_sec": 34, "rgb_path": "assets/cat_07/v_00Dk03Jr70M_1368_2505/0034s_00848_rgb.jpg", "mask_path": "assets/cat_07/v_00Dk03Jr70M_1368_2505/0034s_00848_mask.png", "mask_status": "available" }, { "frame": "00876", "timestamp_sec": 89.7, "offset_sec": 35, "rgb_path": "assets/cat_07/v_00Dk03Jr70M_1368_2505/0035s_00876_rgb.jpg", "mask_path": "assets/cat_07/v_00Dk03Jr70M_1368_2505/0035s_00876_mask.png", "mask_status": "available" }, { "frame": "00900", "timestamp_sec": 90.7, "offset_sec": 36, "rgb_path": "assets/cat_07/v_00Dk03Jr70M_1368_2505/0036s_00900_rgb.jpg", "mask_path": "assets/cat_07/v_00Dk03Jr70M_1368_2505/0036s_00900_mask.png", "mask_status": "available" }, { "frame": "00924", "timestamp_sec": 91.7, "offset_sec": 37, "rgb_path": "assets/cat_07/v_00Dk03Jr70M_1368_2505/0037s_00924_rgb.jpg", "mask_path": "assets/cat_07/v_00Dk03Jr70M_1368_2505/0037s_00924_mask.png", "mask_status": "available" }, { "frame": "00948", "timestamp_sec": 92.7, "offset_sec": 38, "rgb_path": "assets/cat_07/v_00Dk03Jr70M_1368_2505/0038s_00948_rgb.jpg", "mask_path": "assets/cat_07/v_00Dk03Jr70M_1368_2505/0038s_00948_mask.png", "mask_status": "available" }, { "frame": "00976", "timestamp_sec": 93.7, "offset_sec": 39, "rgb_path": "assets/cat_07/v_00Dk03Jr70M_1368_2505/0039s_00976_rgb.jpg", "mask_path": "assets/cat_07/v_00Dk03Jr70M_1368_2505/0039s_00976_mask.png", "mask_status": "available" }, { "frame": "01000", "timestamp_sec": 94.7, "offset_sec": 40, "rgb_path": "assets/cat_07/v_00Dk03Jr70M_1368_2505/0040s_01000_rgb.jpg", "mask_path": "assets/cat_07/v_00Dk03Jr70M_1368_2505/0040s_01000_mask.png", "mask_status": "available" }, { "frame": "01024", "timestamp_sec": 95.7, "offset_sec": 41, "rgb_path": "assets/cat_07/v_00Dk03Jr70M_1368_2505/0041s_01024_rgb.jpg", "mask_path": "assets/cat_07/v_00Dk03Jr70M_1368_2505/0041s_01024_mask.png", "mask_status": "available" }, { "frame": "01048", "timestamp_sec": 96.7, "offset_sec": 42, "rgb_path": "assets/cat_07/v_00Dk03Jr70M_1368_2505/0042s_01048_rgb.jpg", "mask_path": "assets/cat_07/v_00Dk03Jr70M_1368_2505/0042s_01048_mask.png", "mask_status": "available" }, { "frame": "01076", "timestamp_sec": 97.7, "offset_sec": 43, "rgb_path": "assets/cat_07/v_00Dk03Jr70M_1368_2505/0043s_01076_rgb.jpg", "mask_path": "assets/cat_07/v_00Dk03Jr70M_1368_2505/0043s_01076_mask.png", "mask_status": "available" }, { "frame": "01100", "timestamp_sec": 98.7, "offset_sec": 44, "rgb_path": "assets/cat_07/v_00Dk03Jr70M_1368_2505/0044s_01100_rgb.jpg", "mask_path": "assets/cat_07/v_00Dk03Jr70M_1368_2505/0044s_01100_mask.png", "mask_status": "available" }, { "frame": "01124", "timestamp_sec": 99.7, "offset_sec": 45, "rgb_path": "assets/cat_07/v_00Dk03Jr70M_1368_2505/0045s_01124_rgb.jpg", "mask_path": "assets/cat_07/v_00Dk03Jr70M_1368_2505/0045s_01124_mask.png", "mask_status": "available" } ], "obj_id": [ 0 ], "anno_id": [ 155 ], "fps_inferred_from_clip_interval": 24.989 }, "usage": { "input_tokens": 50983, "input_tokens_details": { "cache_write_tokens": 50980, "cached_tokens": 0 }, "output_tokens": 6177, "output_tokens_details": { "reasoning_tokens": 3939 }, "total_tokens": 57160 }, "model": "gpt-5.6-sol", "effort": "medium", "custom_id": "cat07_cap04_v_00Dk03Jr70M_1368_2505", "annotation": { "caption": "He turns back to the wall and continues plastering, moving from the corner outwards. He applies the material in long vertical strips, using the trowel to smooth the surface with broad upward and downward motions.", "entity_groups": [ { "id": "entity_person_1", "label": "person", "spans": [ { "text": "He", "range": [ 0, 2 ] }, { "text": "He", "range": [ 85, 87 ] } ], "source_mask_link": "selected_person" }, { "id": "entity_wall_1", "label": "wall", "spans": [ { "text": "the wall", "range": [ 17, 25 ] }, { "text": "the surface", "range": [ 161, 172 ] } ], "source_mask_link": null }, { "id": "entity_material_1", "label": "plastering material", "spans": [ { "text": "the material", "range": [ 96, 108 ] } ], "source_mask_link": null }, { "id": "entity_strips_1", "label": "material strips", "spans": [ { "text": "long vertical strips", "range": [ 112, 132 ] } ], "source_mask_link": null }, { "id": "entity_trowel_1", "label": "trowel", "spans": [ { "text": "the trowel", "range": [ 140, 150 ] } ], "source_mask_link": null } ], "actions": [ { "id": "action_turn_1", "mentions": [ { "text": "turns", "ranges": [ [ 3, 8 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "type": "simple", "subject_refs": [ "entity_person_1" ], "object_refs": [ "entity_wall_1" ], "confidence": "high", "temporal_status": "localized", "intervals": [ { "start_s": 0.0, "end_s": 2.0, "source_start_s": 54.7, "source_end_s": 56.7 } ], "evidence": "Between samples 0 and 1, the masked person changes from leaning toward the work materials to orienting toward the wall.", "mask_relation": "The action subject matches the supplied selected-person mask.", "source_subject_refs": "entity_person_1" }, { "id": "action_plaster_1", "mentions": [ { "text": "plastering", "ranges": [ [ 40, 50 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "type": "simple", "subject_refs": [ "entity_person_1" ], "object_refs": [ "entity_wall_1", "entity_material_1" ], "confidence": "high", "temporal_status": "localized", "intervals": [ { "start_s": 2.0, "end_s": 45.5, "source_start_s": 56.7, "source_end_s": 100.2 } ], "evidence": "From sample 2 through the clip boundary, the masked person repeatedly works plaster onto the wall, with brief material-loading motions remaining part of the ongoing plastering activity.", "mask_relation": "The action subject matches the supplied selected-person mask.", "source_subject_refs": "entity_person_1" }, { "id": "action_move_1", "mentions": [ { "text": "moving", "ranges": [ [ 52, 58 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "type": "simple", "subject_refs": [ "entity_person_1" ], "object_refs": [], "confidence": "high", "temporal_status": "localized", "intervals": [ { "start_s": 2.0, "end_s": 45.5, "source_start_s": 56.7, "source_end_s": 100.2 } ], "evidence": "Across the sampled sequence, the masked person progresses from the left corner toward more central portions of the wall while working.", "mask_relation": "The action subject matches the supplied selected-person mask.", "source_subject_refs": "entity_person_1" }, { "id": "action_apply_1", "mentions": [ { "text": "applies", "ranges": [ [ 88, 95 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "type": "simple", "subject_refs": [ "entity_person_1" ], "object_refs": [ "entity_material_1", "entity_strips_1" ], "confidence": "high", "temporal_status": "localized", "intervals": [ { "start_s": 2.0, "end_s": 35.0, "source_start_s": 56.7, "source_end_s": 89.7 }, { "start_s": 37.0, "end_s": 45.5, "source_start_s": 91.7, "source_end_s": 100.2 } ], "evidence": "The clean frames show the masked person pressing and spreading white material onto the wall in successive vertical working areas; samples 35–36 primarily show reloading or repositioning material.", "mask_relation": "The action subject matches the supplied selected-person mask.", "source_subject_refs": "entity_person_1" }, { "id": "action_use_1", "mentions": [ { "text": "using", "ranges": [ [ 134, 139 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "type": "simple", "subject_refs": [ "entity_person_1" ], "object_refs": [ "entity_trowel_1" ], "confidence": "high", "temporal_status": "localized", "intervals": [ { "start_s": 2.0, "end_s": 35.0, "source_start_s": 56.7, "source_end_s": 89.7 }, { "start_s": 37.0, "end_s": 45.5, "source_start_s": 91.7, "source_end_s": 100.2 } ], "evidence": "In the supported frames, the masked person holds a trowel against the wall and manipulates it through repeated strokes.", "mask_relation": "The action subject matches the supplied selected-person mask.", "source_subject_refs": "entity_person_1" }, { "id": "action_smooth_1", "mentions": [ { "text": "smooth", "ranges": [ [ 154, 160 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "type": "simple", "subject_refs": [ "entity_person_1" ], "object_refs": [ "entity_wall_1" ], "confidence": "high", "temporal_status": "localized", "intervals": [ { "start_s": 2.0, "end_s": 35.0, "source_start_s": 56.7, "source_end_s": 89.7 }, { "start_s": 37.0, "end_s": 45.5, "source_start_s": 91.7, "source_end_s": 100.2 } ], "evidence": "The masked person repeatedly draws the trowel upward and downward across the plastered wall surface, visibly flattening and smoothing it.", "mask_relation": "The action subject matches the supplied selected-person mask.", "source_subject_refs": "entity_person_1" } ], "notes": [ "The supplied masklet tracks only the person referred to by both occurrences of “He”; all captioned actions therefore have a mask-linked subject.", "No separate masks were supplied for the wall, plastering material, material strips, or trowel, so their source-mask links are null.", "“the surface” is grouped with “the wall” because it unambiguously refers to that wall's surface.", "“the corner” is excluded as a spatial region, and “motions” is excluded as an abstract action description rather than a concrete entity.", "Action intervals are sparse visual-support intervals derived from the sampled clean RGB frames; intervals reaching 45.5 seconds end exactly at the original clip boundary." ] }, "model_output_raw": { "caption": "He turns back to the wall and continues plastering, moving from the corner outwards. He applies the material in long vertical strips, using the trowel to smooth the surface with broad upward and downward motions.", "entity_groups": [ { "id": "entity_person_1", "label": "person", "spans": [ { "text": "He", "range": [ 0, 2 ] }, { "text": "He", "range": [ 85, 87 ] } ], "source_mask_link": { "track_id": "selected_person_mask", "description": "The supplied masklet identifies this person in all sampled frames." } }, { "id": "entity_wall_1", "label": "wall", "spans": [ { "text": "the wall", "range": [ 17, 25 ] }, { "text": "the surface", "range": [ 161, 172 ] } ], "source_mask_link": null }, { "id": "entity_material_1", "label": "plastering material", "spans": [ { "text": "the material", "range": [ 96, 108 ] } ], "source_mask_link": null }, { "id": "entity_strips_1", "label": "material strips", "spans": [ { "text": "long vertical strips", "range": [ 112, 132 ] } ], "source_mask_link": null }, { "id": "entity_trowel_1", "label": "trowel", "spans": [ { "text": "the trowel", "range": [ 140, 150 ] } ], "source_mask_link": null } ], "actions": [ { "id": "action_turn_1", "mentions": [ { "text": "turns", "ranges": [ [ 3, 8 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "type": "simple", "subject_entity_group_id": "entity_person_1", "object_entity_group_ids": [ "entity_wall_1" ], "confidence": "high", "temporal_status": "localized", "clip_relative_intervals": [ { "start_s": 0.0, "end_s": 2.0 } ], "source_video_intervals": [ { "start_s": 54.7, "end_s": 56.7 } ], "evidence": "Between samples 0 and 1, the masked person changes from leaning toward the work materials to orienting toward the wall.", "mask_relation": "The action subject matches the supplied selected-person mask." }, { "id": "action_plaster_1", "mentions": [ { "text": "plastering", "ranges": [ [ 40, 50 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "type": "simple", "subject_entity_group_id": "entity_person_1", "object_entity_group_ids": [ "entity_wall_1", "entity_material_1" ], "confidence": "high", "temporal_status": "localized", "clip_relative_intervals": [ { "start_s": 2.0, "end_s": 45.5 } ], "source_video_intervals": [ { "start_s": 56.7, "end_s": 100.2 } ], "evidence": "From sample 2 through the clip boundary, the masked person repeatedly works plaster onto the wall, with brief material-loading motions remaining part of the ongoing plastering activity.", "mask_relation": "The action subject matches the supplied selected-person mask." }, { "id": "action_move_1", "mentions": [ { "text": "moving", "ranges": [ [ 52, 58 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "type": "simple", "subject_entity_group_id": "entity_person_1", "object_entity_group_ids": [], "confidence": "high", "temporal_status": "localized", "clip_relative_intervals": [ { "start_s": 2.0, "end_s": 45.5 } ], "source_video_intervals": [ { "start_s": 56.7, "end_s": 100.2 } ], "evidence": "Across the sampled sequence, the masked person progresses from the left corner toward more central portions of the wall while working.", "mask_relation": "The action subject matches the supplied selected-person mask." }, { "id": "action_apply_1", "mentions": [ { "text": "applies", "ranges": [ [ 88, 95 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "type": "simple", "subject_entity_group_id": "entity_person_1", "object_entity_group_ids": [ "entity_material_1", "entity_strips_1" ], "confidence": "high", "temporal_status": "localized", "clip_relative_intervals": [ { "start_s": 2.0, "end_s": 35.0 }, { "start_s": 37.0, "end_s": 45.5 } ], "source_video_intervals": [ { "start_s": 56.7, "end_s": 89.7 }, { "start_s": 91.7, "end_s": 100.2 } ], "evidence": "The clean frames show the masked person pressing and spreading white material onto the wall in successive vertical working areas; samples 35–36 primarily show reloading or repositioning material.", "mask_relation": "The action subject matches the supplied selected-person mask." }, { "id": "action_use_1", "mentions": [ { "text": "using", "ranges": [ [ 134, 139 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "type": "simple", "subject_entity_group_id": "entity_person_1", "object_entity_group_ids": [ "entity_trowel_1" ], "confidence": "high", "temporal_status": "localized", "clip_relative_intervals": [ { "start_s": 2.0, "end_s": 35.0 }, { "start_s": 37.0, "end_s": 45.5 } ], "source_video_intervals": [ { "start_s": 56.7, "end_s": 89.7 }, { "start_s": 91.7, "end_s": 100.2 } ], "evidence": "In the supported frames, the masked person holds a trowel against the wall and manipulates it through repeated strokes.", "mask_relation": "The action subject matches the supplied selected-person mask." }, { "id": "action_smooth_1", "mentions": [ { "text": "smooth", "ranges": [ [ 154, 160 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "type": "simple", "subject_entity_group_id": "entity_person_1", "object_entity_group_ids": [ "entity_wall_1" ], "confidence": "high", "temporal_status": "localized", "clip_relative_intervals": [ { "start_s": 2.0, "end_s": 35.0 }, { "start_s": 37.0, "end_s": 45.5 } ], "source_video_intervals": [ { "start_s": 56.7, "end_s": 89.7 }, { "start_s": 91.7, "end_s": 100.2 } ], "evidence": "The masked person repeatedly draws the trowel upward and downward across the plastered wall surface, visibly flattening and smoothing it.", "mask_relation": "The action subject matches the supplied selected-person mask." } ], "notes": [ "The supplied masklet tracks only the person referred to by both occurrences of “He”; all captioned actions therefore have a mask-linked subject.", "No separate masks were supplied for the wall, plastering material, material strips, or trowel, so their source-mask links are null.", "“the surface” is grouped with “the wall” because it unambiguously refers to that wall's surface.", "“the corner” is excluded as a spatial region, and “motions” is excluded as an abstract action description rather than a concrete entity.", "Action intervals are sparse visual-support intervals derived from the sampled clean RGB frames; intervals reaching 45.5 seconds end exactly at the original clip boundary." ] }, "normalization_repairs": 0 }, { "source_video_id": "v_00Dk03Jr70M", "caption_count_category": 7, "clip": { "caption_index": 5, "caption": "He turns away from the wall and moves back to the window area. He bends down once more to scoop a fresh batch of plaster onto his hawk.", "source_interval_sec": [ 100.2, 113.2 ], "loom_clip_id": "v_00Dk03Jr70M_2505_2830", "clip_frame_interval_4fps": [ 2505, 2830 ], "sample_count_1fps": 13, "sampled_frames": [ { "frame": "00000", "timestamp_sec": 100.2, "offset_sec": 0, "rgb_path": "assets/cat_07/v_00Dk03Jr70M_2505_2830/0000s_00000_rgb.jpg", "mask_path": "assets/cat_07/v_00Dk03Jr70M_2505_2830/0000s_00000_mask.png", "mask_status": "available" }, { "frame": "00024", "timestamp_sec": 101.2, "offset_sec": 1, "rgb_path": "assets/cat_07/v_00Dk03Jr70M_2505_2830/0001s_00024_rgb.jpg", "mask_path": "assets/cat_07/v_00Dk03Jr70M_2505_2830/0001s_00024_mask.png", "mask_status": "available" }, { "frame": "00048", "timestamp_sec": 102.2, "offset_sec": 2, "rgb_path": "assets/cat_07/v_00Dk03Jr70M_2505_2830/0002s_00048_rgb.jpg", "mask_path": "assets/cat_07/v_00Dk03Jr70M_2505_2830/0002s_00048_mask.png", "mask_status": "available" }, { "frame": "00076", "timestamp_sec": 103.2, "offset_sec": 3, "rgb_path": "assets/cat_07/v_00Dk03Jr70M_2505_2830/0003s_00076_rgb.jpg", "mask_path": "assets/cat_07/v_00Dk03Jr70M_2505_2830/0003s_00076_mask.png", "mask_status": "available" }, { "frame": "00100", "timestamp_sec": 104.2, "offset_sec": 4, "rgb_path": "assets/cat_07/v_00Dk03Jr70M_2505_2830/0004s_00100_rgb.jpg", "mask_path": "assets/cat_07/v_00Dk03Jr70M_2505_2830/0004s_00100_mask.png", "mask_status": "available" }, { "frame": "00124", "timestamp_sec": 105.2, "offset_sec": 5, "rgb_path": "assets/cat_07/v_00Dk03Jr70M_2505_2830/0005s_00124_rgb.jpg", "mask_path": "assets/cat_07/v_00Dk03Jr70M_2505_2830/0005s_00124_mask.png", "mask_status": "available" }, { "frame": "00148", "timestamp_sec": 106.2, "offset_sec": 6, "rgb_path": "assets/cat_07/v_00Dk03Jr70M_2505_2830/0006s_00148_rgb.jpg", "mask_path": "assets/cat_07/v_00Dk03Jr70M_2505_2830/0006s_00148_mask.png", "mask_status": "available" }, { "frame": "00176", "timestamp_sec": 107.2, "offset_sec": 7, "rgb_path": "assets/cat_07/v_00Dk03Jr70M_2505_2830/0007s_00176_rgb.jpg", "mask_path": "assets/cat_07/v_00Dk03Jr70M_2505_2830/0007s_00176_mask.png", "mask_status": "available" }, { "frame": "00200", "timestamp_sec": 108.2, "offset_sec": 8, "rgb_path": "assets/cat_07/v_00Dk03Jr70M_2505_2830/0008s_00200_rgb.jpg", "mask_path": "assets/cat_07/v_00Dk03Jr70M_2505_2830/0008s_00200_mask.png", "mask_status": "available" }, { "frame": "00224", "timestamp_sec": 109.2, "offset_sec": 9, "rgb_path": "assets/cat_07/v_00Dk03Jr70M_2505_2830/0009s_00224_rgb.jpg", "mask_path": "assets/cat_07/v_00Dk03Jr70M_2505_2830/0009s_00224_mask.png", "mask_status": "available" }, { "frame": "00248", "timestamp_sec": 110.2, "offset_sec": 10, "rgb_path": "assets/cat_07/v_00Dk03Jr70M_2505_2830/0010s_00248_rgb.jpg", "mask_path": "assets/cat_07/v_00Dk03Jr70M_2505_2830/0010s_00248_mask.png", "mask_status": "available" }, { "frame": "00276", "timestamp_sec": 111.2, "offset_sec": 11, "rgb_path": "assets/cat_07/v_00Dk03Jr70M_2505_2830/0011s_00276_rgb.jpg", "mask_path": "assets/cat_07/v_00Dk03Jr70M_2505_2830/0011s_00276_mask.png", "mask_status": "available" }, { "frame": "00300", "timestamp_sec": 112.2, "offset_sec": 12, "rgb_path": "assets/cat_07/v_00Dk03Jr70M_2505_2830/0012s_00300_rgb.jpg", "mask_path": "assets/cat_07/v_00Dk03Jr70M_2505_2830/0012s_00300_mask.png", "mask_status": "available" } ], "obj_id": [ 0 ], "anno_id": [ 156 ], "fps_inferred_from_clip_interval": 25.0 }, "usage": { "input_tokens": 25396, "input_tokens_details": { "cache_write_tokens": 25393, "cached_tokens": 0 }, "output_tokens": 3874, "output_tokens_details": { "reasoning_tokens": 2458 }, "total_tokens": 29270 }, "model": "gpt-5.6-sol", "effort": "medium", "custom_id": "cat07_cap05_v_00Dk03Jr70M_2505_2830", "annotation": { "caption": "He turns away from the wall and moves back to the window area. He bends down once more to scoop a fresh batch of plaster onto his hawk.", "entity_groups": [ { "id": "entity_1", "label": "person", "spans": [ { "text": "He", "range": [ 0, 2 ] }, { "text": "He", "range": [ 63, 65 ] }, { "text": "his", "range": [ 126, 129 ] } ], "source_mask_link": "selected_person" }, { "id": "entity_2", "label": "wall", "spans": [ { "text": "the wall", "range": [ 19, 27 ] } ], "source_mask_link": null }, { "id": "entity_3", "label": "window", "spans": [ { "text": "window", "range": [ 50, 56 ] } ], "source_mask_link": null }, { "id": "entity_4", "label": "plaster", "spans": [ { "text": "a fresh batch of plaster", "range": [ 96, 120 ] } ], "source_mask_link": null }, { "id": "entity_5", "label": "hawk", "spans": [ { "text": "his hawk", "range": [ 126, 134 ] } ], "source_mask_link": null } ], "actions": [ { "id": "action_1", "mentions": [ { "text": "turns", "ranges": [ [ 3, 8 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "type": "simple", "subject_refs": [ "entity_1" ], "object_refs": [ "entity_2" ], "confidence": "high", "temporal_status": "localized", "intervals": [ { "start_s": 1.0, "end_s": 3.0, "source_start_s": 101.2, "source_end_s": 103.2 } ], "evidence": "At 1.0 s the masked person is oriented toward the wall; by 2.0 s he has rotated away from it.", "mask_relation": "subject matches the supplied selected-person mask", "source_subject_refs": "entity_1" }, { "id": "action_2", "mentions": [ { "text": "moves", "ranges": [ [ 32, 37 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "type": "simple", "subject_refs": [ "entity_1" ], "object_refs": [ "entity_3" ], "confidence": "high", "temporal_status": "localized", "intervals": [ { "start_s": 2.0, "end_s": 7.0, "source_start_s": 102.2, "source_end_s": 107.2 } ], "evidence": "Across 2.0–6.0 s, the masked person shifts away from the wall and leftward toward the window-side work area.", "mask_relation": "subject matches the supplied selected-person mask", "source_subject_refs": "entity_1" }, { "id": "action_3", "mentions": [ { "text": "bends", "ranges": [ [ 66, 71 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "type": "simple", "subject_refs": [ "entity_1" ], "object_refs": [], "confidence": "high", "temporal_status": "localized", "intervals": [ { "start_s": 7.0, "end_s": 12.0, "source_start_s": 107.2, "source_end_s": 112.2 } ], "evidence": "The masked person leans and bends over the work surface in the clean frames from 7.0 through 11.0 s.", "mask_relation": "subject matches the supplied selected-person mask", "source_subject_refs": "entity_1" }, { "id": "action_4", "mentions": [ { "text": "scoop", "ranges": [ [ 90, 95 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "type": "simple", "subject_refs": [ "entity_1" ], "object_refs": [ "entity_4", "entity_5" ], "confidence": "high", "temporal_status": "localized", "intervals": [ { "start_s": 9.0, "end_s": 13.0, "source_start_s": 109.2, "source_end_s": 113.2 } ], "evidence": "From 9.0 through 12.0 s, the masked person visibly manipulates a trowel at the material surface and transfers material toward the hawk; support reaches the clip boundary.", "mask_relation": "subject matches the supplied selected-person mask", "source_subject_refs": "entity_1" } ], "notes": [ "The supplied mask consistently identifies the sole person referred to by both instances of “He” and by “his.”", "No separate masks were supplied for the wall, window, plaster, or hawk.", "The abstract spatial head “area” is excluded; the concrete “window” is retained as an entity.", "The plaster mention uses the complete counter-plus-content span “a fresh batch of plaster.”", "Directional modifiers “away,” “back,” and “down” are excluded from the action spans." ] }, "model_output_raw": { "caption": "He turns away from the wall and moves back to the window area. He bends down once more to scoop a fresh batch of plaster onto his hawk.", "entity_groups": [ { "id": "entity_1", "label": "person", "spans": [ { "text": "He", "range": [ 0, 2 ] }, { "text": "He", "range": [ 63, 65 ] }, { "text": "his", "range": [ 126, 129 ] } ], "source_mask": "selected_person_masklet" }, { "id": "entity_2", "label": "wall", "spans": [ { "text": "the wall", "range": [ 19, 27 ] } ], "source_mask": null }, { "id": "entity_3", "label": "window", "spans": [ { "text": "window", "range": [ 50, 56 ] } ], "source_mask": null }, { "id": "entity_4", "label": "plaster", "spans": [ { "text": "a fresh batch of plaster", "range": [ 96, 120 ] } ], "source_mask": null }, { "id": "entity_5", "label": "hawk", "spans": [ { "text": "his hawk", "range": [ 126, 134 ] } ], "source_mask": null } ], "actions": [ { "id": "action_1", "mentions": [ { "text": "turns", "ranges": [ [ 3, 8 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "type": "simple", "subject_ref": "entity_1", "object_refs": [ "entity_2" ], "confidence": "high", "mask_relation": "subject matches the supplied selected-person mask", "temporal_status": "localized", "intervals": [ { "start_s": 1.0, "end_s": 3.0, "source_start_s": 101.2, "source_end_s": 103.2 } ], "evidence": "At 1.0 s the masked person is oriented toward the wall; by 2.0 s he has rotated away from it." }, { "id": "action_2", "mentions": [ { "text": "moves", "ranges": [ [ 32, 37 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "type": "simple", "subject_ref": "entity_1", "object_refs": [ "entity_3" ], "confidence": "high", "mask_relation": "subject matches the supplied selected-person mask", "temporal_status": "localized", "intervals": [ { "start_s": 2.0, "end_s": 7.0, "source_start_s": 102.2, "source_end_s": 107.2 } ], "evidence": "Across 2.0–6.0 s, the masked person shifts away from the wall and leftward toward the window-side work area." }, { "id": "action_3", "mentions": [ { "text": "bends", "ranges": [ [ 66, 71 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "type": "simple", "subject_ref": "entity_1", "object_refs": [], "confidence": "high", "mask_relation": "subject matches the supplied selected-person mask", "temporal_status": "localized", "intervals": [ { "start_s": 7.0, "end_s": 12.0, "source_start_s": 107.2, "source_end_s": 112.2 } ], "evidence": "The masked person leans and bends over the work surface in the clean frames from 7.0 through 11.0 s." }, { "id": "action_4", "mentions": [ { "text": "scoop", "ranges": [ [ 90, 95 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "type": "simple", "subject_ref": "entity_1", "object_refs": [ "entity_4", "entity_5" ], "confidence": "high", "mask_relation": "subject matches the supplied selected-person mask", "temporal_status": "localized", "intervals": [ { "start_s": 9.0, "end_s": 13.0, "source_start_s": 109.2, "source_end_s": 113.2 } ], "evidence": "From 9.0 through 12.0 s, the masked person visibly manipulates a trowel at the material surface and transfers material toward the hawk; support reaches the clip boundary." } ], "notes": [ "The supplied mask consistently identifies the sole person referred to by both instances of “He” and by “his.”", "No separate masks were supplied for the wall, window, plaster, or hawk.", "The abstract spatial head “area” is excluded; the concrete “window” is retained as an entity.", "The plaster mention uses the complete counter-plus-content span “a fresh batch of plaster.”", "Directional modifiers “away,” “back,” and “down” are excluded from the action spans." ] }, "normalization_repairs": 0 }, { "source_video_id": "v_00Dk03Jr70M", "caption_count_category": 7, "clip": { "caption_index": 6, "caption": "Facing the wall again, he applies another large section of plaster next to his previous work. He uses long, sweeping strokes to spread the material and carefully smooths the surface with his trowel.", "source_interval_sec": [ 113.2, 139.2 ], "loom_clip_id": "v_00Dk03Jr70M_2830_3480", "clip_frame_interval_4fps": [ 2830, 3480 ], "sample_count_1fps": 26, "sampled_frames": [ { "frame": "00000", "timestamp_sec": 113.2, "offset_sec": 0, "rgb_path": "assets/cat_07/v_00Dk03Jr70M_2830_3480/0000s_00000_rgb.jpg", "mask_path": "assets/cat_07/v_00Dk03Jr70M_2830_3480/0000s_00000_mask.png", "mask_status": "available" }, { "frame": "00024", "timestamp_sec": 114.2, "offset_sec": 1, "rgb_path": "assets/cat_07/v_00Dk03Jr70M_2830_3480/0001s_00024_rgb.jpg", "mask_path": "assets/cat_07/v_00Dk03Jr70M_2830_3480/0001s_00024_mask.png", "mask_status": "available" }, { "frame": "00052", "timestamp_sec": 115.2, "offset_sec": 2, "rgb_path": "assets/cat_07/v_00Dk03Jr70M_2830_3480/0002s_00052_rgb.jpg", "mask_path": "assets/cat_07/v_00Dk03Jr70M_2830_3480/0002s_00052_mask.png", "mask_status": "available" }, { "frame": "00076", "timestamp_sec": 116.2, "offset_sec": 3, "rgb_path": "assets/cat_07/v_00Dk03Jr70M_2830_3480/0003s_00076_rgb.jpg", "mask_path": "assets/cat_07/v_00Dk03Jr70M_2830_3480/0003s_00076_mask.png", "mask_status": "available" }, { "frame": "00100", "timestamp_sec": 117.2, "offset_sec": 4, "rgb_path": "assets/cat_07/v_00Dk03Jr70M_2830_3480/0004s_00100_rgb.jpg", "mask_path": "assets/cat_07/v_00Dk03Jr70M_2830_3480/0004s_00100_mask.png", "mask_status": "available" }, { "frame": "00124", "timestamp_sec": 118.2, "offset_sec": 5, "rgb_path": "assets/cat_07/v_00Dk03Jr70M_2830_3480/0005s_00124_rgb.jpg", "mask_path": "assets/cat_07/v_00Dk03Jr70M_2830_3480/0005s_00124_mask.png", "mask_status": "available" }, { "frame": "00152", "timestamp_sec": 119.2, "offset_sec": 6, "rgb_path": "assets/cat_07/v_00Dk03Jr70M_2830_3480/0006s_00152_rgb.jpg", "mask_path": "assets/cat_07/v_00Dk03Jr70M_2830_3480/0006s_00152_mask.png", "mask_status": "available" }, { "frame": "00176", "timestamp_sec": 120.2, "offset_sec": 7, "rgb_path": "assets/cat_07/v_00Dk03Jr70M_2830_3480/0007s_00176_rgb.jpg", "mask_path": "assets/cat_07/v_00Dk03Jr70M_2830_3480/0007s_00176_mask.png", "mask_status": "available" }, { "frame": "00200", "timestamp_sec": 121.2, "offset_sec": 8, "rgb_path": "assets/cat_07/v_00Dk03Jr70M_2830_3480/0008s_00200_rgb.jpg", "mask_path": "assets/cat_07/v_00Dk03Jr70M_2830_3480/0008s_00200_mask.png", "mask_status": "available" }, { "frame": "00224", "timestamp_sec": 122.2, "offset_sec": 9, "rgb_path": "assets/cat_07/v_00Dk03Jr70M_2830_3480/0009s_00224_rgb.jpg", "mask_path": "assets/cat_07/v_00Dk03Jr70M_2830_3480/0009s_00224_mask.png", "mask_status": "available" }, { "frame": "00252", "timestamp_sec": 123.2, "offset_sec": 10, "rgb_path": "assets/cat_07/v_00Dk03Jr70M_2830_3480/0010s_00252_rgb.jpg", "mask_path": "assets/cat_07/v_00Dk03Jr70M_2830_3480/0010s_00252_mask.png", "mask_status": "available" }, { "frame": "00276", "timestamp_sec": 124.2, "offset_sec": 11, "rgb_path": "assets/cat_07/v_00Dk03Jr70M_2830_3480/0011s_00276_rgb.jpg", "mask_path": "assets/cat_07/v_00Dk03Jr70M_2830_3480/0011s_00276_mask.png", "mask_status": "available" }, { "frame": "00300", "timestamp_sec": 125.2, "offset_sec": 12, "rgb_path": "assets/cat_07/v_00Dk03Jr70M_2830_3480/0012s_00300_rgb.jpg", "mask_path": "assets/cat_07/v_00Dk03Jr70M_2830_3480/0012s_00300_mask.png", "mask_status": "available" }, { "frame": "00324", "timestamp_sec": 126.2, "offset_sec": 13, "rgb_path": "assets/cat_07/v_00Dk03Jr70M_2830_3480/0013s_00324_rgb.jpg", "mask_path": "assets/cat_07/v_00Dk03Jr70M_2830_3480/0013s_00324_mask.png", "mask_status": "available" }, { "frame": "00352", "timestamp_sec": 127.2, "offset_sec": 14, "rgb_path": "assets/cat_07/v_00Dk03Jr70M_2830_3480/0014s_00352_rgb.jpg", "mask_path": "assets/cat_07/v_00Dk03Jr70M_2830_3480/0014s_00352_mask.png", "mask_status": "available" }, { "frame": "00376", "timestamp_sec": 128.2, "offset_sec": 15, "rgb_path": "assets/cat_07/v_00Dk03Jr70M_2830_3480/0015s_00376_rgb.jpg", "mask_path": "assets/cat_07/v_00Dk03Jr70M_2830_3480/0015s_00376_mask.png", "mask_status": "available" }, { "frame": "00400", "timestamp_sec": 129.2, "offset_sec": 16, "rgb_path": "assets/cat_07/v_00Dk03Jr70M_2830_3480/0016s_00400_rgb.jpg", "mask_path": "assets/cat_07/v_00Dk03Jr70M_2830_3480/0016s_00400_mask.png", "mask_status": "available" }, { "frame": "00424", "timestamp_sec": 130.2, "offset_sec": 17, "rgb_path": "assets/cat_07/v_00Dk03Jr70M_2830_3480/0017s_00424_rgb.jpg", "mask_path": "assets/cat_07/v_00Dk03Jr70M_2830_3480/0017s_00424_mask.png", "mask_status": "available" }, { "frame": "00452", "timestamp_sec": 131.2, "offset_sec": 18, "rgb_path": "assets/cat_07/v_00Dk03Jr70M_2830_3480/0018s_00452_rgb.jpg", "mask_path": "assets/cat_07/v_00Dk03Jr70M_2830_3480/0018s_00452_mask.png", "mask_status": "available" }, { "frame": "00476", "timestamp_sec": 132.2, "offset_sec": 19, "rgb_path": "assets/cat_07/v_00Dk03Jr70M_2830_3480/0019s_00476_rgb.jpg", "mask_path": "assets/cat_07/v_00Dk03Jr70M_2830_3480/0019s_00476_mask.png", "mask_status": "available" }, { "frame": "00500", "timestamp_sec": 133.2, "offset_sec": 20, "rgb_path": "assets/cat_07/v_00Dk03Jr70M_2830_3480/0020s_00500_rgb.jpg", "mask_path": "assets/cat_07/v_00Dk03Jr70M_2830_3480/0020s_00500_mask.png", "mask_status": "available" }, { "frame": "00524", "timestamp_sec": 134.2, "offset_sec": 21, "rgb_path": "assets/cat_07/v_00Dk03Jr70M_2830_3480/0021s_00524_rgb.jpg", "mask_path": "assets/cat_07/v_00Dk03Jr70M_2830_3480/0021s_00524_mask.png", "mask_status": "available" }, { "frame": "00552", "timestamp_sec": 135.2, "offset_sec": 22, "rgb_path": "assets/cat_07/v_00Dk03Jr70M_2830_3480/0022s_00552_rgb.jpg", "mask_path": "assets/cat_07/v_00Dk03Jr70M_2830_3480/0022s_00552_mask.png", "mask_status": "available" }, { "frame": "00576", "timestamp_sec": 136.2, "offset_sec": 23, "rgb_path": "assets/cat_07/v_00Dk03Jr70M_2830_3480/0023s_00576_rgb.jpg", "mask_path": "assets/cat_07/v_00Dk03Jr70M_2830_3480/0023s_00576_mask.png", "mask_status": "available" }, { "frame": "00600", "timestamp_sec": 137.2, "offset_sec": 24, "rgb_path": "assets/cat_07/v_00Dk03Jr70M_2830_3480/0024s_00600_rgb.jpg", "mask_path": "assets/cat_07/v_00Dk03Jr70M_2830_3480/0024s_00600_mask.png", "mask_status": "available" }, { "frame": "00624", "timestamp_sec": 138.2, "offset_sec": 25, "rgb_path": "assets/cat_07/v_00Dk03Jr70M_2830_3480/0025s_00624_rgb.jpg", "mask_path": "assets/cat_07/v_00Dk03Jr70M_2830_3480/0025s_00624_mask.png", "mask_status": "available" } ], "obj_id": [ 0 ], "anno_id": [ 157 ], "fps_inferred_from_clip_interval": 25.0 }, "usage": { "input_tokens": 35480, "input_tokens_details": { "cache_write_tokens": 35477, "cached_tokens": 0 }, "output_tokens": 4032, "output_tokens_details": { "reasoning_tokens": 2588 }, "total_tokens": 39512 }, "model": "gpt-5.6-sol", "effort": "medium", "custom_id": "cat07_cap06_v_00Dk03Jr70M_2830_3480", "annotation": { "caption": "Facing the wall again, he applies another large section of plaster next to his previous work. He uses long, sweeping strokes to spread the material and carefully smooths the surface with his trowel.", "entity_groups": [ { "id": "entity_1", "label": "person", "spans": [ { "text": "he", "range": [ 23, 25 ] }, { "text": "his", "range": [ 75, 78 ] }, { "text": "He", "range": [ 94, 96 ] }, { "text": "his", "range": [ 187, 190 ] } ], "source_mask_link": "selected_person" }, { "id": "entity_2", "label": "wall", "spans": [ { "text": "the wall", "range": [ 7, 15 ] }, { "text": "the surface", "range": [ 170, 181 ] } ], "source_mask_link": null }, { "id": "entity_3", "label": "plaster", "spans": [ { "text": "another large section of plaster", "range": [ 34, 66 ] }, { "text": "the material", "range": [ 135, 147 ] } ], "source_mask_link": null }, { "id": "entity_4", "label": "trowel", "spans": [ { "text": "his trowel", "range": [ 187, 197 ] } ], "source_mask_link": null } ], "actions": [ { "id": "action_1", "mentions": [ { "text": "Facing", "ranges": [ [ 0, 6 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "type": "posture", "subject_refs": [ "entity_1" ], "object_refs": [ "entity_2" ], "confidence": "high", "temporal_status": "localized", "intervals": [ { "start_s": 1.0, "end_s": 25.999999999999986, "source_start_s": 114.2, "source_end_s": 139.2 } ], "evidence": "From 1.0 seconds onward, the masked person is oriented toward the wall while working, with the posture continuing through the final sample.", "mask_relation": "The action subject matches the supplied selected-person mask.", "source_subject_refs": "entity_1" }, { "id": "action_2", "mentions": [ { "text": "applies", "ranges": [ [ 26, 33 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } }, { "text": "spread", "ranges": [ [ 128, 134 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "type": "simple", "subject_refs": [ "entity_1" ], "object_refs": [ "entity_3" ], "confidence": "high", "temporal_status": "localized", "intervals": [ { "start_s": 4.0, "end_s": 21.0, "source_start_s": 117.2, "source_end_s": 134.2 } ], "evidence": "Samples from 4.0 through 20.0 seconds show the masked person repeatedly transferring and spreading plaster across the wall with a hand tool.", "mask_relation": "The action subject matches the supplied selected-person mask.", "source_subject_refs": "entity_1" }, { "id": "action_3", "mentions": [ { "text": "uses long, sweeping strokes", "ranges": [ [ 97, 124 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "type": "light_verb", "subject_refs": [ "entity_1" ], "object_refs": [], "confidence": "high", "temporal_status": "localized", "intervals": [ { "start_s": 4.0, "end_s": 21.0, "source_start_s": 117.2, "source_end_s": 134.2 } ], "evidence": "The masked person makes repeated long upward and downward arm strokes over the wall in samples from 4.0 through 20.0 seconds.", "mask_relation": "The action subject matches the supplied selected-person mask.", "source_subject_refs": "entity_1" }, { "id": "action_4", "mentions": [ { "text": "smooths", "ranges": [ [ 162, 169 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "type": "simple", "subject_refs": [ "entity_1" ], "object_refs": [ "entity_2" ], "confidence": "high", "temporal_status": "localized", "intervals": [ { "start_s": 21.0, "end_s": 25.999999999999986, "source_start_s": 134.2, "source_end_s": 139.2 } ], "evidence": "Samples from 21.0 through 25.0 seconds show the masked person pressing a broad trowel or float against the plastered surface and making smoothing passes; the action remains supported in the final sample.", "mask_relation": "The action subject matches the supplied selected-person mask.", "source_subject_refs": "entity_1" } ], "notes": [ "The supplied mask consistently identifies the sole visible worker and is linked only to the person entity group.", "No source masks were supplied for the wall, plaster, or trowel.", "\"the surface\" is grouped with \"the wall\" because it denotes the worked wall surface.", "\"the material\" corefers with the plaster.", "\"his previous work\" is not annotated as an entity because work is an abstract/result reference rather than an independently segmentable object." ] }, "model_output_raw": { "caption": "Facing the wall again, he applies another large section of plaster next to his previous work. He uses long, sweeping strokes to spread the material and carefully smooths the surface with his trowel.", "entity_groups": [ { "id": "entity_1", "label": "person", "spans": [ { "text": "he", "range": [ 23, 25 ] }, { "text": "his", "range": [ 75, 78 ] }, { "text": "He", "range": [ 94, 96 ] }, { "text": "his", "range": [ 187, 190 ] } ], "source_mask": "selected_person_mask" }, { "id": "entity_2", "label": "wall", "spans": [ { "text": "the wall", "range": [ 7, 15 ] }, { "text": "the surface", "range": [ 170, 181 ] } ], "source_mask": null }, { "id": "entity_3", "label": "plaster", "spans": [ { "text": "another large section of plaster", "range": [ 34, 66 ] }, { "text": "the material", "range": [ 135, 147 ] } ], "source_mask": null }, { "id": "entity_4", "label": "trowel", "spans": [ { "text": "his trowel", "range": [ 187, 197 ] } ], "source_mask": null } ], "actions": [ { "id": "action_1", "mentions": [ { "text": "Facing", "ranges": [ [ 0, 6 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "type": "posture", "subject": "entity_1", "objects": [ "entity_2" ], "confidence": "high", "temporal_status": "localized", "intervals": [ { "clip_relative": [ 1.0, 26.0 ], "source_video": [ 114.2, 139.2 ] } ], "evidence": "From 1.0 seconds onward, the masked person is oriented toward the wall while working, with the posture continuing through the final sample.", "mask_relation": "The action subject matches the supplied selected-person mask." }, { "id": "action_2", "mentions": [ { "text": "applies", "ranges": [ [ 26, 33 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } }, { "text": "spread", "ranges": [ [ 128, 134 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "type": "simple", "subject": "entity_1", "objects": [ "entity_3" ], "confidence": "high", "temporal_status": "localized", "intervals": [ { "clip_relative": [ 4.0, 21.0 ], "source_video": [ 117.2, 134.2 ] } ], "evidence": "Samples from 4.0 through 20.0 seconds show the masked person repeatedly transferring and spreading plaster across the wall with a hand tool.", "mask_relation": "The action subject matches the supplied selected-person mask." }, { "id": "action_3", "mentions": [ { "text": "uses long, sweeping strokes", "ranges": [ [ 97, 124 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "type": "light_verb", "subject": "entity_1", "objects": [], "confidence": "high", "temporal_status": "localized", "intervals": [ { "clip_relative": [ 4.0, 21.0 ], "source_video": [ 117.2, 134.2 ] } ], "evidence": "The masked person makes repeated long upward and downward arm strokes over the wall in samples from 4.0 through 20.0 seconds.", "mask_relation": "The action subject matches the supplied selected-person mask." }, { "id": "action_4", "mentions": [ { "text": "smooths", "ranges": [ [ 162, 169 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "type": "simple", "subject": "entity_1", "objects": [ "entity_2" ], "instrument": "entity_4", "confidence": "high", "temporal_status": "localized", "intervals": [ { "clip_relative": [ 21.0, 26.0 ], "source_video": [ 134.2, 139.2 ] } ], "evidence": "Samples from 21.0 through 25.0 seconds show the masked person pressing a broad trowel or float against the plastered surface and making smoothing passes; the action remains supported in the final sample.", "mask_relation": "The action subject matches the supplied selected-person mask." } ], "notes": [ "The supplied mask consistently identifies the sole visible worker and is linked only to the person entity group.", "No source masks were supplied for the wall, plaster, or trowel.", "\"the surface\" is grouped with \"the wall\" because it denotes the worked wall surface.", "\"the material\" corefers with the plaster.", "\"his previous work\" is not annotated as an entity because work is an abstract/result reference rather than an independently segmentable object." ] }, "normalization_repairs": 0 }, { "source_video_id": "v_00Dk03Jr70M", "caption_count_category": 7, "clip": { "caption_index": 7, "caption": "He makes a final smoothing motion, scrapes his trowel clean on the hawk, and then walks out of the frame to the left.", "source_interval_sec": [ 139.2, 148.8 ], "loom_clip_id": "v_00Dk03Jr70M_3480_3720", "clip_frame_interval_4fps": [ 3480, 3720 ], "sample_count_1fps": 10, "sampled_frames": [ { "frame": "00000", "timestamp_sec": 139.2, "offset_sec": 0, "rgb_path": "assets/cat_07/v_00Dk03Jr70M_3480_3720/0000s_00000_rgb.jpg", "mask_path": "assets/cat_07/v_00Dk03Jr70M_3480_3720/0000s_00000_mask.png", "mask_status": "available" }, { "frame": "00024", "timestamp_sec": 140.2, "offset_sec": 1, "rgb_path": "assets/cat_07/v_00Dk03Jr70M_3480_3720/0001s_00024_rgb.jpg", "mask_path": "assets/cat_07/v_00Dk03Jr70M_3480_3720/0001s_00024_mask.png", "mask_status": "available" }, { "frame": "00048", "timestamp_sec": 141.2, "offset_sec": 2, "rgb_path": "assets/cat_07/v_00Dk03Jr70M_3480_3720/0002s_00048_rgb.jpg", "mask_path": "assets/cat_07/v_00Dk03Jr70M_3480_3720/0002s_00048_mask.png", "mask_status": "available" }, { "frame": "00076", "timestamp_sec": 142.2, "offset_sec": 3, "rgb_path": "assets/cat_07/v_00Dk03Jr70M_3480_3720/0003s_00076_rgb.jpg", "mask_path": "assets/cat_07/v_00Dk03Jr70M_3480_3720/0003s_00076_mask.png", "mask_status": "available" }, { "frame": "00100", "timestamp_sec": 143.2, "offset_sec": 4, "rgb_path": "assets/cat_07/v_00Dk03Jr70M_3480_3720/0004s_00100_rgb.jpg", "mask_path": "assets/cat_07/v_00Dk03Jr70M_3480_3720/0004s_00100_mask.png", "mask_status": "available" }, { "frame": "00124", "timestamp_sec": 144.2, "offset_sec": 5, "rgb_path": "assets/cat_07/v_00Dk03Jr70M_3480_3720/0005s_00124_rgb.jpg", "mask_path": "assets/cat_07/v_00Dk03Jr70M_3480_3720/0005s_00124_mask.png", "mask_status": "available" }, { "frame": "00148", "timestamp_sec": 145.2, "offset_sec": 6, "rgb_path": "assets/cat_07/v_00Dk03Jr70M_3480_3720/0006s_00148_rgb.jpg", "mask_path": "assets/cat_07/v_00Dk03Jr70M_3480_3720/0006s_00148_mask.png", "mask_status": "available" }, { "frame": "00176", "timestamp_sec": 146.2, "offset_sec": 7, "rgb_path": "assets/cat_07/v_00Dk03Jr70M_3480_3720/0007s_00176_rgb.jpg", "mask_path": "assets/cat_07/v_00Dk03Jr70M_3480_3720/0007s_00176_mask.png", "mask_status": "available" }, { "frame": "00200", "timestamp_sec": 147.2, "offset_sec": 8, "rgb_path": "assets/cat_07/v_00Dk03Jr70M_3480_3720/0008s_00200_rgb.jpg", "mask_path": "assets/cat_07/v_00Dk03Jr70M_3480_3720/0008s_00200_mask.png", "mask_status": "available" }, { "frame": "00224", "timestamp_sec": 148.2, "offset_sec": 9, "rgb_path": "assets/cat_07/v_00Dk03Jr70M_3480_3720/0009s_00224_rgb.jpg", "mask_path": "assets/cat_07/v_00Dk03Jr70M_3480_3720/0009s_00224_mask.png", "mask_status": "available" } ], "obj_id": [ 0 ], "anno_id": [ 159 ], "fps_inferred_from_clip_interval": 25.0 }, "usage": { "input_tokens": 23070, "input_tokens_details": { "cache_write_tokens": 23067, "cached_tokens": 0 }, "output_tokens": 2557, "output_tokens_details": { "reasoning_tokens": 1603 }, "total_tokens": 25627 }, "model": "gpt-5.6-sol", "effort": "medium", "custom_id": "cat07_cap07_v_00Dk03Jr70M_3480_3720", "annotation": { "caption": "He makes a final smoothing motion, scrapes his trowel clean on the hawk, and then walks out of the frame to the left.", "entity_groups": [ { "id": "entity_1", "label": "person", "spans": [ { "text": "He", "range": [ 0, 2 ] }, { "text": "his", "range": [ 43, 46 ] } ], "source_mask_link": "selected_person" }, { "id": "entity_2", "label": "trowel", "spans": [ { "text": "his trowel", "range": [ 43, 53 ] } ], "source_mask_link": null }, { "id": "entity_3", "label": "hawk", "spans": [ { "text": "the hawk", "range": [ 63, 71 ] } ], "source_mask_link": null } ], "actions": [ { "id": "action_1", "mentions": [ { "text": "makes a final smoothing motion", "ranges": [ [ 3, 33 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "type": "light_verb", "subject_refs": [ "entity_1" ], "object_refs": [], "confidence": "high", "temporal_status": "localized", "intervals": [ { "start_s": 0.0, "end_s": 1.0, "source_start_s": 139.2, "source_end_s": 140.2 } ], "evidence": "At 0.000 s, the masked person presses and moves a trowel against the wall in a smoothing posture.", "mask_relation": "subject matches the supplied selected-person mask", "source_subject_refs": "entity_1" }, { "id": "action_2", "mentions": [ { "text": "scrapes", "ranges": [ [ 35, 42 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "type": "simple", "subject_refs": [ "entity_1" ], "object_refs": [ "entity_2", "entity_3" ], "confidence": "high", "temporal_status": "localized", "intervals": [ { "start_s": 1.0, "end_s": 3.0, "source_start_s": 140.2, "source_end_s": 142.2 }, { "start_s": 4.0, "end_s": 7.0, "source_start_s": 143.2, "source_end_s": 146.2 } ], "evidence": "At 1.000–2.000 s and 4.000–6.000 s, the masked person holds the hawk and repeatedly brings the trowel across or against it.", "mask_relation": "subject matches the supplied selected-person mask", "source_subject_refs": "entity_1" }, { "id": "action_3", "mentions": [ { "text": "walks out", "ranges": [ [ 82, 91 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "type": "phrasal", "subject_refs": [ "entity_1" ], "object_refs": [], "confidence": "high", "temporal_status": "localized", "intervals": [ { "start_s": 7.0, "end_s": 9.600000000000023, "source_start_s": 146.2, "source_end_s": 148.8 } ], "evidence": "From 7.000 through 9.000 s, the masked person moves progressively leftward and is only partially visible at the left boundary in the final sample.", "mask_relation": "subject matches the supplied selected-person mask", "source_subject_refs": "entity_1" } ], "notes": [ "The supplied mask track consistently identifies the person referred to by “He” and “his.”", "No separate source masks were supplied for the trowel or hawk, so their source-mask links are null.", "“the frame” and “the left” are excluded as meta/spatial references rather than concrete entity instances." ] }, "model_output_raw": { "caption": "He makes a final smoothing motion, scrapes his trowel clean on the hawk, and then walks out of the frame to the left.", "entity_groups": [ { "id": "entity_1", "label": "person", "spans": [ { "text": "He", "range": [ 0, 2 ] }, { "text": "his", "range": [ 43, 46 ] } ], "source_mask": "supplied_selected_person_mask" }, { "id": "entity_2", "label": "trowel", "spans": [ { "text": "his trowel", "range": [ 43, 53 ] } ], "source_mask": null }, { "id": "entity_3", "label": "hawk", "spans": [ { "text": "the hawk", "range": [ 63, 71 ] } ], "source_mask": null } ], "actions": [ { "mentions": [ { "text": "makes a final smoothing motion", "ranges": [ [ 3, 33 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "type": "light_verb", "subject_ref": "entity_1", "object_refs": [], "confidence": "high", "temporal_status": "localized", "intervals": [ { "start_s": 0.0, "end_s": 1.0 } ], "mask_relation": "subject matches the supplied selected-person mask", "evidence": "At 0.000 s, the masked person presses and moves a trowel against the wall in a smoothing posture." }, { "mentions": [ { "text": "scrapes", "ranges": [ [ 35, 42 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "type": "simple", "subject_ref": "entity_1", "object_refs": [ "entity_2", "entity_3" ], "confidence": "high", "temporal_status": "localized", "intervals": [ { "start_s": 1.0, "end_s": 3.0 }, { "start_s": 4.0, "end_s": 7.0 } ], "mask_relation": "subject matches the supplied selected-person mask", "evidence": "At 1.000–2.000 s and 4.000–6.000 s, the masked person holds the hawk and repeatedly brings the trowel across or against it." }, { "mentions": [ { "text": "walks out", "ranges": [ [ 82, 91 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "type": "phrasal", "subject_ref": "entity_1", "object_refs": [], "confidence": "high", "temporal_status": "localized", "intervals": [ { "start_s": 7.0, "end_s": 9.6 } ], "mask_relation": "subject matches the supplied selected-person mask", "evidence": "From 7.000 through 9.000 s, the masked person moves progressively leftward and is only partially visible at the left boundary in the final sample." } ], "notes": [ "The supplied mask track consistently identifies the person referred to by “He” and “his.”", "No separate source masks were supplied for the trowel or hawk, so their source-mask links are null.", "“the frame” and “the left” are excluded as meta/spatial references rather than concrete entity instances." ] }, "normalization_repairs": 0 }, { "source_video_id": "v_-0NfmccuL9Q", "caption_count_category": 8, "clip": { "caption_index": 1, "caption": "A person wearing a black jacket, blue jeans, and tan gloves enters from the right and crouches next to a car's rear wheel. They pick up a power tool from the ground. They then reach for a car jack and position it underneath the car.", "source_interval_sec": [ 6.17, 18.17 ], "loom_clip_id": "v_-0NfmccuL9Q_185_545", "clip_frame_interval_4fps": [ 185, 545 ], "sample_count_1fps": 13, "sampled_frames": [ { "frame": "00000", "timestamp_sec": 6.17, "offset_sec": 0, "rgb_path": "assets/cat_08/v_-0NfmccuL9Q_185_545/0000s_00000_rgb.jpg", "mask_path": "assets/cat_08/v_-0NfmccuL9Q_185_545/0000s_00000_mask.png", "mask_status": "available" }, { "frame": "00028", "timestamp_sec": 7.17, "offset_sec": 1, "rgb_path": "assets/cat_08/v_-0NfmccuL9Q_185_545/0001s_00028_rgb.jpg", "mask_path": "assets/cat_08/v_-0NfmccuL9Q_185_545/0001s_00028_mask.png", "mask_status": "available" }, { "frame": "00060", "timestamp_sec": 8.17, "offset_sec": 2, "rgb_path": "assets/cat_08/v_-0NfmccuL9Q_185_545/0002s_00060_rgb.jpg", "mask_path": "assets/cat_08/v_-0NfmccuL9Q_185_545/0002s_00060_mask.png", "mask_status": "available" }, { "frame": "00088", "timestamp_sec": 9.17, "offset_sec": 3, "rgb_path": "assets/cat_08/v_-0NfmccuL9Q_185_545/0003s_00088_rgb.jpg", "mask_path": "assets/cat_08/v_-0NfmccuL9Q_185_545/0003s_00088_mask.png", "mask_status": "available" }, { "frame": "00120", "timestamp_sec": 10.17, "offset_sec": 4, "rgb_path": "assets/cat_08/v_-0NfmccuL9Q_185_545/0004s_00120_rgb.jpg", "mask_path": "assets/cat_08/v_-0NfmccuL9Q_185_545/0004s_00120_mask.png", "mask_status": "available" }, { "frame": "00148", "timestamp_sec": 11.17, "offset_sec": 5, "rgb_path": "assets/cat_08/v_-0NfmccuL9Q_185_545/0005s_00148_rgb.jpg", "mask_path": "assets/cat_08/v_-0NfmccuL9Q_185_545/0005s_00148_mask.png", "mask_status": "available" }, { "frame": "00180", "timestamp_sec": 12.17, "offset_sec": 6, "rgb_path": "assets/cat_08/v_-0NfmccuL9Q_185_545/0006s_00180_rgb.jpg", "mask_path": "assets/cat_08/v_-0NfmccuL9Q_185_545/0006s_00180_mask.png", "mask_status": "available" }, { "frame": "00208", "timestamp_sec": 13.17, "offset_sec": 7, "rgb_path": "assets/cat_08/v_-0NfmccuL9Q_185_545/0007s_00208_rgb.jpg", "mask_path": "assets/cat_08/v_-0NfmccuL9Q_185_545/0007s_00208_mask.png", "mask_status": "available" }, { "frame": "00240", "timestamp_sec": 14.17, "offset_sec": 8, "rgb_path": "assets/cat_08/v_-0NfmccuL9Q_185_545/0008s_00240_rgb.jpg", "mask_path": "assets/cat_08/v_-0NfmccuL9Q_185_545/0008s_00240_mask.png", "mask_status": "available" }, { "frame": "00268", "timestamp_sec": 15.17, "offset_sec": 9, "rgb_path": "assets/cat_08/v_-0NfmccuL9Q_185_545/0009s_00268_rgb.jpg", "mask_path": "assets/cat_08/v_-0NfmccuL9Q_185_545/0009s_00268_mask.png", "mask_status": "available" }, { "frame": "00300", "timestamp_sec": 16.17, "offset_sec": 10, "rgb_path": "assets/cat_08/v_-0NfmccuL9Q_185_545/0010s_00300_rgb.jpg", "mask_path": "assets/cat_08/v_-0NfmccuL9Q_185_545/0010s_00300_mask.png", "mask_status": "available" }, { "frame": "00328", "timestamp_sec": 17.17, "offset_sec": 11, "rgb_path": "assets/cat_08/v_-0NfmccuL9Q_185_545/0011s_00328_rgb.jpg", "mask_path": "assets/cat_08/v_-0NfmccuL9Q_185_545/0011s_00328_mask.png", "mask_status": "available" }, { "frame": "00360", "timestamp_sec": 18.17, "offset_sec": 12.0, "rgb_path": "assets/cat_08/v_-0NfmccuL9Q_185_545/0012s_00360_rgb.jpg", "mask_path": "assets/cat_08/v_-0NfmccuL9Q_185_545/0012s_00360_mask.png", "mask_status": "available" } ], "obj_id": [ 0 ], "anno_id": [ 7 ], "fps_inferred_from_clip_interval": 30.0 }, "usage": { "input_tokens": 79705, "input_tokens_details": { "cache_write_tokens": 79702, "cached_tokens": 0 }, "output_tokens": 4430, "output_tokens_details": { "reasoning_tokens": 2311 }, "total_tokens": 84135 }, "model": "gpt-5.6-sol", "effort": "medium", "custom_id": "cat08_cap01_v_-0NfmccuL9Q_185_545", "annotation": { "caption": "A person wearing a black jacket, blue jeans, and tan gloves enters from the right and crouches next to a car's rear wheel. They pick up a power tool from the ground. They then reach for a car jack and position it underneath the car.", "entity_groups": [ { "id": "person_1", "label": "person", "spans": [ { "text": "A person", "range": [ 0, 8 ] }, { "text": "They", "range": [ 123, 127 ] }, { "text": "They", "range": [ 166, 170 ] } ], "source_mask_link": "selected_person" }, { "id": "jacket_1", "label": "jacket", "spans": [ { "text": "a black jacket", "range": [ 17, 31 ] } ], "source_mask_link": null }, { "id": "jeans_1", "label": "jeans", "spans": [ { "text": "blue jeans", "range": [ 33, 43 ] } ], "source_mask_link": null }, { "id": "gloves_1", "label": "gloves", "spans": [ { "text": "tan gloves", "range": [ 49, 59 ] } ], "source_mask_link": null }, { "id": "car_1", "label": "car", "spans": [ { "text": "a car's", "range": [ 103, 110 ] }, { "text": "the car", "range": [ 224, 231 ] } ], "source_mask_link": null }, { "id": "rear_wheel_1", "label": "wheel", "spans": [ { "text": "a car's rear wheel", "range": [ 103, 121 ] } ], "source_mask_link": null }, { "id": "power_tool_1", "label": "power_tool", "spans": [ { "text": "a power tool", "range": [ 136, 148 ] } ], "source_mask_link": null }, { "id": "ground_1", "label": "ground", "spans": [ { "text": "the ground", "range": [ 154, 164 ] } ], "source_mask_link": null }, { "id": "car_jack_1", "label": "car_jack", "spans": [ { "text": "a car jack", "range": [ 186, 196 ] }, { "text": "it", "range": [ 210, 212 ] } ], "source_mask_link": null } ], "actions": [ { "id": "action_1", "mentions": [ { "text": "wearing", "ranges": [ [ 9, 16 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "type": "simple", "subject_refs": [ "person_1" ], "object_refs": [ "jacket_1", "jeans_1", "gloves_1" ], "confidence": "high", "temporal_status": "supported", "intervals": [ { "start_s": 4.0, "end_s": 12.000000000000002, "source_start_s": 10.17, "source_end_s": 18.17 } ], "evidence": "From local 4.0 s onward, the mask-linked person is visibly wearing the black jacket, blue jeans, and tan gloves; this remains visible through the clip boundary.", "mask_relation": "subject_matches_supplied_selected_person_mask", "source_subject_refs": "person_1" }, { "id": "action_2", "mentions": [ { "text": "enters", "ranges": [ [ 60, 66 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "type": "simple", "subject_refs": [ "person_1" ], "object_refs": [], "confidence": "high", "temporal_status": "supported", "intervals": [ { "start_s": 1.0, "end_s": 6.0, "source_start_s": 7.17, "source_end_s": 12.17 } ], "evidence": "At local 1.0–5.0 s, the selected person progressively enters from the right edge, changing from a narrow partial view to a substantially visible body.", "mask_relation": "subject_matches_supplied_selected_person_mask", "source_subject_refs": "person_1" }, { "id": "action_3", "mentions": [ { "text": "crouches", "ranges": [ [ 86, 94 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "type": "posture", "subject_refs": [ "person_1" ], "object_refs": [], "confidence": "high", "temporal_status": "supported", "intervals": [ { "start_s": 6.0, "end_s": 12.000000000000002, "source_start_s": 12.17, "source_end_s": 18.17 } ], "evidence": "The mask-linked person is visibly crouched beside the rear wheel from local 6.0 s through the clip's right boundary.", "mask_relation": "subject_matches_supplied_selected_person_mask", "source_subject_refs": "person_1" }, { "id": "action_4", "mentions": [ { "text": "pick up", "ranges": [ [ 128, 135 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "type": "phrasal", "subject_refs": [ "person_1" ], "object_refs": [ "power_tool_1" ], "confidence": "high", "temporal_status": "supported", "intervals": [ { "start_s": 5.0, "end_s": 8.0, "source_start_s": 11.17, "source_end_s": 14.17 } ], "evidence": "At local 5.0–7.0 s, the selected person bends to the power tool, grasps it, and lifts it from the ground.", "mask_relation": "subject_matches_supplied_selected_person_mask", "source_subject_refs": "person_1" }, { "id": "action_5", "mentions": [ { "text": "reach", "ranges": [ [ 176, 181 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "type": "prepositional", "subject_refs": [ "person_1" ], "object_refs": [ "car_jack_1" ], "confidence": "medium", "temporal_status": "supported", "intervals": [ { "start_s": 7.0, "end_s": 9.0, "source_start_s": 13.17, "source_end_s": 15.17 } ], "evidence": "At local 7.0–8.0 s, the mask-linked person extends a gloved hand toward the car jack while crouched beside the wheel.", "mask_relation": "subject_matches_supplied_selected_person_mask", "source_subject_refs": "person_1" }, { "id": "action_6", "mentions": [ { "text": "position", "ranges": [ [ 201, 209 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "type": "simple", "subject_refs": [ "person_1" ], "object_refs": [ "car_jack_1" ], "confidence": "high", "temporal_status": "supported", "intervals": [ { "start_s": 8.0, "end_s": 12.000000000000002, "source_start_s": 14.17, "source_end_s": 18.17 } ], "evidence": "From local 8.0 s through the right boundary, the selected person manipulates and slides the jack into place underneath the car.", "mask_relation": "subject_matches_supplied_selected_person_mask", "source_subject_refs": "person_1" } ], "notes": [ "The supplied masklet identifies the single person entering from the right and later crouching beside the car.", "No separate source masks were supplied for the clothing, car, rear wheel, power tool, ground, or car jack, so those entity groups have null source-mask links.", "Action intervals are sparse clip-relative support intervals derived from the sampled clean RGB frames. The original source interval is [6.170000, 18.170000), corresponding to local [0.000000, 12.000000).", "The sampled frame at local 12.0 s lies at the clip boundary; actions visibly continuing there end exactly at 12.0 s and are not extended beyond the source interval." ] }, "model_output_raw": { "caption": "A person wearing a black jacket, blue jeans, and tan gloves enters from the right and crouches next to a car's rear wheel. They pick up a power tool from the ground. They then reach for a car jack and position it underneath the car.", "entity_groups": [ { "id": "person_1", "entity_type": "person", "spans": [ { "text": "A person", "range": [ 0, 8 ] }, { "text": "They", "range": [ 123, 127 ] }, { "text": "They", "range": [ 166, 170 ] } ], "source_mask_link": "selected_person_mask" }, { "id": "jacket_1", "entity_type": "jacket", "spans": [ { "text": "a black jacket", "range": [ 17, 31 ] } ], "source_mask_link": null }, { "id": "jeans_1", "entity_type": "jeans", "spans": [ { "text": "blue jeans", "range": [ 33, 43 ] } ], "source_mask_link": null }, { "id": "gloves_1", "entity_type": "gloves", "spans": [ { "text": "tan gloves", "range": [ 49, 59 ] } ], "source_mask_link": null }, { "id": "car_1", "entity_type": "car", "spans": [ { "text": "a car's", "range": [ 103, 110 ] }, { "text": "the car", "range": [ 224, 231 ] } ], "source_mask_link": null }, { "id": "rear_wheel_1", "entity_type": "wheel", "spans": [ { "text": "a car's rear wheel", "range": [ 103, 121 ] } ], "source_mask_link": null }, { "id": "power_tool_1", "entity_type": "power_tool", "spans": [ { "text": "a power tool", "range": [ 136, 148 ] } ], "source_mask_link": null }, { "id": "ground_1", "entity_type": "ground", "spans": [ { "text": "the ground", "range": [ 154, 164 ] } ], "source_mask_link": null }, { "id": "car_jack_1", "entity_type": "car_jack", "spans": [ { "text": "a car jack", "range": [ 186, 196 ] }, { "text": "it", "range": [ 210, 212 ] } ], "source_mask_link": null } ], "actions": [ { "id": "action_1", "mentions": [ { "text": "wearing", "ranges": [ [ 9, 16 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "type": "simple", "subject_entity_group": "person_1", "object_entity_groups": [ "jacket_1", "jeans_1", "gloves_1" ], "confidence": "high", "temporal_status": "supported", "intervals": [ { "start_s": 4.0, "end_s": 12.0 } ], "mask_relation": "subject_matches_supplied_selected_person_mask", "evidence": "From local 4.0 s onward, the mask-linked person is visibly wearing the black jacket, blue jeans, and tan gloves; this remains visible through the clip boundary." }, { "id": "action_2", "mentions": [ { "text": "enters", "ranges": [ [ 60, 66 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "type": "simple", "subject_entity_group": "person_1", "object_entity_groups": [], "confidence": "high", "temporal_status": "supported", "intervals": [ { "start_s": 1.0, "end_s": 6.0 } ], "mask_relation": "subject_matches_supplied_selected_person_mask", "evidence": "At local 1.0–5.0 s, the selected person progressively enters from the right edge, changing from a narrow partial view to a substantially visible body." }, { "id": "action_3", "mentions": [ { "text": "crouches", "ranges": [ [ 86, 94 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "type": "posture", "subject_entity_group": "person_1", "object_entity_groups": [], "confidence": "high", "temporal_status": "supported", "intervals": [ { "start_s": 6.0, "end_s": 12.0 } ], "mask_relation": "subject_matches_supplied_selected_person_mask", "evidence": "The mask-linked person is visibly crouched beside the rear wheel from local 6.0 s through the clip's right boundary." }, { "id": "action_4", "mentions": [ { "text": "pick up", "ranges": [ [ 128, 135 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "type": "phrasal", "subject_entity_group": "person_1", "object_entity_groups": [ "power_tool_1" ], "confidence": "high", "temporal_status": "supported", "intervals": [ { "start_s": 5.0, "end_s": 8.0 } ], "mask_relation": "subject_matches_supplied_selected_person_mask", "evidence": "At local 5.0–7.0 s, the selected person bends to the power tool, grasps it, and lifts it from the ground." }, { "id": "action_5", "mentions": [ { "text": "reach", "ranges": [ [ 176, 181 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "type": "prepositional", "subject_entity_group": "person_1", "object_entity_groups": [ "car_jack_1" ], "confidence": "medium", "temporal_status": "supported", "intervals": [ { "start_s": 7.0, "end_s": 9.0 } ], "mask_relation": "subject_matches_supplied_selected_person_mask", "evidence": "At local 7.0–8.0 s, the mask-linked person extends a gloved hand toward the car jack while crouched beside the wheel." }, { "id": "action_6", "mentions": [ { "text": "position", "ranges": [ [ 201, 209 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "type": "simple", "subject_entity_group": "person_1", "object_entity_groups": [ "car_jack_1" ], "context_entity_groups": [ "car_1" ], "confidence": "high", "temporal_status": "supported", "intervals": [ { "start_s": 8.0, "end_s": 12.0 } ], "mask_relation": "subject_matches_supplied_selected_person_mask", "evidence": "From local 8.0 s through the right boundary, the selected person manipulates and slides the jack into place underneath the car." } ], "notes": [ "The supplied masklet identifies the single person entering from the right and later crouching beside the car.", "No separate source masks were supplied for the clothing, car, rear wheel, power tool, ground, or car jack, so those entity groups have null source-mask links.", "Action intervals are sparse clip-relative support intervals derived from the sampled clean RGB frames. The original source interval is [6.170000, 18.170000), corresponding to local [0.000000, 12.000000).", "The sampled frame at local 12.0 s lies at the clip boundary; actions visibly continuing there end exactly at 12.0 s and are not extended beyond the source interval." ] }, "normalization_repairs": 0 }, { "source_video_id": "v_-0NfmccuL9Q", "caption_count_category": 8, "clip": { "caption_index": 2, "caption": "While kneeling, the person uses the power tool to operate the car jack, steadily raising the car's rear wheel off the ground. Their left hand rests on top of the tire as it lifts. They remain focused on the jack and the tool during this action.", "source_interval_sec": [ 18.17, 36.67 ], "loom_clip_id": "v_-0NfmccuL9Q_545_1099", "clip_frame_interval_4fps": [ 545, 1099 ], "sample_count_1fps": 19, "sampled_frames": [ { "frame": "00000", "timestamp_sec": 18.17, "offset_sec": 0, "rgb_path": "assets/cat_08/v_-0NfmccuL9Q_545_1099/0000s_00000_rgb.jpg", "mask_path": "assets/cat_08/v_-0NfmccuL9Q_545_1099/0000s_00000_mask.png", "mask_status": "available" }, { "frame": "00028", "timestamp_sec": 19.17, "offset_sec": 1, "rgb_path": "assets/cat_08/v_-0NfmccuL9Q_545_1099/0001s_00028_rgb.jpg", "mask_path": "assets/cat_08/v_-0NfmccuL9Q_545_1099/0001s_00028_mask.png", "mask_status": "available" }, { "frame": "00060", "timestamp_sec": 20.17, "offset_sec": 2, "rgb_path": "assets/cat_08/v_-0NfmccuL9Q_545_1099/0002s_00060_rgb.jpg", "mask_path": "assets/cat_08/v_-0NfmccuL9Q_545_1099/0002s_00060_mask.png", "mask_status": "available" }, { "frame": "00088", "timestamp_sec": 21.17, "offset_sec": 3, "rgb_path": "assets/cat_08/v_-0NfmccuL9Q_545_1099/0003s_00088_rgb.jpg", "mask_path": "assets/cat_08/v_-0NfmccuL9Q_545_1099/0003s_00088_mask.png", "mask_status": "available" }, { "frame": "00120", "timestamp_sec": 22.17, "offset_sec": 4, "rgb_path": "assets/cat_08/v_-0NfmccuL9Q_545_1099/0004s_00120_rgb.jpg", "mask_path": "assets/cat_08/v_-0NfmccuL9Q_545_1099/0004s_00120_mask.png", "mask_status": "available" }, { "frame": "00148", "timestamp_sec": 23.17, "offset_sec": 5, "rgb_path": "assets/cat_08/v_-0NfmccuL9Q_545_1099/0005s_00148_rgb.jpg", "mask_path": "assets/cat_08/v_-0NfmccuL9Q_545_1099/0005s_00148_mask.png", "mask_status": "available" }, { "frame": "00180", "timestamp_sec": 24.17, "offset_sec": 6, "rgb_path": "assets/cat_08/v_-0NfmccuL9Q_545_1099/0006s_00180_rgb.jpg", "mask_path": "assets/cat_08/v_-0NfmccuL9Q_545_1099/0006s_00180_mask.png", "mask_status": "available" }, { "frame": "00208", "timestamp_sec": 25.17, "offset_sec": 7, "rgb_path": "assets/cat_08/v_-0NfmccuL9Q_545_1099/0007s_00208_rgb.jpg", "mask_path": "assets/cat_08/v_-0NfmccuL9Q_545_1099/0007s_00208_mask.png", "mask_status": "available" }, { "frame": "00240", "timestamp_sec": 26.17, "offset_sec": 8, "rgb_path": "assets/cat_08/v_-0NfmccuL9Q_545_1099/0008s_00240_rgb.jpg", "mask_path": "assets/cat_08/v_-0NfmccuL9Q_545_1099/0008s_00240_mask.png", "mask_status": "available" }, { "frame": "00268", "timestamp_sec": 27.17, "offset_sec": 9, "rgb_path": "assets/cat_08/v_-0NfmccuL9Q_545_1099/0009s_00268_rgb.jpg", "mask_path": "assets/cat_08/v_-0NfmccuL9Q_545_1099/0009s_00268_mask.png", "mask_status": "available" }, { "frame": "00300", "timestamp_sec": 28.17, "offset_sec": 10, "rgb_path": "assets/cat_08/v_-0NfmccuL9Q_545_1099/0010s_00300_rgb.jpg", "mask_path": "assets/cat_08/v_-0NfmccuL9Q_545_1099/0010s_00300_mask.png", "mask_status": "available" }, { "frame": "00328", "timestamp_sec": 29.17, "offset_sec": 11, "rgb_path": "assets/cat_08/v_-0NfmccuL9Q_545_1099/0011s_00328_rgb.jpg", "mask_path": "assets/cat_08/v_-0NfmccuL9Q_545_1099/0011s_00328_mask.png", "mask_status": "available" }, { "frame": "00360", "timestamp_sec": 30.17, "offset_sec": 12, "rgb_path": "assets/cat_08/v_-0NfmccuL9Q_545_1099/0012s_00360_rgb.jpg", "mask_path": "assets/cat_08/v_-0NfmccuL9Q_545_1099/0012s_00360_mask.png", "mask_status": "available" }, { "frame": "00388", "timestamp_sec": 31.17, "offset_sec": 13, "rgb_path": "assets/cat_08/v_-0NfmccuL9Q_545_1099/0013s_00388_rgb.jpg", "mask_path": "assets/cat_08/v_-0NfmccuL9Q_545_1099/0013s_00388_mask.png", "mask_status": "available" }, { "frame": "00420", "timestamp_sec": 32.17, "offset_sec": 14, "rgb_path": "assets/cat_08/v_-0NfmccuL9Q_545_1099/0014s_00420_rgb.jpg", "mask_path": "assets/cat_08/v_-0NfmccuL9Q_545_1099/0014s_00420_mask.png", "mask_status": "available" }, { "frame": "00448", "timestamp_sec": 33.17, "offset_sec": 15, "rgb_path": "assets/cat_08/v_-0NfmccuL9Q_545_1099/0015s_00448_rgb.jpg", "mask_path": "assets/cat_08/v_-0NfmccuL9Q_545_1099/0015s_00448_mask.png", "mask_status": "available" }, { "frame": "00480", "timestamp_sec": 34.17, "offset_sec": 16, "rgb_path": "assets/cat_08/v_-0NfmccuL9Q_545_1099/0016s_00480_rgb.jpg", "mask_path": "assets/cat_08/v_-0NfmccuL9Q_545_1099/0016s_00480_mask.png", "mask_status": "available" }, { "frame": "00508", "timestamp_sec": 35.17, "offset_sec": 17, "rgb_path": "assets/cat_08/v_-0NfmccuL9Q_545_1099/0017s_00508_rgb.jpg", "mask_path": "assets/cat_08/v_-0NfmccuL9Q_545_1099/0017s_00508_mask.png", "mask_status": "available" }, { "frame": "00540", "timestamp_sec": 36.17, "offset_sec": 18, "rgb_path": "assets/cat_08/v_-0NfmccuL9Q_545_1099/0018s_00540_rgb.jpg", "mask_path": "assets/cat_08/v_-0NfmccuL9Q_545_1099/0018s_00540_mask.png", "mask_status": "available" } ], "obj_id": [ 0 ], "anno_id": [ 12 ], "fps_inferred_from_clip_interval": 29.9459 }, "usage": { "input_tokens": 109412, "input_tokens_details": { "cache_write_tokens": 109409, "cached_tokens": 0 }, "output_tokens": 4177, "output_tokens_details": { "reasoning_tokens": 2588 }, "total_tokens": 113589 }, "model": "gpt-5.6-sol", "effort": "medium", "custom_id": "cat08_cap02_v_-0NfmccuL9Q_545_1099", "annotation": { "caption": "While kneeling, the person uses the power tool to operate the car jack, steadily raising the car's rear wheel off the ground. Their left hand rests on top of the tire as it lifts. They remain focused on the jack and the tool during this action.", "entity_groups": [ { "id": "person_1", "label": "person", "spans": [ { "text": "the person", "range": [ 16, 26 ] }, { "text": "Their", "range": [ 126, 131 ] }, { "text": "They", "range": [ 180, 184 ] } ], "source_mask_link": "selected_person" }, { "id": "power_tool_1", "label": "power tool", "spans": [ { "text": "the power tool", "range": [ 32, 46 ] }, { "text": "the tool", "range": [ 216, 224 ] } ], "source_mask_link": null }, { "id": "car_jack_1", "label": "car jack", "spans": [ { "text": "the car jack", "range": [ 58, 70 ] }, { "text": "the jack", "range": [ 203, 211 ] } ], "source_mask_link": null }, { "id": "car_1", "label": "car", "spans": [ { "text": "the car's", "range": [ 89, 98 ] } ], "source_mask_link": null }, { "id": "rear_wheel_1", "label": "rear wheel/tire", "spans": [ { "text": "the car's rear wheel", "range": [ 89, 109 ] }, { "text": "the tire", "range": [ 158, 166 ] }, { "text": "it", "range": [ 170, 172 ] } ], "source_mask_link": null }, { "id": "ground_1", "label": "ground", "spans": [ { "text": "the ground", "range": [ 114, 124 ] } ], "source_mask_link": null }, { "id": "left_hand_1", "label": "left hand", "spans": [ { "text": "Their left hand", "range": [ 126, 141 ] } ], "source_mask_link": null } ], "actions": [ { "id": "action_1", "mentions": [ { "text": "kneeling", "ranges": [ [ 6, 14 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "type": "posture", "subject_refs": [ "person_1" ], "object_refs": [], "confidence": "high", "temporal_status": "supported", "intervals": [ { "start_s": 0.0, "end_s": 18.5, "source_start_s": 18.17, "source_end_s": 36.67 } ], "evidence": "The selected person remains in a low kneeling posture beside the wheel in every sampled clean frame from 0.0 through 18.0 seconds.", "mask_relation": "subject_matches_supplied_selected_person_mask", "source_subject_refs": [ "person_1" ] }, { "id": "action_2", "mentions": [ { "text": "uses", "ranges": [ [ 27, 31 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "type": "simple", "subject_refs": [ "person_1" ], "object_refs": [ "power_tool_1" ], "confidence": "high", "temporal_status": "supported", "intervals": [ { "start_s": 0.0, "end_s": 18.5, "source_start_s": 18.17, "source_end_s": 36.67 } ], "evidence": "Across all sampled clean frames, the selected person grips and applies the powered drill-like tool.", "mask_relation": "subject_matches_supplied_selected_person_mask", "source_subject_refs": [ "person_1" ] }, { "id": "action_3", "mentions": [ { "text": "operate", "ranges": [ [ 50, 57 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "type": "simple", "subject_refs": [ "person_1" ], "object_refs": [ "car_jack_1" ], "confidence": "high", "temporal_status": "supported", "intervals": [ { "start_s": 0.0, "end_s": 18.5, "source_start_s": 18.17, "source_end_s": 36.67 } ], "evidence": "The tool is visibly coupled to the jack mechanism while the selected person operates it throughout the sampled sequence.", "mask_relation": "subject_matches_supplied_selected_person_mask", "source_subject_refs": [ "person_1" ] }, { "id": "action_4", "mentions": [ { "text": "raising", "ranges": [ [ 81, 88 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "type": "simple", "subject_refs": [ "person_1" ], "object_refs": [ "rear_wheel_1" ], "confidence": "high", "temporal_status": "supported", "intervals": [ { "start_s": 0.0, "end_s": 18.5, "source_start_s": 18.17, "source_end_s": 36.67 } ], "evidence": "The clean-frame sequence shows the jack extending and the rear wheel becoming progressively more elevated while the selected person operates the tool.", "mask_relation": "subject_matches_supplied_selected_person_mask", "source_subject_refs": [ "person_1" ] } ], "notes": [ "The supplied masklet consistently identifies the person; no separate masks were supplied for the tool, jack, car, wheel/tire, ground, or hand.", "\"rests\" was not included because its explicit subject is the separately annotated left hand, which has no supplied source mask.", "\"lifts\" was not included because its explicit subject is the wheel/tire, which has no supplied source mask.", "\"remain focused\" describes a mental state rather than a visually meaningful physical action." ] }, "model_output_raw": { "caption": "While kneeling, the person uses the power tool to operate the car jack, steadily raising the car's rear wheel off the ground. Their left hand rests on top of the tire as it lifts. They remain focused on the jack and the tool during this action.", "entity_groups": [ { "id": "person_1", "label": "person", "spans": [ { "text": "the person", "range": [ 16, 26 ] }, { "text": "Their", "range": [ 126, 131 ] }, { "text": "They", "range": [ 180, 184 ] } ], "source_mask": "selected_person_masklet" }, { "id": "power_tool_1", "label": "power tool", "spans": [ { "text": "the power tool", "range": [ 32, 46 ] }, { "text": "the tool", "range": [ 216, 224 ] } ], "source_mask": null }, { "id": "car_jack_1", "label": "car jack", "spans": [ { "text": "the car jack", "range": [ 58, 70 ] }, { "text": "the jack", "range": [ 203, 211 ] } ], "source_mask": null }, { "id": "car_1", "label": "car", "spans": [ { "text": "the car's", "range": [ 89, 98 ] } ], "source_mask": null }, { "id": "rear_wheel_1", "label": "rear wheel/tire", "spans": [ { "text": "the car's rear wheel", "range": [ 89, 109 ] }, { "text": "the tire", "range": [ 158, 166 ] }, { "text": "it", "range": [ 170, 172 ] } ], "source_mask": null }, { "id": "ground_1", "label": "ground", "spans": [ { "text": "the ground", "range": [ 114, 124 ] } ], "source_mask": null }, { "id": "left_hand_1", "label": "left hand", "spans": [ { "text": "Their left hand", "range": [ 126, 141 ] } ], "source_mask": null } ], "actions": [ { "id": "action_1", "mentions": [ { "text": "kneeling", "ranges": [ [ 6, 14 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "type": "posture", "subject_refs": [ "person_1" ], "object_refs": [], "confidence": "high", "temporal_status": "supported", "intervals": [ { "clip_relative_s": [ 0.0, 18.5 ], "source_video_s": [ 18.17, 36.67 ] } ], "evidence": "The selected person remains in a low kneeling posture beside the wheel in every sampled clean frame from 0.0 through 18.0 seconds.", "mask_relation": "subject_matches_supplied_selected_person_mask" }, { "id": "action_2", "mentions": [ { "text": "uses", "ranges": [ [ 27, 31 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "type": "simple", "subject_refs": [ "person_1" ], "object_refs": [ "power_tool_1" ], "confidence": "high", "temporal_status": "supported", "intervals": [ { "clip_relative_s": [ 0.0, 18.5 ], "source_video_s": [ 18.17, 36.67 ] } ], "evidence": "Across all sampled clean frames, the selected person grips and applies the powered drill-like tool.", "mask_relation": "subject_matches_supplied_selected_person_mask" }, { "id": "action_3", "mentions": [ { "text": "operate", "ranges": [ [ 50, 57 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "type": "simple", "subject_refs": [ "person_1" ], "object_refs": [ "car_jack_1" ], "confidence": "high", "temporal_status": "supported", "intervals": [ { "clip_relative_s": [ 0.0, 18.5 ], "source_video_s": [ 18.17, 36.67 ] } ], "evidence": "The tool is visibly coupled to the jack mechanism while the selected person operates it throughout the sampled sequence.", "mask_relation": "subject_matches_supplied_selected_person_mask" }, { "id": "action_4", "mentions": [ { "text": "raising", "ranges": [ [ 81, 88 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "type": "simple", "subject_refs": [ "person_1" ], "object_refs": [ "rear_wheel_1" ], "confidence": "high", "temporal_status": "supported", "intervals": [ { "clip_relative_s": [ 0.0, 18.5 ], "source_video_s": [ 18.17, 36.67 ] } ], "evidence": "The clean-frame sequence shows the jack extending and the rear wheel becoming progressively more elevated while the selected person operates the tool.", "mask_relation": "subject_matches_supplied_selected_person_mask" } ], "notes": [ "The supplied masklet consistently identifies the person; no separate masks were supplied for the tool, jack, car, wheel/tire, ground, or hand.", "\"rests\" was not included because its explicit subject is the separately annotated left hand, which has no supplied source mask.", "\"lifts\" was not included because its explicit subject is the wheel/tire, which has no supplied source mask.", "\"remain focused\" describes a mental state rather than a visually meaningful physical action." ] }, "normalization_repairs": 0 }, { "source_video_id": "v_-0NfmccuL9Q", "caption_count_category": 8, "clip": { "caption_index": 3, "caption": "The person attaches a socket to the power tool and uses it to systematically remove all the lug nuts from the wheel. After placing the nuts and the tool on the ground, they grip the wheel with both hands and pull it off the car. They then roll the removed tire away to the right.", "source_interval_sec": [ 36.67, 79.58 ], "loom_clip_id": "v_-0NfmccuL9Q_1099_2385", "clip_frame_interval_4fps": [ 1099, 2385 ], "sample_count_1fps": 43, "sampled_frames": [ { "frame": "00000", "timestamp_sec": 36.67, "offset_sec": 0, "rgb_path": "assets/cat_08/v_-0NfmccuL9Q_1099_2385/0000s_00000_rgb.jpg", "mask_path": "assets/cat_08/v_-0NfmccuL9Q_1099_2385/0000s_00000_mask.png", "mask_status": "available" }, { "frame": "00028", "timestamp_sec": 37.67, "offset_sec": 1, "rgb_path": "assets/cat_08/v_-0NfmccuL9Q_1099_2385/0001s_00028_rgb.jpg", "mask_path": "assets/cat_08/v_-0NfmccuL9Q_1099_2385/0001s_00028_mask.png", "mask_status": "available" }, { "frame": "00060", "timestamp_sec": 38.67, "offset_sec": 2, "rgb_path": "assets/cat_08/v_-0NfmccuL9Q_1099_2385/0002s_00060_rgb.jpg", "mask_path": "assets/cat_08/v_-0NfmccuL9Q_1099_2385/0002s_00060_mask.png", "mask_status": "available" }, { "frame": "00088", "timestamp_sec": 39.67, "offset_sec": 3, "rgb_path": "assets/cat_08/v_-0NfmccuL9Q_1099_2385/0003s_00088_rgb.jpg", "mask_path": "assets/cat_08/v_-0NfmccuL9Q_1099_2385/0003s_00088_mask.png", "mask_status": "available" }, { "frame": "00120", "timestamp_sec": 40.67, "offset_sec": 4, "rgb_path": "assets/cat_08/v_-0NfmccuL9Q_1099_2385/0004s_00120_rgb.jpg", "mask_path": "assets/cat_08/v_-0NfmccuL9Q_1099_2385/0004s_00120_mask.png", "mask_status": "available" }, { "frame": "00148", "timestamp_sec": 41.67, "offset_sec": 5, "rgb_path": "assets/cat_08/v_-0NfmccuL9Q_1099_2385/0005s_00148_rgb.jpg", "mask_path": "assets/cat_08/v_-0NfmccuL9Q_1099_2385/0005s_00148_mask.png", "mask_status": "available" }, { "frame": "00180", "timestamp_sec": 42.67, "offset_sec": 6, "rgb_path": "assets/cat_08/v_-0NfmccuL9Q_1099_2385/0006s_00180_rgb.jpg", "mask_path": "assets/cat_08/v_-0NfmccuL9Q_1099_2385/0006s_00180_mask.png", "mask_status": "available" }, { "frame": "00208", "timestamp_sec": 43.67, "offset_sec": 7, "rgb_path": "assets/cat_08/v_-0NfmccuL9Q_1099_2385/0007s_00208_rgb.jpg", "mask_path": "assets/cat_08/v_-0NfmccuL9Q_1099_2385/0007s_00208_mask.png", "mask_status": "available" }, { "frame": "00240", "timestamp_sec": 44.67, "offset_sec": 8, "rgb_path": "assets/cat_08/v_-0NfmccuL9Q_1099_2385/0008s_00240_rgb.jpg", "mask_path": "assets/cat_08/v_-0NfmccuL9Q_1099_2385/0008s_00240_mask.png", "mask_status": "available" }, { "frame": "00268", "timestamp_sec": 45.67, "offset_sec": 9, "rgb_path": "assets/cat_08/v_-0NfmccuL9Q_1099_2385/0009s_00268_rgb.jpg", "mask_path": "assets/cat_08/v_-0NfmccuL9Q_1099_2385/0009s_00268_mask.png", "mask_status": "available" }, { "frame": "00300", "timestamp_sec": 46.67, "offset_sec": 10, "rgb_path": "assets/cat_08/v_-0NfmccuL9Q_1099_2385/0010s_00300_rgb.jpg", "mask_path": "assets/cat_08/v_-0NfmccuL9Q_1099_2385/0010s_00300_mask.png", "mask_status": "available" }, { "frame": "00328", "timestamp_sec": 47.67, "offset_sec": 11, "rgb_path": "assets/cat_08/v_-0NfmccuL9Q_1099_2385/0011s_00328_rgb.jpg", "mask_path": "assets/cat_08/v_-0NfmccuL9Q_1099_2385/0011s_00328_mask.png", "mask_status": "available" }, { "frame": "00360", "timestamp_sec": 48.67, "offset_sec": 12, "rgb_path": "assets/cat_08/v_-0NfmccuL9Q_1099_2385/0012s_00360_rgb.jpg", "mask_path": "assets/cat_08/v_-0NfmccuL9Q_1099_2385/0012s_00360_mask.png", "mask_status": "available" }, { "frame": "00388", "timestamp_sec": 49.67, "offset_sec": 13, "rgb_path": "assets/cat_08/v_-0NfmccuL9Q_1099_2385/0013s_00388_rgb.jpg", "mask_path": "assets/cat_08/v_-0NfmccuL9Q_1099_2385/0013s_00388_mask.png", "mask_status": "available" }, { "frame": "00420", "timestamp_sec": 50.67, "offset_sec": 14, "rgb_path": "assets/cat_08/v_-0NfmccuL9Q_1099_2385/0014s_00420_rgb.jpg", "mask_path": "assets/cat_08/v_-0NfmccuL9Q_1099_2385/0014s_00420_mask.png", "mask_status": "available" }, { "frame": "00448", "timestamp_sec": 51.67, "offset_sec": 15, "rgb_path": "assets/cat_08/v_-0NfmccuL9Q_1099_2385/0015s_00448_rgb.jpg", "mask_path": "assets/cat_08/v_-0NfmccuL9Q_1099_2385/0015s_00448_mask.png", "mask_status": "available" }, { "frame": "00480", "timestamp_sec": 52.67, "offset_sec": 16, "rgb_path": "assets/cat_08/v_-0NfmccuL9Q_1099_2385/0016s_00480_rgb.jpg", "mask_path": "assets/cat_08/v_-0NfmccuL9Q_1099_2385/0016s_00480_mask.png", "mask_status": "available" }, { "frame": "00508", "timestamp_sec": 53.67, "offset_sec": 17, "rgb_path": "assets/cat_08/v_-0NfmccuL9Q_1099_2385/0017s_00508_rgb.jpg", "mask_path": "assets/cat_08/v_-0NfmccuL9Q_1099_2385/0017s_00508_mask.png", "mask_status": "available" }, { "frame": "00540", "timestamp_sec": 54.67, "offset_sec": 18, "rgb_path": "assets/cat_08/v_-0NfmccuL9Q_1099_2385/0018s_00540_rgb.jpg", "mask_path": "assets/cat_08/v_-0NfmccuL9Q_1099_2385/0018s_00540_mask.png", "mask_status": "available" }, { "frame": "00568", "timestamp_sec": 55.67, "offset_sec": 19, "rgb_path": "assets/cat_08/v_-0NfmccuL9Q_1099_2385/0019s_00568_rgb.jpg", "mask_path": "assets/cat_08/v_-0NfmccuL9Q_1099_2385/0019s_00568_mask.png", "mask_status": "available" }, { "frame": "00600", "timestamp_sec": 56.67, "offset_sec": 20, "rgb_path": "assets/cat_08/v_-0NfmccuL9Q_1099_2385/0020s_00600_rgb.jpg", "mask_path": "assets/cat_08/v_-0NfmccuL9Q_1099_2385/0020s_00600_mask.png", "mask_status": "available" }, { "frame": "00628", "timestamp_sec": 57.67, "offset_sec": 21, "rgb_path": "assets/cat_08/v_-0NfmccuL9Q_1099_2385/0021s_00628_rgb.jpg", "mask_path": "assets/cat_08/v_-0NfmccuL9Q_1099_2385/0021s_00628_mask.png", "mask_status": "available" }, { "frame": "00660", "timestamp_sec": 58.67, "offset_sec": 22, "rgb_path": "assets/cat_08/v_-0NfmccuL9Q_1099_2385/0022s_00660_rgb.jpg", "mask_path": "assets/cat_08/v_-0NfmccuL9Q_1099_2385/0022s_00660_mask.png", "mask_status": "available" }, { "frame": "00688", "timestamp_sec": 59.67, "offset_sec": 23, "rgb_path": "assets/cat_08/v_-0NfmccuL9Q_1099_2385/0023s_00688_rgb.jpg", "mask_path": "assets/cat_08/v_-0NfmccuL9Q_1099_2385/0023s_00688_mask.png", "mask_status": "available" }, { "frame": "00720", "timestamp_sec": 60.67, "offset_sec": 24, "rgb_path": "assets/cat_08/v_-0NfmccuL9Q_1099_2385/0024s_00720_rgb.jpg", "mask_path": "assets/cat_08/v_-0NfmccuL9Q_1099_2385/0024s_00720_mask.png", "mask_status": "available" }, { "frame": "00748", "timestamp_sec": 61.67, "offset_sec": 25, "rgb_path": "assets/cat_08/v_-0NfmccuL9Q_1099_2385/0025s_00748_rgb.jpg", "mask_path": "assets/cat_08/v_-0NfmccuL9Q_1099_2385/0025s_00748_mask.png", "mask_status": "available" }, { "frame": "00780", "timestamp_sec": 62.67, "offset_sec": 26, "rgb_path": "assets/cat_08/v_-0NfmccuL9Q_1099_2385/0026s_00780_rgb.jpg", "mask_path": "assets/cat_08/v_-0NfmccuL9Q_1099_2385/0026s_00780_mask.png", "mask_status": "available" }, { "frame": "00808", "timestamp_sec": 63.67, "offset_sec": 27, "rgb_path": "assets/cat_08/v_-0NfmccuL9Q_1099_2385/0027s_00808_rgb.jpg", "mask_path": "assets/cat_08/v_-0NfmccuL9Q_1099_2385/0027s_00808_mask.png", "mask_status": "available" }, { "frame": "00840", "timestamp_sec": 64.67, "offset_sec": 28, "rgb_path": "assets/cat_08/v_-0NfmccuL9Q_1099_2385/0028s_00840_rgb.jpg", "mask_path": "assets/cat_08/v_-0NfmccuL9Q_1099_2385/0028s_00840_mask.png", "mask_status": "available" }, { "frame": "00868", "timestamp_sec": 65.67, "offset_sec": 29, "rgb_path": "assets/cat_08/v_-0NfmccuL9Q_1099_2385/0029s_00868_rgb.jpg", "mask_path": "assets/cat_08/v_-0NfmccuL9Q_1099_2385/0029s_00868_mask.png", "mask_status": "available" }, { "frame": "00900", "timestamp_sec": 66.67, "offset_sec": 30, "rgb_path": "assets/cat_08/v_-0NfmccuL9Q_1099_2385/0030s_00900_rgb.jpg", "mask_path": "assets/cat_08/v_-0NfmccuL9Q_1099_2385/0030s_00900_mask.png", "mask_status": "available" }, { "frame": "00928", "timestamp_sec": 67.67, "offset_sec": 31, "rgb_path": "assets/cat_08/v_-0NfmccuL9Q_1099_2385/0031s_00928_rgb.jpg", "mask_path": "assets/cat_08/v_-0NfmccuL9Q_1099_2385/0031s_00928_mask.png", "mask_status": "available" }, { "frame": "00960", "timestamp_sec": 68.67, "offset_sec": 32, "rgb_path": "assets/cat_08/v_-0NfmccuL9Q_1099_2385/0032s_00960_rgb.jpg", "mask_path": "assets/cat_08/v_-0NfmccuL9Q_1099_2385/0032s_00960_mask.png", "mask_status": "available" }, { "frame": "00988", "timestamp_sec": 69.67, "offset_sec": 33, "rgb_path": "assets/cat_08/v_-0NfmccuL9Q_1099_2385/0033s_00988_rgb.jpg", "mask_path": "assets/cat_08/v_-0NfmccuL9Q_1099_2385/0033s_00988_mask.png", "mask_status": "available" }, { "frame": "01020", "timestamp_sec": 70.67, "offset_sec": 34, "rgb_path": "assets/cat_08/v_-0NfmccuL9Q_1099_2385/0034s_01020_rgb.jpg", "mask_path": "assets/cat_08/v_-0NfmccuL9Q_1099_2385/0034s_01020_mask.png", "mask_status": "available" }, { "frame": "01048", "timestamp_sec": 71.67, "offset_sec": 35, "rgb_path": "assets/cat_08/v_-0NfmccuL9Q_1099_2385/0035s_01048_rgb.jpg", "mask_path": "assets/cat_08/v_-0NfmccuL9Q_1099_2385/0035s_01048_mask.png", "mask_status": "available" }, { "frame": "01080", "timestamp_sec": 72.67, "offset_sec": 36, "rgb_path": "assets/cat_08/v_-0NfmccuL9Q_1099_2385/0036s_01080_rgb.jpg", "mask_path": "assets/cat_08/v_-0NfmccuL9Q_1099_2385/0036s_01080_mask.png", "mask_status": "available" }, { "frame": "01108", "timestamp_sec": 73.67, "offset_sec": 37, "rgb_path": "assets/cat_08/v_-0NfmccuL9Q_1099_2385/0037s_01108_rgb.jpg", "mask_path": "assets/cat_08/v_-0NfmccuL9Q_1099_2385/0037s_01108_mask.png", "mask_status": "available" }, { "frame": "01140", "timestamp_sec": 74.67, "offset_sec": 38, "rgb_path": "assets/cat_08/v_-0NfmccuL9Q_1099_2385/0038s_01140_rgb.jpg", "mask_path": "assets/cat_08/v_-0NfmccuL9Q_1099_2385/0038s_01140_mask.png", "mask_status": "available" }, { "frame": "01168", "timestamp_sec": 75.67, "offset_sec": 39, "rgb_path": "assets/cat_08/v_-0NfmccuL9Q_1099_2385/0039s_01168_rgb.jpg", "mask_path": "assets/cat_08/v_-0NfmccuL9Q_1099_2385/0039s_01168_mask.png", "mask_status": "available" }, { "frame": "01200", "timestamp_sec": 76.67, "offset_sec": 40, "rgb_path": "assets/cat_08/v_-0NfmccuL9Q_1099_2385/0040s_01200_rgb.jpg", "mask_path": "assets/cat_08/v_-0NfmccuL9Q_1099_2385/0040s_01200_mask.png", "mask_status": "available" }, { "frame": "01228", "timestamp_sec": 77.67, "offset_sec": 41, "rgb_path": "assets/cat_08/v_-0NfmccuL9Q_1099_2385/0041s_01228_rgb.jpg", "mask_path": "assets/cat_08/v_-0NfmccuL9Q_1099_2385/0041s_01228_mask.png", "mask_status": "available" }, { "frame": "01260", "timestamp_sec": 78.67, "offset_sec": 42, "rgb_path": "assets/cat_08/v_-0NfmccuL9Q_1099_2385/0042s_01260_rgb.jpg", "mask_path": "assets/cat_08/v_-0NfmccuL9Q_1099_2385/0042s_01260_mask.png", "mask_status": "available" } ], "obj_id": [ 0 ], "anno_id": [ 6 ], "fps_inferred_from_clip_interval": 29.9697 }, "usage": { "input_tokens": 228242, "input_tokens_details": { "cache_write_tokens": 228239, "cached_tokens": 0 }, "output_tokens": 5141, "output_tokens_details": { "reasoning_tokens": 2634 }, "total_tokens": 233383 }, "model": "gpt-5.6-sol", "effort": "medium", "custom_id": "cat08_cap03_v_-0NfmccuL9Q_1099_2385", "annotation": { "caption": "The person attaches a socket to the power tool and uses it to systematically remove all the lug nuts from the wheel. After placing the nuts and the tool on the ground, they grip the wheel with both hands and pull it off the car. They then roll the removed tire away to the right.", "entity_groups": [ { "id": "person_1", "label": "person", "spans": [ { "text": "The person", "range": [ 0, 10 ] }, { "text": "they", "range": [ 168, 172 ] }, { "text": "They", "range": [ 229, 233 ] } ], "source_mask_link": "selected_person" }, { "id": "socket_1", "label": "socket", "spans": [ { "text": "a socket", "range": [ 20, 28 ] } ], "source_mask_link": null }, { "id": "power_tool_1", "label": "power tool", "spans": [ { "text": "the power tool", "range": [ 32, 46 ] }, { "text": "it", "range": [ 56, 58 ] }, { "text": "the tool", "range": [ 144, 152 ] } ], "source_mask_link": null }, { "id": "lug_nuts_1", "label": "lug nuts", "spans": [ { "text": "all the lug nuts", "range": [ 84, 100 ] }, { "text": "the nuts", "range": [ 131, 139 ] } ], "source_mask_link": null }, { "id": "wheel_tire_1", "label": "wheel/tire", "spans": [ { "text": "the wheel", "range": [ 106, 115 ] }, { "text": "the wheel", "range": [ 178, 187 ] }, { "text": "it", "range": [ 213, 215 ] }, { "text": "the removed tire", "range": [ 244, 260 ] } ], "source_mask_link": null }, { "id": "ground_1", "label": "ground", "spans": [ { "text": "the ground", "range": [ 156, 166 ] } ], "source_mask_link": null }, { "id": "hands_1", "label": "hands", "spans": [ { "text": "both hands", "range": [ 193, 203 ] } ], "source_mask_link": null }, { "id": "car_1", "label": "car", "spans": [ { "text": "the car", "range": [ 220, 227 ] } ], "source_mask_link": null } ], "actions": [ { "id": "action_1", "mentions": [ { "text": "attaches", "ranges": [ [ 11, 19 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1, "ranges": [] } } ], "type": "simple", "subject_refs": [ "person_1" ], "object_refs": [ "socket_1", "power_tool_1" ], "confidence": "high", "temporal_status": "supported", "intervals": [ { "start_s": 4.0, "end_s": 7.0, "source_start_s": 40.67, "source_end_s": 43.67 } ], "evidence": "At 4–6 seconds, the masked person manipulates the socket at the front of the power tool before bringing the assembled tool to the wheel.", "mask_relation": "subject matches the supplied selected-person mask", "source_subject_refs": "person_1" }, { "id": "action_2", "mentions": [ { "text": "uses", "ranges": [ [ 51, 55 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1, "ranges": [] } } ], "type": "simple", "subject_refs": [ "person_1" ], "object_refs": [ "power_tool_1" ], "confidence": "high", "temporal_status": "supported", "intervals": [ { "start_s": 7.0, "end_s": 36.0, "source_start_s": 43.67, "source_end_s": 72.67 } ], "evidence": "Across samples 7–35 seconds, the masked person repeatedly holds and operates the power tool against the wheel's lug positions.", "mask_relation": "subject matches the supplied selected-person mask", "source_subject_refs": "person_1" }, { "id": "action_3", "mentions": [ { "text": "remove", "ranges": [ [ 77, 83 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1, "ranges": [] } } ], "type": "simple", "subject_refs": [ "person_1" ], "object_refs": [ "lug_nuts_1" ], "confidence": "high", "temporal_status": "supported", "intervals": [ { "start_s": 7.0, "end_s": 36.0, "source_start_s": 43.67, "source_end_s": 72.67 } ], "evidence": "The masked person applies the socket tool sequentially to the lug positions, while an increasing number of removed lug nuts becomes visible on the ground.", "mask_relation": "subject matches the supplied selected-person mask", "source_subject_refs": "person_1" }, { "id": "action_4", "mentions": [ { "text": "placing", "ranges": [ [ 123, 130 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1, "ranges": [] } } ], "type": "simple", "subject_refs": [ "person_1" ], "object_refs": [ "lug_nuts_1", "power_tool_1", "ground_1" ], "confidence": "high", "temporal_status": "supported", "intervals": [ { "start_s": 37.0, "end_s": 39.0, "source_start_s": 73.67, "source_end_s": 75.67 } ], "evidence": "At 37 seconds the masked person lowers the power tool; by 38 seconds it is on the ground beside the removed lug nuts.", "mask_relation": "subject matches the supplied selected-person mask", "source_subject_refs": "person_1" }, { "id": "action_5", "mentions": [ { "text": "grip", "ranges": [ [ 173, 177 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1, "ranges": [] } } ], "type": "simple", "subject_refs": [ "person_1" ], "object_refs": [ "wheel_tire_1", "hands_1" ], "confidence": "high", "temporal_status": "supported", "intervals": [ { "start_s": 39.0, "end_s": 42.0, "source_start_s": 75.67, "source_end_s": 78.67 } ], "evidence": "From 39 seconds onward, the masked person visibly holds opposite sides of the wheel with both hands.", "mask_relation": "subject matches the supplied selected-person mask", "source_subject_refs": "person_1" }, { "id": "action_6", "mentions": [ { "text": "pull", "ranges": [ [ 208, 212 ] ], "occurrence": 1, "particle": { "text": "off", "occurrence": 1, "ranges": [ [ 216, 219 ] ] } } ], "type": "phrasal", "subject_refs": [ "person_1" ], "object_refs": [ "wheel_tire_1", "car_1" ], "confidence": "high", "temporal_status": "supported", "intervals": [ { "start_s": 40.0, "end_s": 42.0, "source_start_s": 76.67, "source_end_s": 78.67 } ], "evidence": "At 40–41 seconds, the masked person draws the wheel outward; the exposed brake assembly confirms that the wheel has separated from the car.", "mask_relation": "subject matches the supplied selected-person mask", "source_subject_refs": "person_1" }, { "id": "action_7", "mentions": [ { "text": "roll", "ranges": [ [ 239, 243 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1, "ranges": [] } } ], "type": "simple", "subject_refs": [ "person_1" ], "object_refs": [ "wheel_tire_1" ], "confidence": "medium", "temporal_status": "sparsely supported at clip boundary", "intervals": [ { "start_s": 42.0, "end_s": 42.91, "source_start_s": 78.67, "source_end_s": 79.58 } ], "evidence": "In the final sampled frame, the masked person has the detached wheel upright and is moving it toward the right; the clip ends immediately afterward.", "mask_relation": "subject matches the supplied selected-person mask", "source_subject_refs": "person_1" } ], "notes": [ "The supplied masklet consistently identifies the single person described by “The person,” “they,” and “They.”", "Only the person has a supplied source mask. The socket, power tool, lug nuts, wheel/tire, ground, hands, and car are text-groundable entities but have no separate supplied masks.", "“the wheel,” the pronominal “it,” and “the removed tire” are grouped as the same wheel-and-tire assembly.", "The final rolling action has only sparse support in the last sampled frame, so its interval ends exactly at the clip boundary of 42.910 seconds." ] }, "model_output_raw": { "caption": "The person attaches a socket to the power tool and uses it to systematically remove all the lug nuts from the wheel. After placing the nuts and the tool on the ground, they grip the wheel with both hands and pull it off the car. They then roll the removed tire away to the right.", "entity_groups": [ { "id": "person_1", "label": "person", "spans": [ { "text": "The person", "range": [ 0, 10 ] }, { "text": "they", "range": [ 168, 172 ] }, { "text": "They", "range": [ 229, 233 ] } ], "source_mask": "selected_person_masklet" }, { "id": "socket_1", "label": "socket", "spans": [ { "text": "a socket", "range": [ 20, 28 ] } ], "source_mask": null }, { "id": "power_tool_1", "label": "power tool", "spans": [ { "text": "the power tool", "range": [ 32, 46 ] }, { "text": "it", "range": [ 56, 58 ] }, { "text": "the tool", "range": [ 144, 152 ] } ], "source_mask": null }, { "id": "lug_nuts_1", "label": "lug nuts", "spans": [ { "text": "all the lug nuts", "range": [ 84, 100 ] }, { "text": "the nuts", "range": [ 131, 139 ] } ], "source_mask": null }, { "id": "wheel_tire_1", "label": "wheel/tire", "spans": [ { "text": "the wheel", "range": [ 106, 115 ] }, { "text": "the wheel", "range": [ 178, 187 ] }, { "text": "it", "range": [ 213, 215 ] }, { "text": "the removed tire", "range": [ 244, 260 ] } ], "source_mask": null }, { "id": "ground_1", "label": "ground", "spans": [ { "text": "the ground", "range": [ 156, 166 ] } ], "source_mask": null }, { "id": "hands_1", "label": "hands", "spans": [ { "text": "both hands", "range": [ 193, 203 ] } ], "source_mask": null }, { "id": "car_1", "label": "car", "spans": [ { "text": "the car", "range": [ 220, 227 ] } ], "source_mask": null } ], "actions": [ { "id": "action_1", "mentions": [ { "text": "attaches", "occurrence": 1, "ranges": [ [ 11, 19 ] ], "particle": { "text": null, "occurrence": 1, "ranges": [] } } ], "type": "simple", "subject": "person_1", "objects": [ "socket_1", "power_tool_1" ], "confidence": "high", "mask_relation": "subject matches the supplied selected-person mask", "temporal_status": "supported", "intervals": [ { "start_s": 4.0, "end_s": 7.0, "source_start_s": 40.67, "source_end_s": 43.67 } ], "evidence": "At 4–6 seconds, the masked person manipulates the socket at the front of the power tool before bringing the assembled tool to the wheel." }, { "id": "action_2", "mentions": [ { "text": "uses", "occurrence": 1, "ranges": [ [ 51, 55 ] ], "particle": { "text": null, "occurrence": 1, "ranges": [] } } ], "type": "simple", "subject": "person_1", "objects": [ "power_tool_1" ], "confidence": "high", "mask_relation": "subject matches the supplied selected-person mask", "temporal_status": "supported", "intervals": [ { "start_s": 7.0, "end_s": 36.0, "source_start_s": 43.67, "source_end_s": 72.67 } ], "evidence": "Across samples 7–35 seconds, the masked person repeatedly holds and operates the power tool against the wheel's lug positions." }, { "id": "action_3", "mentions": [ { "text": "remove", "occurrence": 1, "ranges": [ [ 77, 83 ] ], "particle": { "text": null, "occurrence": 1, "ranges": [] } } ], "type": "simple", "subject": "person_1", "objects": [ "lug_nuts_1" ], "confidence": "high", "mask_relation": "subject matches the supplied selected-person mask", "temporal_status": "supported", "intervals": [ { "start_s": 7.0, "end_s": 36.0, "source_start_s": 43.67, "source_end_s": 72.67 } ], "evidence": "The masked person applies the socket tool sequentially to the lug positions, while an increasing number of removed lug nuts becomes visible on the ground." }, { "id": "action_4", "mentions": [ { "text": "placing", "occurrence": 1, "ranges": [ [ 123, 130 ] ], "particle": { "text": null, "occurrence": 1, "ranges": [] } } ], "type": "simple", "subject": "person_1", "objects": [ "lug_nuts_1", "power_tool_1", "ground_1" ], "confidence": "high", "mask_relation": "subject matches the supplied selected-person mask", "temporal_status": "supported", "intervals": [ { "start_s": 37.0, "end_s": 39.0, "source_start_s": 73.67, "source_end_s": 75.67 } ], "evidence": "At 37 seconds the masked person lowers the power tool; by 38 seconds it is on the ground beside the removed lug nuts." }, { "id": "action_5", "mentions": [ { "text": "grip", "occurrence": 1, "ranges": [ [ 173, 177 ] ], "particle": { "text": null, "occurrence": 1, "ranges": [] } } ], "type": "simple", "subject": "person_1", "objects": [ "wheel_tire_1", "hands_1" ], "confidence": "high", "mask_relation": "subject matches the supplied selected-person mask", "temporal_status": "supported", "intervals": [ { "start_s": 39.0, "end_s": 42.0, "source_start_s": 75.67, "source_end_s": 78.67 } ], "evidence": "From 39 seconds onward, the masked person visibly holds opposite sides of the wheel with both hands." }, { "id": "action_6", "mentions": [ { "text": "pull", "occurrence": 1, "ranges": [ [ 208, 212 ] ], "particle": { "text": "off", "occurrence": 1, "ranges": [ [ 216, 219 ] ] } } ], "type": "phrasal", "subject": "person_1", "objects": [ "wheel_tire_1", "car_1" ], "confidence": "high", "mask_relation": "subject matches the supplied selected-person mask", "temporal_status": "supported", "intervals": [ { "start_s": 40.0, "end_s": 42.0, "source_start_s": 76.67, "source_end_s": 78.67 } ], "evidence": "At 40–41 seconds, the masked person draws the wheel outward; the exposed brake assembly confirms that the wheel has separated from the car." }, { "id": "action_7", "mentions": [ { "text": "roll", "occurrence": 1, "ranges": [ [ 239, 243 ] ], "particle": { "text": null, "occurrence": 1, "ranges": [] } } ], "type": "simple", "subject": "person_1", "objects": [ "wheel_tire_1" ], "confidence": "medium", "mask_relation": "subject matches the supplied selected-person mask", "temporal_status": "sparsely supported at clip boundary", "intervals": [ { "start_s": 42.0, "end_s": 42.91, "source_start_s": 78.67, "source_end_s": 79.58 } ], "evidence": "In the final sampled frame, the masked person has the detached wheel upright and is moving it toward the right; the clip ends immediately afterward." } ], "notes": [ "The supplied masklet consistently identifies the single person described by “The person,” “they,” and “They.”", "Only the person has a supplied source mask. The socket, power tool, lug nuts, wheel/tire, ground, hands, and car are text-groundable entities but have no separate supplied masks.", "“the wheel,” the pronominal “it,” and “the removed tire” are grouped as the same wheel-and-tire assembly.", "The final rolling action has only sparse support in the last sampled frame, so its interval ends exactly at the clip boundary of 42.910 seconds." ] }, "normalization_repairs": 0 }, { "source_video_id": "v_-0NfmccuL9Q", "caption_count_category": 8, "clip": { "caption_index": 4, "caption": "A person wearing a black jacket, blue jeans, and tan gloves carries a tire toward a jacked-up car. They bend over, align the tire with the wheel hub, and carefully lift it into place.", "source_interval_sec": [ 88.12, 99.62 ], "loom_clip_id": "v_-0NfmccuL9Q_2641_2986", "clip_frame_interval_4fps": [ 2641, 2986 ], "sample_count_1fps": 12, "sampled_frames": [ { "frame": "00000", "timestamp_sec": 88.12, "offset_sec": 0, "rgb_path": "assets/cat_08/v_-0NfmccuL9Q_2641_2986/0000s_00000_rgb.jpg", "mask_path": "assets/cat_08/v_-0NfmccuL9Q_2641_2986/0000s_00000_mask.png", "mask_status": "available" }, { "frame": "00028", "timestamp_sec": 89.12, "offset_sec": 1, "rgb_path": "assets/cat_08/v_-0NfmccuL9Q_2641_2986/0001s_00028_rgb.jpg", "mask_path": "assets/cat_08/v_-0NfmccuL9Q_2641_2986/0001s_00028_mask.png", "mask_status": "available" }, { "frame": "00060", "timestamp_sec": 90.12, "offset_sec": 2, "rgb_path": "assets/cat_08/v_-0NfmccuL9Q_2641_2986/0002s_00060_rgb.jpg", "mask_path": "assets/cat_08/v_-0NfmccuL9Q_2641_2986/0002s_00060_mask.png", "mask_status": "available" }, { "frame": "00088", "timestamp_sec": 91.12, "offset_sec": 3, "rgb_path": "assets/cat_08/v_-0NfmccuL9Q_2641_2986/0003s_00088_rgb.jpg", "mask_path": "assets/cat_08/v_-0NfmccuL9Q_2641_2986/0003s_00088_mask.png", "mask_status": "available" }, { "frame": "00120", "timestamp_sec": 92.12, "offset_sec": 4, "rgb_path": "assets/cat_08/v_-0NfmccuL9Q_2641_2986/0004s_00120_rgb.jpg", "mask_path": "assets/cat_08/v_-0NfmccuL9Q_2641_2986/0004s_00120_mask.png", "mask_status": "available" }, { "frame": "00148", "timestamp_sec": 93.12, "offset_sec": 5, "rgb_path": "assets/cat_08/v_-0NfmccuL9Q_2641_2986/0005s_00148_rgb.jpg", "mask_path": "assets/cat_08/v_-0NfmccuL9Q_2641_2986/0005s_00148_mask.png", "mask_status": "available" }, { "frame": "00180", "timestamp_sec": 94.12, "offset_sec": 6, "rgb_path": "assets/cat_08/v_-0NfmccuL9Q_2641_2986/0006s_00180_rgb.jpg", "mask_path": "assets/cat_08/v_-0NfmccuL9Q_2641_2986/0006s_00180_mask.png", "mask_status": "available" }, { "frame": "00208", "timestamp_sec": 95.12, "offset_sec": 7, "rgb_path": "assets/cat_08/v_-0NfmccuL9Q_2641_2986/0007s_00208_rgb.jpg", "mask_path": "assets/cat_08/v_-0NfmccuL9Q_2641_2986/0007s_00208_mask.png", "mask_status": "available" }, { "frame": "00240", "timestamp_sec": 96.12, "offset_sec": 8, "rgb_path": "assets/cat_08/v_-0NfmccuL9Q_2641_2986/0008s_00240_rgb.jpg", "mask_path": "assets/cat_08/v_-0NfmccuL9Q_2641_2986/0008s_00240_mask.png", "mask_status": "available" }, { "frame": "00268", "timestamp_sec": 97.12, "offset_sec": 9, "rgb_path": "assets/cat_08/v_-0NfmccuL9Q_2641_2986/0009s_00268_rgb.jpg", "mask_path": "assets/cat_08/v_-0NfmccuL9Q_2641_2986/0009s_00268_mask.png", "mask_status": "available" }, { "frame": "00300", "timestamp_sec": 98.12, "offset_sec": 10, "rgb_path": "assets/cat_08/v_-0NfmccuL9Q_2641_2986/0010s_00300_rgb.jpg", "mask_path": "assets/cat_08/v_-0NfmccuL9Q_2641_2986/0010s_00300_mask.png", "mask_status": "available" }, { "frame": "00328", "timestamp_sec": 99.12, "offset_sec": 11, "rgb_path": "assets/cat_08/v_-0NfmccuL9Q_2641_2986/0011s_00328_rgb.jpg", "mask_path": "assets/cat_08/v_-0NfmccuL9Q_2641_2986/0011s_00328_mask.png", "mask_status": "available" } ], "obj_id": [ 0 ], "anno_id": [ 8 ], "fps_inferred_from_clip_interval": 30.0 }, "usage": { "input_tokens": 74744, "input_tokens_details": { "cache_write_tokens": 74741, "cached_tokens": 0 }, "output_tokens": 4419, "output_tokens_details": { "reasoning_tokens": 2588 }, "total_tokens": 79163 }, "model": "gpt-5.6-sol", "effort": "medium", "custom_id": "cat08_cap04_v_-0NfmccuL9Q_2641_2986", "annotation": { "caption": "A person wearing a black jacket, blue jeans, and tan gloves carries a tire toward a jacked-up car. They bend over, align the tire with the wheel hub, and carefully lift it into place.", "entity_groups": [ { "id": "person_1", "label": "person", "spans": [ { "text": "A person", "range": [ 0, 8 ] }, { "text": "They", "range": [ 99, 103 ] } ], "source_mask_link": "selected_person" }, { "id": "jacket_1", "label": "jacket", "spans": [ { "text": "a black jacket", "range": [ 17, 31 ] } ], "source_mask_link": null }, { "id": "jeans_1", "label": "jeans", "spans": [ { "text": "blue jeans", "range": [ 33, 43 ] } ], "source_mask_link": null }, { "id": "gloves_1", "label": "gloves", "spans": [ { "text": "tan gloves", "range": [ 49, 59 ] } ], "source_mask_link": null }, { "id": "tire_1", "label": "tire", "spans": [ { "text": "a tire", "range": [ 68, 74 ] }, { "text": "the tire", "range": [ 121, 129 ] }, { "text": "it", "range": [ 169, 171 ] } ], "source_mask_link": null }, { "id": "car_1", "label": "car", "spans": [ { "text": "a jacked-up car", "range": [ 82, 97 ] } ], "source_mask_link": null }, { "id": "wheel_hub_1", "label": "wheel hub", "spans": [ { "text": "the wheel hub", "range": [ 135, 148 ] } ], "source_mask_link": null } ], "actions": [ { "id": "action_wearing_1", "mentions": [ { "text": "wearing", "ranges": [ [ 9, 16 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "type": "simple", "subject_refs": [ "person_1" ], "object_refs": [ "jacket_1", "jeans_1", "gloves_1" ], "confidence": "high", "temporal_status": "localized_sparse", "intervals": [ { "start_s": 1.0, "end_s": 11.5, "source_start_s": 89.12, "source_end_s": 99.62 } ], "evidence": "From 1.000 seconds through the clip boundary, the masked person is visibly dressed in the black jacket, blue jeans, and tan gloves.", "mask_relation": "subject_matches_selected_person_mask", "source_subject_refs": "person_1" }, { "id": "action_carries_1", "mentions": [ { "text": "carries", "ranges": [ [ 60, 67 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "type": "simple", "subject_refs": [ "person_1" ], "object_refs": [ "tire_1" ], "confidence": "high", "temporal_status": "localized_sparse", "intervals": [ { "start_s": 1.0, "end_s": 3.0, "source_start_s": 89.12, "source_end_s": 91.12 } ], "evidence": "At 1.000 and 2.000 seconds, the masked person holds and advances the tire toward the raised car.", "mask_relation": "subject_matches_selected_person_mask", "source_subject_refs": "person_1" }, { "id": "action_bend_over_1", "mentions": [ { "text": "bend over", "ranges": [ [ 104, 113 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "type": "phrasal", "subject_refs": [ "person_1" ], "object_refs": [], "confidence": "high", "temporal_status": "localized_sparse", "intervals": [ { "start_s": 4.0, "end_s": 11.5, "source_start_s": 92.12, "source_end_s": 99.62 } ], "evidence": "At sampled times 4.000–11.000 seconds, the masked person remains visibly bent over the tire and wheel area.", "mask_relation": "subject_matches_selected_person_mask", "source_subject_refs": "person_1" }, { "id": "action_align_1", "mentions": [ { "text": "align", "ranges": [ [ 115, 120 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "type": "simple", "subject_refs": [ "person_1" ], "object_refs": [ "tire_1" ], "confidence": "high", "temporal_status": "localized_sparse", "intervals": [ { "start_s": 3.0, "end_s": 11.5, "source_start_s": 91.12, "source_end_s": 99.62 } ], "evidence": "From 3.000 seconds onward, the masked person maneuvers and centers the tire against the exposed wheel hub.", "mask_relation": "subject_matches_selected_person_mask", "source_subject_refs": "person_1" }, { "id": "action_lift_1", "mentions": [ { "text": "lift", "ranges": [ [ 164, 168 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "type": "simple", "subject_refs": [ "person_1" ], "object_refs": [ "tire_1" ], "confidence": "high", "temporal_status": "localized_sparse", "intervals": [ { "start_s": 3.0, "end_s": 11.5, "source_start_s": 91.12, "source_end_s": 99.62 } ], "evidence": "Across samples 3.000–11.000 seconds, the masked person raises and seats the tire at hub height, with placement still supported at the clip boundary.", "mask_relation": "subject_matches_selected_person_mask", "source_subject_refs": "person_1" } ], "notes": [ "The supplied masklet identifies the selected person; it is empty at 0.000 seconds because the person is not yet visible and tracks the person from 1.000 seconds onward.", "No separate masks were supplied for the jacket, jeans, gloves, tire, car, or wheel hub, so those entity groups have null source-mask links.", "Action intervals are sparse sampled-frame support intervals rather than exact continuous-video onset and offset boundaries." ] }, "model_output_raw": { "caption": "A person wearing a black jacket, blue jeans, and tan gloves carries a tire toward a jacked-up car. They bend over, align the tire with the wheel hub, and carefully lift it into place.", "entity_groups": [ { "id": "person_1", "entity_type": "person", "spans": [ { "text": "A person", "range": [ 0, 8 ] }, { "text": "They", "range": [ 99, 103 ] } ], "source_mask_link": "selected_person_masklet" }, { "id": "jacket_1", "entity_type": "jacket", "spans": [ { "text": "a black jacket", "range": [ 17, 31 ] } ], "source_mask_link": null }, { "id": "jeans_1", "entity_type": "jeans", "spans": [ { "text": "blue jeans", "range": [ 33, 43 ] } ], "source_mask_link": null }, { "id": "gloves_1", "entity_type": "gloves", "spans": [ { "text": "tan gloves", "range": [ 49, 59 ] } ], "source_mask_link": null }, { "id": "tire_1", "entity_type": "tire", "spans": [ { "text": "a tire", "range": [ 68, 74 ] }, { "text": "the tire", "range": [ 121, 129 ] }, { "text": "it", "range": [ 169, 171 ] } ], "source_mask_link": null }, { "id": "car_1", "entity_type": "car", "spans": [ { "text": "a jacked-up car", "range": [ 82, 97 ] } ], "source_mask_link": null }, { "id": "wheel_hub_1", "entity_type": "wheel hub", "spans": [ { "text": "the wheel hub", "range": [ 135, 148 ] } ], "source_mask_link": null } ], "actions": [ { "id": "action_wearing_1", "mentions": [ { "text": "wearing", "ranges": [ [ 9, 16 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "type": "simple", "subject_ref": "person_1", "object_refs": [ "jacket_1", "jeans_1", "gloves_1" ], "confidence": "high", "mask_relation": "subject_matches_selected_person_mask", "temporal_status": "localized_sparse", "intervals": [ { "start_s": 1.0, "end_s": 11.5, "source_start_s": 89.12, "source_end_s": 99.62 } ], "evidence": "From 1.000 seconds through the clip boundary, the masked person is visibly dressed in the black jacket, blue jeans, and tan gloves." }, { "id": "action_carries_1", "mentions": [ { "text": "carries", "ranges": [ [ 60, 67 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "type": "simple", "subject_ref": "person_1", "object_refs": [ "tire_1" ], "target_refs": [ "car_1" ], "confidence": "high", "mask_relation": "subject_matches_selected_person_mask", "temporal_status": "localized_sparse", "intervals": [ { "start_s": 1.0, "end_s": 3.0, "source_start_s": 89.12, "source_end_s": 91.12 } ], "evidence": "At 1.000 and 2.000 seconds, the masked person holds and advances the tire toward the raised car." }, { "id": "action_bend_over_1", "mentions": [ { "text": "bend over", "ranges": [ [ 104, 113 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "type": "phrasal", "subject_ref": "person_1", "object_refs": [], "confidence": "high", "mask_relation": "subject_matches_selected_person_mask", "temporal_status": "localized_sparse", "intervals": [ { "start_s": 4.0, "end_s": 11.5, "source_start_s": 92.12, "source_end_s": 99.62 } ], "evidence": "At sampled times 4.000–11.000 seconds, the masked person remains visibly bent over the tire and wheel area." }, { "id": "action_align_1", "mentions": [ { "text": "align", "ranges": [ [ 115, 120 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "type": "simple", "subject_ref": "person_1", "object_refs": [ "tire_1" ], "target_refs": [ "wheel_hub_1" ], "confidence": "high", "mask_relation": "subject_matches_selected_person_mask", "temporal_status": "localized_sparse", "intervals": [ { "start_s": 3.0, "end_s": 11.5, "source_start_s": 91.12, "source_end_s": 99.62 } ], "evidence": "From 3.000 seconds onward, the masked person maneuvers and centers the tire against the exposed wheel hub." }, { "id": "action_lift_1", "mentions": [ { "text": "lift", "ranges": [ [ 164, 168 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "type": "simple", "subject_ref": "person_1", "object_refs": [ "tire_1" ], "confidence": "high", "mask_relation": "subject_matches_selected_person_mask", "temporal_status": "localized_sparse", "intervals": [ { "start_s": 3.0, "end_s": 11.5, "source_start_s": 91.12, "source_end_s": 99.62 } ], "evidence": "Across samples 3.000–11.000 seconds, the masked person raises and seats the tire at hub height, with placement still supported at the clip boundary." } ], "notes": [ "The supplied masklet identifies the selected person; it is empty at 0.000 seconds because the person is not yet visible and tracks the person from 1.000 seconds onward.", "No separate masks were supplied for the jacket, jeans, gloves, tire, car, or wheel hub, so those entity groups have null source-mask links.", "Action intervals are sparse sampled-frame support intervals rather than exact continuous-video onset and offset boundaries." ] }, "normalization_repairs": 0 }, { "source_video_id": "v_-0NfmccuL9Q", "caption_count_category": 8, "clip": { "caption_index": 5, "caption": "Squatting beside the newly mounted tire, the person picks up individual lug nuts from the ground. They screw each nut onto the studs by hand, occasionally using a small cloth to wipe the area. After securing all the nuts, the person removes their gloves.", "source_interval_sec": [ 99.62, 148.12 ], "loom_clip_id": "v_-0NfmccuL9Q_2986_4439", "clip_frame_interval_4fps": [ 2986, 4439 ], "sample_count_1fps": 49, "sampled_frames": [ { "frame": "00000", "timestamp_sec": 99.62, "offset_sec": 0, "rgb_path": "assets/cat_08/v_-0NfmccuL9Q_2986_4439/0000s_00000_rgb.jpg", "mask_path": "assets/cat_08/v_-0NfmccuL9Q_2986_4439/0000s_00000_mask.png", "mask_status": "available" }, { "frame": "00028", "timestamp_sec": 100.62, "offset_sec": 1, "rgb_path": "assets/cat_08/v_-0NfmccuL9Q_2986_4439/0001s_00028_rgb.jpg", "mask_path": "assets/cat_08/v_-0NfmccuL9Q_2986_4439/0001s_00028_mask.png", "mask_status": "available" }, { "frame": "00060", "timestamp_sec": 101.62, "offset_sec": 2, "rgb_path": "assets/cat_08/v_-0NfmccuL9Q_2986_4439/0002s_00060_rgb.jpg", "mask_path": "assets/cat_08/v_-0NfmccuL9Q_2986_4439/0002s_00060_mask.png", "mask_status": "available" }, { "frame": "00088", "timestamp_sec": 102.62, "offset_sec": 3, "rgb_path": "assets/cat_08/v_-0NfmccuL9Q_2986_4439/0003s_00088_rgb.jpg", "mask_path": "assets/cat_08/v_-0NfmccuL9Q_2986_4439/0003s_00088_mask.png", "mask_status": "available" }, { "frame": "00120", "timestamp_sec": 103.62, "offset_sec": 4, "rgb_path": "assets/cat_08/v_-0NfmccuL9Q_2986_4439/0004s_00120_rgb.jpg", "mask_path": "assets/cat_08/v_-0NfmccuL9Q_2986_4439/0004s_00120_mask.png", "mask_status": "available" }, { "frame": "00148", "timestamp_sec": 104.62, "offset_sec": 5, "rgb_path": "assets/cat_08/v_-0NfmccuL9Q_2986_4439/0005s_00148_rgb.jpg", "mask_path": "assets/cat_08/v_-0NfmccuL9Q_2986_4439/0005s_00148_mask.png", "mask_status": "available" }, { "frame": "00180", "timestamp_sec": 105.62, "offset_sec": 6, "rgb_path": "assets/cat_08/v_-0NfmccuL9Q_2986_4439/0006s_00180_rgb.jpg", "mask_path": "assets/cat_08/v_-0NfmccuL9Q_2986_4439/0006s_00180_mask.png", "mask_status": "available" }, { "frame": "00208", "timestamp_sec": 106.62, "offset_sec": 7, "rgb_path": "assets/cat_08/v_-0NfmccuL9Q_2986_4439/0007s_00208_rgb.jpg", "mask_path": "assets/cat_08/v_-0NfmccuL9Q_2986_4439/0007s_00208_mask.png", "mask_status": "available" }, { "frame": "00240", "timestamp_sec": 107.62, "offset_sec": 8, "rgb_path": "assets/cat_08/v_-0NfmccuL9Q_2986_4439/0008s_00240_rgb.jpg", "mask_path": "assets/cat_08/v_-0NfmccuL9Q_2986_4439/0008s_00240_mask.png", "mask_status": "available" }, { "frame": "00268", "timestamp_sec": 108.62, "offset_sec": 9, "rgb_path": "assets/cat_08/v_-0NfmccuL9Q_2986_4439/0009s_00268_rgb.jpg", "mask_path": "assets/cat_08/v_-0NfmccuL9Q_2986_4439/0009s_00268_mask.png", "mask_status": "available" }, { "frame": "00300", "timestamp_sec": 109.62, "offset_sec": 10, "rgb_path": "assets/cat_08/v_-0NfmccuL9Q_2986_4439/0010s_00300_rgb.jpg", "mask_path": "assets/cat_08/v_-0NfmccuL9Q_2986_4439/0010s_00300_mask.png", "mask_status": "available" }, { "frame": "00328", "timestamp_sec": 110.62, "offset_sec": 11, "rgb_path": "assets/cat_08/v_-0NfmccuL9Q_2986_4439/0011s_00328_rgb.jpg", "mask_path": "assets/cat_08/v_-0NfmccuL9Q_2986_4439/0011s_00328_mask.png", "mask_status": "available" }, { "frame": "00360", "timestamp_sec": 111.62, "offset_sec": 12, "rgb_path": "assets/cat_08/v_-0NfmccuL9Q_2986_4439/0012s_00360_rgb.jpg", "mask_path": "assets/cat_08/v_-0NfmccuL9Q_2986_4439/0012s_00360_mask.png", "mask_status": "available" }, { "frame": "00388", "timestamp_sec": 112.62, "offset_sec": 13, "rgb_path": "assets/cat_08/v_-0NfmccuL9Q_2986_4439/0013s_00388_rgb.jpg", "mask_path": "assets/cat_08/v_-0NfmccuL9Q_2986_4439/0013s_00388_mask.png", "mask_status": "available" }, { "frame": "00420", "timestamp_sec": 113.62, "offset_sec": 14, "rgb_path": "assets/cat_08/v_-0NfmccuL9Q_2986_4439/0014s_00420_rgb.jpg", "mask_path": "assets/cat_08/v_-0NfmccuL9Q_2986_4439/0014s_00420_mask.png", "mask_status": "available" }, { "frame": "00448", "timestamp_sec": 114.62, "offset_sec": 15, "rgb_path": "assets/cat_08/v_-0NfmccuL9Q_2986_4439/0015s_00448_rgb.jpg", "mask_path": "assets/cat_08/v_-0NfmccuL9Q_2986_4439/0015s_00448_mask.png", "mask_status": "available" }, { "frame": "00480", "timestamp_sec": 115.62, "offset_sec": 16, "rgb_path": "assets/cat_08/v_-0NfmccuL9Q_2986_4439/0016s_00480_rgb.jpg", "mask_path": "assets/cat_08/v_-0NfmccuL9Q_2986_4439/0016s_00480_mask.png", "mask_status": "available" }, { "frame": "00508", "timestamp_sec": 116.62, "offset_sec": 17, "rgb_path": "assets/cat_08/v_-0NfmccuL9Q_2986_4439/0017s_00508_rgb.jpg", "mask_path": "assets/cat_08/v_-0NfmccuL9Q_2986_4439/0017s_00508_mask.png", "mask_status": "available" }, { "frame": "00540", "timestamp_sec": 117.62, "offset_sec": 18, "rgb_path": "assets/cat_08/v_-0NfmccuL9Q_2986_4439/0018s_00540_rgb.jpg", "mask_path": "assets/cat_08/v_-0NfmccuL9Q_2986_4439/0018s_00540_mask.png", "mask_status": "available" }, { "frame": "00568", "timestamp_sec": 118.62, "offset_sec": 19, "rgb_path": "assets/cat_08/v_-0NfmccuL9Q_2986_4439/0019s_00568_rgb.jpg", "mask_path": "assets/cat_08/v_-0NfmccuL9Q_2986_4439/0019s_00568_mask.png", "mask_status": "available" }, { "frame": "00600", "timestamp_sec": 119.62, "offset_sec": 20, "rgb_path": "assets/cat_08/v_-0NfmccuL9Q_2986_4439/0020s_00600_rgb.jpg", "mask_path": "assets/cat_08/v_-0NfmccuL9Q_2986_4439/0020s_00600_mask.png", "mask_status": "available" }, { "frame": "00628", "timestamp_sec": 120.62, "offset_sec": 21, "rgb_path": "assets/cat_08/v_-0NfmccuL9Q_2986_4439/0021s_00628_rgb.jpg", "mask_path": "assets/cat_08/v_-0NfmccuL9Q_2986_4439/0021s_00628_mask.png", "mask_status": "available" }, { "frame": "00660", "timestamp_sec": 121.62, "offset_sec": 22, "rgb_path": "assets/cat_08/v_-0NfmccuL9Q_2986_4439/0022s_00660_rgb.jpg", "mask_path": "assets/cat_08/v_-0NfmccuL9Q_2986_4439/0022s_00660_mask.png", "mask_status": "available" }, { "frame": "00688", "timestamp_sec": 122.62, "offset_sec": 23, "rgb_path": "assets/cat_08/v_-0NfmccuL9Q_2986_4439/0023s_00688_rgb.jpg", "mask_path": "assets/cat_08/v_-0NfmccuL9Q_2986_4439/0023s_00688_mask.png", "mask_status": "available" }, { "frame": "00720", "timestamp_sec": 123.62, "offset_sec": 24, "rgb_path": "assets/cat_08/v_-0NfmccuL9Q_2986_4439/0024s_00720_rgb.jpg", "mask_path": "assets/cat_08/v_-0NfmccuL9Q_2986_4439/0024s_00720_mask.png", "mask_status": "available" }, { "frame": "00748", "timestamp_sec": 124.62, "offset_sec": 25, "rgb_path": "assets/cat_08/v_-0NfmccuL9Q_2986_4439/0025s_00748_rgb.jpg", "mask_path": "assets/cat_08/v_-0NfmccuL9Q_2986_4439/0025s_00748_mask.png", "mask_status": "available" }, { "frame": "00780", "timestamp_sec": 125.62, "offset_sec": 26, "rgb_path": "assets/cat_08/v_-0NfmccuL9Q_2986_4439/0026s_00780_rgb.jpg", "mask_path": "assets/cat_08/v_-0NfmccuL9Q_2986_4439/0026s_00780_mask.png", "mask_status": "available" }, { "frame": "00808", "timestamp_sec": 126.62, "offset_sec": 27, "rgb_path": "assets/cat_08/v_-0NfmccuL9Q_2986_4439/0027s_00808_rgb.jpg", "mask_path": "assets/cat_08/v_-0NfmccuL9Q_2986_4439/0027s_00808_mask.png", "mask_status": "available" }, { "frame": "00840", "timestamp_sec": 127.62, "offset_sec": 28, "rgb_path": "assets/cat_08/v_-0NfmccuL9Q_2986_4439/0028s_00840_rgb.jpg", "mask_path": "assets/cat_08/v_-0NfmccuL9Q_2986_4439/0028s_00840_mask.png", "mask_status": "available" }, { "frame": "00868", "timestamp_sec": 128.62, "offset_sec": 29, "rgb_path": "assets/cat_08/v_-0NfmccuL9Q_2986_4439/0029s_00868_rgb.jpg", "mask_path": "assets/cat_08/v_-0NfmccuL9Q_2986_4439/0029s_00868_mask.png", "mask_status": "available" }, { "frame": "00900", "timestamp_sec": 129.62, "offset_sec": 30, "rgb_path": "assets/cat_08/v_-0NfmccuL9Q_2986_4439/0030s_00900_rgb.jpg", "mask_path": "assets/cat_08/v_-0NfmccuL9Q_2986_4439/0030s_00900_mask.png", "mask_status": "available" }, { "frame": "00928", "timestamp_sec": 130.62, "offset_sec": 31, "rgb_path": "assets/cat_08/v_-0NfmccuL9Q_2986_4439/0031s_00928_rgb.jpg", "mask_path": "assets/cat_08/v_-0NfmccuL9Q_2986_4439/0031s_00928_mask.png", "mask_status": "available" }, { "frame": "00960", "timestamp_sec": 131.62, "offset_sec": 32, "rgb_path": "assets/cat_08/v_-0NfmccuL9Q_2986_4439/0032s_00960_rgb.jpg", "mask_path": "assets/cat_08/v_-0NfmccuL9Q_2986_4439/0032s_00960_mask.png", "mask_status": "available" }, { "frame": "00988", "timestamp_sec": 132.62, "offset_sec": 33, "rgb_path": "assets/cat_08/v_-0NfmccuL9Q_2986_4439/0033s_00988_rgb.jpg", "mask_path": "assets/cat_08/v_-0NfmccuL9Q_2986_4439/0033s_00988_mask.png", "mask_status": "available" }, { "frame": "01020", "timestamp_sec": 133.62, "offset_sec": 34, "rgb_path": "assets/cat_08/v_-0NfmccuL9Q_2986_4439/0034s_01020_rgb.jpg", "mask_path": "assets/cat_08/v_-0NfmccuL9Q_2986_4439/0034s_01020_mask.png", "mask_status": "available" }, { "frame": "01048", "timestamp_sec": 134.62, "offset_sec": 35, "rgb_path": "assets/cat_08/v_-0NfmccuL9Q_2986_4439/0035s_01048_rgb.jpg", "mask_path": "assets/cat_08/v_-0NfmccuL9Q_2986_4439/0035s_01048_mask.png", "mask_status": "available" }, { "frame": "01080", "timestamp_sec": 135.62, "offset_sec": 36, "rgb_path": "assets/cat_08/v_-0NfmccuL9Q_2986_4439/0036s_01080_rgb.jpg", "mask_path": "assets/cat_08/v_-0NfmccuL9Q_2986_4439/0036s_01080_mask.png", "mask_status": "available" }, { "frame": "01108", "timestamp_sec": 136.62, "offset_sec": 37, "rgb_path": "assets/cat_08/v_-0NfmccuL9Q_2986_4439/0037s_01108_rgb.jpg", "mask_path": "assets/cat_08/v_-0NfmccuL9Q_2986_4439/0037s_01108_mask.png", "mask_status": "available" }, { "frame": "01140", "timestamp_sec": 137.62, "offset_sec": 38, "rgb_path": "assets/cat_08/v_-0NfmccuL9Q_2986_4439/0038s_01140_rgb.jpg", "mask_path": "assets/cat_08/v_-0NfmccuL9Q_2986_4439/0038s_01140_mask.png", "mask_status": "available" }, { "frame": "01168", "timestamp_sec": 138.62, "offset_sec": 39, "rgb_path": "assets/cat_08/v_-0NfmccuL9Q_2986_4439/0039s_01168_rgb.jpg", "mask_path": "assets/cat_08/v_-0NfmccuL9Q_2986_4439/0039s_01168_mask.png", "mask_status": "available" }, { "frame": "01200", "timestamp_sec": 139.62, "offset_sec": 40, "rgb_path": "assets/cat_08/v_-0NfmccuL9Q_2986_4439/0040s_01200_rgb.jpg", "mask_path": "assets/cat_08/v_-0NfmccuL9Q_2986_4439/0040s_01200_mask.png", "mask_status": "available" }, { "frame": "01228", "timestamp_sec": 140.62, "offset_sec": 41, "rgb_path": "assets/cat_08/v_-0NfmccuL9Q_2986_4439/0041s_01228_rgb.jpg", "mask_path": "assets/cat_08/v_-0NfmccuL9Q_2986_4439/0041s_01228_mask.png", "mask_status": "available" }, { "frame": "01260", "timestamp_sec": 141.62, "offset_sec": 42, "rgb_path": "assets/cat_08/v_-0NfmccuL9Q_2986_4439/0042s_01260_rgb.jpg", "mask_path": "assets/cat_08/v_-0NfmccuL9Q_2986_4439/0042s_01260_mask.png", "mask_status": "available" }, { "frame": "01288", "timestamp_sec": 142.62, "offset_sec": 43, "rgb_path": "assets/cat_08/v_-0NfmccuL9Q_2986_4439/0043s_01288_rgb.jpg", "mask_path": "assets/cat_08/v_-0NfmccuL9Q_2986_4439/0043s_01288_mask.png", "mask_status": "available" }, { "frame": "01320", "timestamp_sec": 143.62, "offset_sec": 44, "rgb_path": "assets/cat_08/v_-0NfmccuL9Q_2986_4439/0044s_01320_rgb.jpg", "mask_path": "assets/cat_08/v_-0NfmccuL9Q_2986_4439/0044s_01320_mask.png", "mask_status": "available" }, { "frame": "01348", "timestamp_sec": 144.62, "offset_sec": 45, "rgb_path": "assets/cat_08/v_-0NfmccuL9Q_2986_4439/0045s_01348_rgb.jpg", "mask_path": "assets/cat_08/v_-0NfmccuL9Q_2986_4439/0045s_01348_mask.png", "mask_status": "available" }, { "frame": "01380", "timestamp_sec": 145.62, "offset_sec": 46, "rgb_path": "assets/cat_08/v_-0NfmccuL9Q_2986_4439/0046s_01380_rgb.jpg", "mask_path": "assets/cat_08/v_-0NfmccuL9Q_2986_4439/0046s_01380_mask.png", "mask_status": "available" }, { "frame": "01408", "timestamp_sec": 146.62, "offset_sec": 47, "rgb_path": "assets/cat_08/v_-0NfmccuL9Q_2986_4439/0047s_01408_rgb.jpg", "mask_path": "assets/cat_08/v_-0NfmccuL9Q_2986_4439/0047s_01408_mask.png", "mask_status": "available" }, { "frame": "01440", "timestamp_sec": 147.62, "offset_sec": 48, "rgb_path": "assets/cat_08/v_-0NfmccuL9Q_2986_4439/0048s_01440_rgb.jpg", "mask_path": "assets/cat_08/v_-0NfmccuL9Q_2986_4439/0048s_01440_mask.png", "mask_status": "available" } ], "obj_id": [ 0 ], "anno_id": [ 9 ], "fps_inferred_from_clip_interval": 29.9588 }, "usage": { "input_tokens": 257940, "input_tokens_details": { "cache_write_tokens": 257937, "cached_tokens": 0 }, "output_tokens": 5047, "output_tokens_details": { "reasoning_tokens": 2948 }, "total_tokens": 262987 }, "model": "gpt-5.6-sol", "effort": "medium", "custom_id": "cat08_cap05_v_-0NfmccuL9Q_2986_4439", "annotation": { "caption": "Squatting beside the newly mounted tire, the person picks up individual lug nuts from the ground. They screw each nut onto the studs by hand, occasionally using a small cloth to wipe the area. After securing all the nuts, the person removes their gloves.", "entity_groups": [ { "id": "E1", "label": "person", "spans": [ { "text": "the person", "range": [ 41, 51 ] }, { "text": "They", "range": [ 98, 102 ] }, { "text": "the person", "range": [ 222, 232 ] }, { "text": "their", "range": [ 241, 246 ] } ], "source_mask_link": "selected_person" }, { "id": "E2", "label": "tire", "spans": [ { "text": "the newly mounted tire", "range": [ 17, 39 ] } ], "source_mask_link": null }, { "id": "E3", "label": "lug nuts", "spans": [ { "text": "individual lug nuts", "range": [ 61, 80 ] }, { "text": "each nut", "range": [ 109, 117 ] }, { "text": "all the nuts", "range": [ 208, 220 ] } ], "source_mask_link": null }, { "id": "E4", "label": "ground", "spans": [ { "text": "the ground", "range": [ 86, 96 ] } ], "source_mask_link": null }, { "id": "E5", "label": "studs", "spans": [ { "text": "the studs", "range": [ 123, 132 ] } ], "source_mask_link": null }, { "id": "E6", "label": "hand", "spans": [ { "text": "hand", "range": [ 136, 140 ] } ], "source_mask_link": null }, { "id": "E7", "label": "cloth", "spans": [ { "text": "a small cloth", "range": [ 161, 174 ] } ], "source_mask_link": null }, { "id": "E8", "label": "gloves", "spans": [ { "text": "their gloves", "range": [ 241, 253 ] } ], "source_mask_link": null } ], "actions": [ { "id": "A1", "mentions": [ { "text": "Squatting", "ranges": [ [ 0, 9 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "type": "posture", "subject_refs": [ "E1" ], "object_refs": [], "confidence": "high", "temporal_status": "localized", "intervals": [ { "start_s": 0.0, "end_s": 48.5, "source_start_s": 99.62, "source_end_s": 148.12 } ], "evidence": "The mask-linked person remains in a low squatting posture beside the wheel throughout all sampled frames, including the final sample.", "mask_relation": "The action subject matches the supplied selected-person mask.", "source_subject_refs": [ "E1" ] }, { "id": "A2", "mentions": [ { "text": "picks up", "ranges": [ [ 52, 60 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "type": "phrasal", "subject_refs": [ "E1" ], "object_refs": [ "E3" ], "confidence": "high", "temporal_status": "localized", "intervals": [ { "start_s": 2.0, "end_s": 5.0, "source_start_s": 101.62, "source_end_s": 104.62 } ], "evidence": "At 2–4 seconds, the mask-linked person reaches down toward the loose hardware and brings a small item back toward the wheel.", "mask_relation": "The action subject matches the supplied selected-person mask.", "source_subject_refs": [ "E1" ] }, { "id": "A3", "mentions": [ { "text": "screw", "ranges": [ [ 103, 108 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } }, { "text": "securing", "ranges": [ [ 199, 207 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "type": "simple", "subject_refs": [ "E1" ], "object_refs": [ "E3", "E5" ], "confidence": "high", "temporal_status": "localized", "intervals": [ { "start_s": 0.0, "end_s": 2.0, "source_start_s": 99.62, "source_end_s": 101.62 }, { "start_s": 4.0, "end_s": 15.0, "source_start_s": 103.62, "source_end_s": 114.62 }, { "start_s": 16.0, "end_s": 22.0, "source_start_s": 115.62, "source_end_s": 121.62 }, { "start_s": 24.0, "end_s": 30.0, "source_start_s": 123.62, "source_end_s": 129.62 }, { "start_s": 31.0, "end_s": 36.0, "source_start_s": 130.62, "source_end_s": 135.62 }, { "start_s": 37.0, "end_s": 45.0, "source_start_s": 136.62, "source_end_s": 144.62 } ], "evidence": "Across these sampled intervals, the mask-linked person repeatedly holds and rotates small lug nuts at the wheel studs by hand. “securing” summarizes the same fastening event.", "mask_relation": "The action subject matches the supplied selected-person mask.", "source_subject_refs": [ "E1" ] }, { "id": "A4", "mentions": [ { "text": "using", "ranges": [ [ 155, 160 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "type": "simple", "subject_refs": [ "E1" ], "object_refs": [ "E7" ], "confidence": "medium", "temporal_status": "localized", "intervals": [ { "start_s": 16.0, "end_s": 23.0, "source_start_s": 115.62, "source_end_s": 122.62 } ], "evidence": "A small orange cloth is visible in the mask-linked person's hands against the wheel area during the supported samples.", "mask_relation": "The action subject matches the supplied selected-person mask.", "source_subject_refs": [ "E1" ] }, { "id": "A5", "mentions": [ { "text": "wipe", "ranges": [ [ 178, 182 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "type": "simple", "subject_refs": [ "E1" ], "object_refs": [], "confidence": "medium", "temporal_status": "localized", "intervals": [ { "start_s": 16.0, "end_s": 23.0, "source_start_s": 115.62, "source_end_s": 122.62 } ], "evidence": "During 16–22 seconds, the cloth is held against and moved around the wheel's lug area; sparse frames support wiping but do not reveal exact stroke boundaries.", "mask_relation": "The action subject matches the supplied selected-person mask.", "source_subject_refs": [ "E1" ] }, { "id": "A6", "mentions": [ { "text": "removes", "ranges": [ [ 233, 240 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "type": "simple", "subject_refs": [ "E1" ], "object_refs": [ "E8" ], "confidence": "high", "temporal_status": "localized", "intervals": [ { "start_s": 46.0, "end_s": 48.5, "source_start_s": 145.62, "source_end_s": 148.12 } ], "evidence": "At 46–48 seconds, the mask-linked person visibly pulls the gloves from the hands; the removal continues through the clip's right boundary.", "mask_relation": "The action subject matches the supplied selected-person mask.", "source_subject_refs": [ "E1" ] } ], "notes": [ "The supplied masklet consistently identifies the captioned person, so only that entity group is linked to a source mask.", "No separate masks were supplied for the tire, lug nuts, ground, studs, hand, cloth, or gloves.", "The result-state premodifier “mounted” was not annotated as an action. “area” was excluded as an abstract spatial reference.", "Action intervals are sparse clip-relative visual-support intervals derived from the sampled clean RGB frames, not exact continuous-video boundaries." ] }, "model_output_raw": { "caption": "Squatting beside the newly mounted tire, the person picks up individual lug nuts from the ground. They screw each nut onto the studs by hand, occasionally using a small cloth to wipe the area. After securing all the nuts, the person removes their gloves.", "entity_groups": [ { "id": "E1", "entity": "person", "spans": [ { "text": "the person", "range": [ 41, 51 ] }, { "text": "They", "range": [ 98, 102 ] }, { "text": "the person", "range": [ 222, 232 ] }, { "text": "their", "range": [ 241, 246 ] } ], "source_mask": "selected_person_masklet" }, { "id": "E2", "entity": "tire", "spans": [ { "text": "the newly mounted tire", "range": [ 17, 39 ] } ], "source_mask": null }, { "id": "E3", "entity": "lug nuts", "spans": [ { "text": "individual lug nuts", "range": [ 61, 80 ] }, { "text": "each nut", "range": [ 109, 117 ] }, { "text": "all the nuts", "range": [ 208, 220 ] } ], "source_mask": null }, { "id": "E4", "entity": "ground", "spans": [ { "text": "the ground", "range": [ 86, 96 ] } ], "source_mask": null }, { "id": "E5", "entity": "studs", "spans": [ { "text": "the studs", "range": [ 123, 132 ] } ], "source_mask": null }, { "id": "E6", "entity": "hand", "spans": [ { "text": "hand", "range": [ 136, 140 ] } ], "source_mask": null }, { "id": "E7", "entity": "cloth", "spans": [ { "text": "a small cloth", "range": [ 161, 174 ] } ], "source_mask": null }, { "id": "E8", "entity": "gloves", "spans": [ { "text": "their gloves", "range": [ 241, 253 ] } ], "source_mask": null } ], "actions": [ { "id": "A1", "mentions": [ { "text": "Squatting", "ranges": [ [ 0, 9 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "type": "posture", "subject_refs": [ "E1" ], "object_refs": [], "confidence": "high", "temporal_status": "localized", "intervals": [ { "start_s": 0.0, "end_s": 48.5 } ], "evidence": "The mask-linked person remains in a low squatting posture beside the wheel throughout all sampled frames, including the final sample.", "mask_relation": "The action subject matches the supplied selected-person mask." }, { "id": "A2", "mentions": [ { "text": "picks up", "ranges": [ [ 52, 60 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "type": "phrasal", "subject_refs": [ "E1" ], "object_refs": [ "E3" ], "confidence": "high", "temporal_status": "localized", "intervals": [ { "start_s": 2.0, "end_s": 5.0 } ], "evidence": "At 2–4 seconds, the mask-linked person reaches down toward the loose hardware and brings a small item back toward the wheel.", "mask_relation": "The action subject matches the supplied selected-person mask." }, { "id": "A3", "mentions": [ { "text": "screw", "ranges": [ [ 103, 108 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } }, { "text": "securing", "ranges": [ [ 199, 207 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "type": "simple", "subject_refs": [ "E1" ], "object_refs": [ "E3", "E5" ], "confidence": "high", "temporal_status": "localized", "intervals": [ { "start_s": 0.0, "end_s": 2.0 }, { "start_s": 4.0, "end_s": 15.0 }, { "start_s": 16.0, "end_s": 22.0 }, { "start_s": 24.0, "end_s": 30.0 }, { "start_s": 31.0, "end_s": 36.0 }, { "start_s": 37.0, "end_s": 45.0 } ], "evidence": "Across these sampled intervals, the mask-linked person repeatedly holds and rotates small lug nuts at the wheel studs by hand. “securing” summarizes the same fastening event.", "mask_relation": "The action subject matches the supplied selected-person mask." }, { "id": "A4", "mentions": [ { "text": "using", "ranges": [ [ 155, 160 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "type": "simple", "subject_refs": [ "E1" ], "object_refs": [ "E7" ], "confidence": "medium", "temporal_status": "localized", "intervals": [ { "start_s": 16.0, "end_s": 23.0 } ], "evidence": "A small orange cloth is visible in the mask-linked person's hands against the wheel area during the supported samples.", "mask_relation": "The action subject matches the supplied selected-person mask." }, { "id": "A5", "mentions": [ { "text": "wipe", "ranges": [ [ 178, 182 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "type": "simple", "subject_refs": [ "E1" ], "object_refs": [], "instrument_refs": [ "E7" ], "confidence": "medium", "temporal_status": "localized", "intervals": [ { "start_s": 16.0, "end_s": 23.0 } ], "evidence": "During 16–22 seconds, the cloth is held against and moved around the wheel's lug area; sparse frames support wiping but do not reveal exact stroke boundaries.", "mask_relation": "The action subject matches the supplied selected-person mask." }, { "id": "A6", "mentions": [ { "text": "removes", "ranges": [ [ 233, 240 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "type": "simple", "subject_refs": [ "E1" ], "object_refs": [ "E8" ], "confidence": "high", "temporal_status": "localized", "intervals": [ { "start_s": 46.0, "end_s": 48.5 } ], "evidence": "At 46–48 seconds, the mask-linked person visibly pulls the gloves from the hands; the removal continues through the clip's right boundary.", "mask_relation": "The action subject matches the supplied selected-person mask." } ], "notes": [ "The supplied masklet consistently identifies the captioned person, so only that entity group is linked to a source mask.", "No separate masks were supplied for the tire, lug nuts, ground, studs, hand, cloth, or gloves.", "The result-state premodifier “mounted” was not annotated as an action. “area” was excluded as an abstract spatial reference.", "Action intervals are sparse clip-relative visual-support intervals derived from the sampled clean RGB frames, not exact continuous-video boundaries." ] }, "normalization_repairs": 0 }, { "source_video_id": "v_-0NfmccuL9Q", "caption_count_category": 8, "clip": { "caption_index": 6, "caption": "The person picks up an impact wrench from the ground and attaches a socket to it. Crouching down, they use the power tool to systematically tighten each lug nut on the wheel. Their left hand rests on top of the tire for stability during this process.", "source_interval_sec": [ 148.12, 179.12 ], "loom_clip_id": "v_-0NfmccuL9Q_4439_5368", "clip_frame_interval_4fps": [ 4439, 5368 ], "sample_count_1fps": 31, "sampled_frames": [ { "frame": "00000", "timestamp_sec": 148.12, "offset_sec": 0, "rgb_path": "assets/cat_08/v_-0NfmccuL9Q_4439_5368/0000s_00000_rgb.jpg", "mask_path": "assets/cat_08/v_-0NfmccuL9Q_4439_5368/0000s_00000_mask.png", "mask_status": "available" }, { "frame": "00028", "timestamp_sec": 149.12, "offset_sec": 1, "rgb_path": "assets/cat_08/v_-0NfmccuL9Q_4439_5368/0001s_00028_rgb.jpg", "mask_path": "assets/cat_08/v_-0NfmccuL9Q_4439_5368/0001s_00028_mask.png", "mask_status": "available" }, { "frame": "00060", "timestamp_sec": 150.12, "offset_sec": 2, "rgb_path": "assets/cat_08/v_-0NfmccuL9Q_4439_5368/0002s_00060_rgb.jpg", "mask_path": "assets/cat_08/v_-0NfmccuL9Q_4439_5368/0002s_00060_mask.png", "mask_status": "available" }, { "frame": "00088", "timestamp_sec": 151.12, "offset_sec": 3, "rgb_path": "assets/cat_08/v_-0NfmccuL9Q_4439_5368/0003s_00088_rgb.jpg", "mask_path": "assets/cat_08/v_-0NfmccuL9Q_4439_5368/0003s_00088_mask.png", "mask_status": "available" }, { "frame": "00120", "timestamp_sec": 152.12, "offset_sec": 4, "rgb_path": "assets/cat_08/v_-0NfmccuL9Q_4439_5368/0004s_00120_rgb.jpg", "mask_path": "assets/cat_08/v_-0NfmccuL9Q_4439_5368/0004s_00120_mask.png", "mask_status": "available" }, { "frame": "00148", "timestamp_sec": 153.12, "offset_sec": 5, "rgb_path": "assets/cat_08/v_-0NfmccuL9Q_4439_5368/0005s_00148_rgb.jpg", "mask_path": "assets/cat_08/v_-0NfmccuL9Q_4439_5368/0005s_00148_mask.png", "mask_status": "available" }, { "frame": "00180", "timestamp_sec": 154.12, "offset_sec": 6, "rgb_path": "assets/cat_08/v_-0NfmccuL9Q_4439_5368/0006s_00180_rgb.jpg", "mask_path": "assets/cat_08/v_-0NfmccuL9Q_4439_5368/0006s_00180_mask.png", "mask_status": "available" }, { "frame": "00208", "timestamp_sec": 155.12, "offset_sec": 7, "rgb_path": "assets/cat_08/v_-0NfmccuL9Q_4439_5368/0007s_00208_rgb.jpg", "mask_path": "assets/cat_08/v_-0NfmccuL9Q_4439_5368/0007s_00208_mask.png", "mask_status": "available" }, { "frame": "00240", "timestamp_sec": 156.12, "offset_sec": 8, "rgb_path": "assets/cat_08/v_-0NfmccuL9Q_4439_5368/0008s_00240_rgb.jpg", "mask_path": "assets/cat_08/v_-0NfmccuL9Q_4439_5368/0008s_00240_mask.png", "mask_status": "available" }, { "frame": "00268", "timestamp_sec": 157.12, "offset_sec": 9, "rgb_path": "assets/cat_08/v_-0NfmccuL9Q_4439_5368/0009s_00268_rgb.jpg", "mask_path": "assets/cat_08/v_-0NfmccuL9Q_4439_5368/0009s_00268_mask.png", "mask_status": "available" }, { "frame": "00300", "timestamp_sec": 158.12, "offset_sec": 10, "rgb_path": "assets/cat_08/v_-0NfmccuL9Q_4439_5368/0010s_00300_rgb.jpg", "mask_path": "assets/cat_08/v_-0NfmccuL9Q_4439_5368/0010s_00300_mask.png", "mask_status": "available" }, { "frame": "00328", "timestamp_sec": 159.12, "offset_sec": 11, "rgb_path": "assets/cat_08/v_-0NfmccuL9Q_4439_5368/0011s_00328_rgb.jpg", "mask_path": "assets/cat_08/v_-0NfmccuL9Q_4439_5368/0011s_00328_mask.png", "mask_status": "available" }, { "frame": "00360", "timestamp_sec": 160.12, "offset_sec": 12, "rgb_path": "assets/cat_08/v_-0NfmccuL9Q_4439_5368/0012s_00360_rgb.jpg", "mask_path": "assets/cat_08/v_-0NfmccuL9Q_4439_5368/0012s_00360_mask.png", "mask_status": "available" }, { "frame": "00388", "timestamp_sec": 161.12, "offset_sec": 13, "rgb_path": "assets/cat_08/v_-0NfmccuL9Q_4439_5368/0013s_00388_rgb.jpg", "mask_path": "assets/cat_08/v_-0NfmccuL9Q_4439_5368/0013s_00388_mask.png", "mask_status": "available" }, { "frame": "00420", "timestamp_sec": 162.12, "offset_sec": 14, "rgb_path": "assets/cat_08/v_-0NfmccuL9Q_4439_5368/0014s_00420_rgb.jpg", "mask_path": "assets/cat_08/v_-0NfmccuL9Q_4439_5368/0014s_00420_mask.png", "mask_status": "available" }, { "frame": "00448", "timestamp_sec": 163.12, "offset_sec": 15, "rgb_path": "assets/cat_08/v_-0NfmccuL9Q_4439_5368/0015s_00448_rgb.jpg", "mask_path": "assets/cat_08/v_-0NfmccuL9Q_4439_5368/0015s_00448_mask.png", "mask_status": "available" }, { "frame": "00480", "timestamp_sec": 164.12, "offset_sec": 16, "rgb_path": "assets/cat_08/v_-0NfmccuL9Q_4439_5368/0016s_00480_rgb.jpg", "mask_path": "assets/cat_08/v_-0NfmccuL9Q_4439_5368/0016s_00480_mask.png", "mask_status": "available" }, { "frame": "00508", "timestamp_sec": 165.12, "offset_sec": 17, "rgb_path": "assets/cat_08/v_-0NfmccuL9Q_4439_5368/0017s_00508_rgb.jpg", "mask_path": "assets/cat_08/v_-0NfmccuL9Q_4439_5368/0017s_00508_mask.png", "mask_status": "available" }, { "frame": "00540", "timestamp_sec": 166.12, "offset_sec": 18, "rgb_path": "assets/cat_08/v_-0NfmccuL9Q_4439_5368/0018s_00540_rgb.jpg", "mask_path": "assets/cat_08/v_-0NfmccuL9Q_4439_5368/0018s_00540_mask.png", "mask_status": "available" }, { "frame": "00568", "timestamp_sec": 167.12, "offset_sec": 19, "rgb_path": "assets/cat_08/v_-0NfmccuL9Q_4439_5368/0019s_00568_rgb.jpg", "mask_path": "assets/cat_08/v_-0NfmccuL9Q_4439_5368/0019s_00568_mask.png", "mask_status": "available" }, { "frame": "00600", "timestamp_sec": 168.12, "offset_sec": 20, "rgb_path": "assets/cat_08/v_-0NfmccuL9Q_4439_5368/0020s_00600_rgb.jpg", "mask_path": "assets/cat_08/v_-0NfmccuL9Q_4439_5368/0020s_00600_mask.png", "mask_status": "available" }, { "frame": "00628", "timestamp_sec": 169.12, "offset_sec": 21, "rgb_path": "assets/cat_08/v_-0NfmccuL9Q_4439_5368/0021s_00628_rgb.jpg", "mask_path": "assets/cat_08/v_-0NfmccuL9Q_4439_5368/0021s_00628_mask.png", "mask_status": "available" }, { "frame": "00660", "timestamp_sec": 170.12, "offset_sec": 22, "rgb_path": "assets/cat_08/v_-0NfmccuL9Q_4439_5368/0022s_00660_rgb.jpg", "mask_path": "assets/cat_08/v_-0NfmccuL9Q_4439_5368/0022s_00660_mask.png", "mask_status": "available" }, { "frame": "00688", "timestamp_sec": 171.12, "offset_sec": 23, "rgb_path": "assets/cat_08/v_-0NfmccuL9Q_4439_5368/0023s_00688_rgb.jpg", "mask_path": "assets/cat_08/v_-0NfmccuL9Q_4439_5368/0023s_00688_mask.png", "mask_status": "available" }, { "frame": "00720", "timestamp_sec": 172.12, "offset_sec": 24, "rgb_path": "assets/cat_08/v_-0NfmccuL9Q_4439_5368/0024s_00720_rgb.jpg", "mask_path": "assets/cat_08/v_-0NfmccuL9Q_4439_5368/0024s_00720_mask.png", "mask_status": "available" }, { "frame": "00748", "timestamp_sec": 173.12, "offset_sec": 25, "rgb_path": "assets/cat_08/v_-0NfmccuL9Q_4439_5368/0025s_00748_rgb.jpg", "mask_path": "assets/cat_08/v_-0NfmccuL9Q_4439_5368/0025s_00748_mask.png", "mask_status": "available" }, { "frame": "00780", "timestamp_sec": 174.12, "offset_sec": 26, "rgb_path": "assets/cat_08/v_-0NfmccuL9Q_4439_5368/0026s_00780_rgb.jpg", "mask_path": "assets/cat_08/v_-0NfmccuL9Q_4439_5368/0026s_00780_mask.png", "mask_status": "available" }, { "frame": "00808", "timestamp_sec": 175.12, "offset_sec": 27, "rgb_path": "assets/cat_08/v_-0NfmccuL9Q_4439_5368/0027s_00808_rgb.jpg", "mask_path": "assets/cat_08/v_-0NfmccuL9Q_4439_5368/0027s_00808_mask.png", "mask_status": "available" }, { "frame": "00840", "timestamp_sec": 176.12, "offset_sec": 28, "rgb_path": "assets/cat_08/v_-0NfmccuL9Q_4439_5368/0028s_00840_rgb.jpg", "mask_path": "assets/cat_08/v_-0NfmccuL9Q_4439_5368/0028s_00840_mask.png", "mask_status": "available" }, { "frame": "00868", "timestamp_sec": 177.12, "offset_sec": 29, "rgb_path": "assets/cat_08/v_-0NfmccuL9Q_4439_5368/0029s_00868_rgb.jpg", "mask_path": "assets/cat_08/v_-0NfmccuL9Q_4439_5368/0029s_00868_mask.png", "mask_status": "available" }, { "frame": "00900", "timestamp_sec": 178.12, "offset_sec": 30, "rgb_path": "assets/cat_08/v_-0NfmccuL9Q_4439_5368/0030s_00900_rgb.jpg", "mask_path": "assets/cat_08/v_-0NfmccuL9Q_4439_5368/0030s_00900_mask.png", "mask_status": "available" } ], "obj_id": [ 0 ], "anno_id": [ 10 ], "fps_inferred_from_clip_interval": 29.9677 }, "usage": { "input_tokens": 168823, "input_tokens_details": { "cache_write_tokens": 168820, "cached_tokens": 0 }, "output_tokens": 4458, "output_tokens_details": { "reasoning_tokens": 2588 }, "total_tokens": 173281 }, "model": "gpt-5.6-sol", "effort": "medium", "custom_id": "cat08_cap06_v_-0NfmccuL9Q_4439_5368", "annotation": { "caption": "The person picks up an impact wrench from the ground and attaches a socket to it. Crouching down, they use the power tool to systematically tighten each lug nut on the wheel. Their left hand rests on top of the tire for stability during this process.", "entity_groups": [ { "id": "person_1", "label": "person", "spans": [ { "text": "The person", "range": [ 0, 10 ] }, { "text": "they", "range": [ 98, 102 ] }, { "text": "Their", "range": [ 175, 180 ] } ], "source_mask_link": "selected_person" }, { "id": "impact_wrench_1", "label": "impact wrench", "spans": [ { "text": "an impact wrench", "range": [ 20, 36 ] }, { "text": "it", "range": [ 78, 80 ] }, { "text": "the power tool", "range": [ 107, 121 ] } ], "source_mask_link": null }, { "id": "ground_1", "label": "ground", "spans": [ { "text": "the ground", "range": [ 42, 52 ] } ], "source_mask_link": null }, { "id": "socket_1", "label": "socket", "spans": [ { "text": "a socket", "range": [ 66, 74 ] } ], "source_mask_link": null }, { "id": "lug_nut_group_1", "label": "lug nut", "spans": [ { "text": "each lug nut", "range": [ 148, 160 ] } ], "source_mask_link": null }, { "id": "wheel_1", "label": "wheel", "spans": [ { "text": "the wheel", "range": [ 164, 173 ] } ], "source_mask_link": null }, { "id": "left_hand_1", "label": "left hand", "spans": [ { "text": "Their left hand", "range": [ 175, 190 ] } ], "source_mask_link": null }, { "id": "tire_1", "label": "tire", "spans": [ { "text": "the tire", "range": [ 207, 215 ] } ], "source_mask_link": null } ], "actions": [ { "id": "action_1", "mentions": [ { "text": "picks up", "ranges": [ [ 11, 19 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "type": "phrasal", "subject_refs": [ "person_1" ], "object_refs": [ "impact_wrench_1" ], "confidence": "medium", "temporal_status": "supported", "intervals": [ { "start_s": 2.0, "end_s": 3.0, "source_start_s": 150.12, "source_end_s": 151.12 } ], "evidence": "The wrench is on the ground at 0–1 seconds and is visibly held by the masked person at 2 seconds.", "mask_relation": "The action subject matches the supplied selected-person mask.", "source_subject_refs": "person_1" }, { "id": "action_2", "mentions": [ { "text": "attaches", "ranges": [ [ 57, 65 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "type": "simple", "subject_refs": [ "person_1" ], "object_refs": [ "socket_1", "impact_wrench_1" ], "confidence": "medium", "temporal_status": "supported", "intervals": [ { "start_s": 2.0, "end_s": 5.0, "source_start_s": 150.12, "source_end_s": 153.12 } ], "evidence": "At 2–4 seconds the masked person holds the wrench away from the wheel and manipulates its front fitting with both hands before placing it on a lug nut.", "mask_relation": "The action subject matches the supplied selected-person mask.", "source_subject_refs": "person_1" }, { "id": "action_3", "mentions": [ { "text": "Crouching", "ranges": [ [ 82, 91 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "type": "posture", "subject_refs": [ "person_1" ], "object_refs": [], "confidence": "high", "temporal_status": "supported", "intervals": [ { "start_s": 0.0, "end_s": 31.0, "source_start_s": 148.12, "source_end_s": 179.12 } ], "evidence": "The masked person maintains a low crouching or kneeling posture beside the wheel throughout all sampled frames.", "mask_relation": "The action subject matches the supplied selected-person mask.", "source_subject_refs": "person_1" }, { "id": "action_4", "mentions": [ { "text": "use", "ranges": [ [ 103, 106 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "type": "simple", "subject_refs": [ "person_1" ], "object_refs": [ "impact_wrench_1" ], "confidence": "high", "temporal_status": "supported", "intervals": [ { "start_s": 5.0, "end_s": 31.0, "source_start_s": 153.12, "source_end_s": 179.12 } ], "evidence": "From 5 seconds through the final frame, the masked person holds and operates the impact wrench against successive wheel fasteners.", "mask_relation": "The action subject matches the supplied selected-person mask.", "source_subject_refs": "person_1" }, { "id": "action_5", "mentions": [ { "text": "tighten", "ranges": [ [ 140, 147 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "type": "simple", "subject_refs": [ "person_1" ], "object_refs": [ "lug_nut_group_1" ], "confidence": "high", "temporal_status": "supported", "intervals": [ { "start_s": 5.0, "end_s": 31.0, "source_start_s": 153.12, "source_end_s": 179.12 } ], "evidence": "At every sampled frame from 5–30 seconds, the masked person places the impact wrench on lug positions around the wheel, with the final frame still supporting the operation.", "mask_relation": "The action subject matches the supplied selected-person mask.", "source_subject_refs": "person_1" } ], "notes": [ "Only the person has a supplied source mask; all tool, vehicle-part, body-part, and scene-entity groups therefore have null source-mask links.", "The predicate \"rests\" at [191,196) was not included as an action because its grammatical subject is the separately annotated left hand, which has no corresponding supplied instance mask.", "The directional word \"down\" following \"Crouching\" is excluded from the action span.", "The tightening/use intervals reach the clip boundary because the final sampled frame at 30.0 seconds still visibly supports those actions." ] }, "model_output_raw": { "caption": "The person picks up an impact wrench from the ground and attaches a socket to it. Crouching down, they use the power tool to systematically tighten each lug nut on the wheel. Their left hand rests on top of the tire for stability during this process.", "entity_groups": [ { "id": "person_1", "label": "person", "spans": [ { "text": "The person", "range": [ 0, 10 ] }, { "text": "they", "range": [ 98, 102 ] }, { "text": "Their", "range": [ 175, 180 ] } ], "source_mask": { "track_id": "selected_person", "description": "The supplied binary mask tracks the crouching person in all sampled frames." } }, { "id": "impact_wrench_1", "label": "impact wrench", "spans": [ { "text": "an impact wrench", "range": [ 20, 36 ] }, { "text": "it", "range": [ 78, 80 ] }, { "text": "the power tool", "range": [ 107, 121 ] } ], "source_mask": null }, { "id": "ground_1", "label": "ground", "spans": [ { "text": "the ground", "range": [ 42, 52 ] } ], "source_mask": null }, { "id": "socket_1", "label": "socket", "spans": [ { "text": "a socket", "range": [ 66, 74 ] } ], "source_mask": null }, { "id": "lug_nut_group_1", "label": "lug nut", "spans": [ { "text": "each lug nut", "range": [ 148, 160 ] } ], "source_mask": null }, { "id": "wheel_1", "label": "wheel", "spans": [ { "text": "the wheel", "range": [ 164, 173 ] } ], "source_mask": null }, { "id": "left_hand_1", "label": "left hand", "spans": [ { "text": "Their left hand", "range": [ 175, 190 ] } ], "source_mask": null }, { "id": "tire_1", "label": "tire", "spans": [ { "text": "the tire", "range": [ 207, 215 ] } ], "source_mask": null } ], "actions": [ { "id": "action_1", "mentions": [ { "text": "picks up", "ranges": [ [ 11, 19 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "type": "phrasal", "subject": "person_1", "objects": [ "impact_wrench_1" ], "confidence": "medium", "temporal_status": "supported", "intervals": [ { "clip_relative_s": [ 2.0, 3.0 ], "source_video_s": [ 150.12, 151.12 ] } ], "evidence": "The wrench is on the ground at 0–1 seconds and is visibly held by the masked person at 2 seconds.", "mask_relation": "The action subject matches the supplied selected-person mask." }, { "id": "action_2", "mentions": [ { "text": "attaches", "ranges": [ [ 57, 65 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "type": "simple", "subject": "person_1", "objects": [ "socket_1", "impact_wrench_1" ], "confidence": "medium", "temporal_status": "supported", "intervals": [ { "clip_relative_s": [ 2.0, 5.0 ], "source_video_s": [ 150.12, 153.12 ] } ], "evidence": "At 2–4 seconds the masked person holds the wrench away from the wheel and manipulates its front fitting with both hands before placing it on a lug nut.", "mask_relation": "The action subject matches the supplied selected-person mask." }, { "id": "action_3", "mentions": [ { "text": "Crouching", "ranges": [ [ 82, 91 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "type": "posture", "subject": "person_1", "objects": [], "confidence": "high", "temporal_status": "supported", "intervals": [ { "clip_relative_s": [ 0.0, 31.0 ], "source_video_s": [ 148.12, 179.12 ] } ], "evidence": "The masked person maintains a low crouching or kneeling posture beside the wheel throughout all sampled frames.", "mask_relation": "The action subject matches the supplied selected-person mask." }, { "id": "action_4", "mentions": [ { "text": "use", "ranges": [ [ 103, 106 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "type": "simple", "subject": "person_1", "objects": [ "impact_wrench_1" ], "confidence": "high", "temporal_status": "supported", "intervals": [ { "clip_relative_s": [ 5.0, 31.0 ], "source_video_s": [ 153.12, 179.12 ] } ], "evidence": "From 5 seconds through the final frame, the masked person holds and operates the impact wrench against successive wheel fasteners.", "mask_relation": "The action subject matches the supplied selected-person mask." }, { "id": "action_5", "mentions": [ { "text": "tighten", "ranges": [ [ 140, 147 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "type": "simple", "subject": "person_1", "objects": [ "lug_nut_group_1" ], "confidence": "high", "temporal_status": "supported", "intervals": [ { "clip_relative_s": [ 5.0, 31.0 ], "source_video_s": [ 153.12, 179.12 ] } ], "evidence": "At every sampled frame from 5–30 seconds, the masked person places the impact wrench on lug positions around the wheel, with the final frame still supporting the operation.", "mask_relation": "The action subject matches the supplied selected-person mask." } ], "notes": [ "Only the person has a supplied source mask; all tool, vehicle-part, body-part, and scene-entity groups therefore have null source-mask links.", "The predicate \"rests\" at [191,196) was not included as an action because its grammatical subject is the separately annotated left hand, which has no corresponding supplied instance mask.", "The directional word \"down\" following \"Crouching\" is excluded from the action span.", "The tightening/use intervals reach the clip boundary because the final sampled frame at 30.0 seconds still visibly supports those actions." ] }, "normalization_repairs": 0 }, { "source_video_id": "v_-0NfmccuL9Q", "caption_count_category": 8, "clip": { "caption_index": 7, "caption": "After placing the impact wrench on the ground, the person crouches to operate the car jack. They use the jack's mechanism to slowly lower the car until the tire is resting firmly on the pavement.", "source_interval_sec": [ 179.12, 206.12 ], "loom_clip_id": "v_-0NfmccuL9Q_5368_6177", "clip_frame_interval_4fps": [ 5368, 6177 ], "sample_count_1fps": 27, "sampled_frames": [ { "frame": "00000", "timestamp_sec": 179.12, "offset_sec": 0, "rgb_path": "assets/cat_08/v_-0NfmccuL9Q_5368_6177/0000s_00000_rgb.jpg", "mask_path": "assets/cat_08/v_-0NfmccuL9Q_5368_6177/0000s_00000_mask.png", "mask_status": "available" }, { "frame": "00028", "timestamp_sec": 180.12, "offset_sec": 1, "rgb_path": "assets/cat_08/v_-0NfmccuL9Q_5368_6177/0001s_00028_rgb.jpg", "mask_path": "assets/cat_08/v_-0NfmccuL9Q_5368_6177/0001s_00028_mask.png", "mask_status": "available" }, { "frame": "00060", "timestamp_sec": 181.12, "offset_sec": 2, "rgb_path": "assets/cat_08/v_-0NfmccuL9Q_5368_6177/0002s_00060_rgb.jpg", "mask_path": "assets/cat_08/v_-0NfmccuL9Q_5368_6177/0002s_00060_mask.png", "mask_status": "available" }, { "frame": "00088", "timestamp_sec": 182.12, "offset_sec": 3, "rgb_path": "assets/cat_08/v_-0NfmccuL9Q_5368_6177/0003s_00088_rgb.jpg", "mask_path": "assets/cat_08/v_-0NfmccuL9Q_5368_6177/0003s_00088_mask.png", "mask_status": "available" }, { "frame": "00120", "timestamp_sec": 183.12, "offset_sec": 4, "rgb_path": "assets/cat_08/v_-0NfmccuL9Q_5368_6177/0004s_00120_rgb.jpg", "mask_path": "assets/cat_08/v_-0NfmccuL9Q_5368_6177/0004s_00120_mask.png", "mask_status": "available" }, { "frame": "00148", "timestamp_sec": 184.12, "offset_sec": 5, "rgb_path": "assets/cat_08/v_-0NfmccuL9Q_5368_6177/0005s_00148_rgb.jpg", "mask_path": "assets/cat_08/v_-0NfmccuL9Q_5368_6177/0005s_00148_mask.png", "mask_status": "available" }, { "frame": "00180", "timestamp_sec": 185.12, "offset_sec": 6, "rgb_path": "assets/cat_08/v_-0NfmccuL9Q_5368_6177/0006s_00180_rgb.jpg", "mask_path": "assets/cat_08/v_-0NfmccuL9Q_5368_6177/0006s_00180_mask.png", "mask_status": "available" }, { "frame": "00208", "timestamp_sec": 186.12, "offset_sec": 7, "rgb_path": "assets/cat_08/v_-0NfmccuL9Q_5368_6177/0007s_00208_rgb.jpg", "mask_path": "assets/cat_08/v_-0NfmccuL9Q_5368_6177/0007s_00208_mask.png", "mask_status": "available" }, { "frame": "00240", "timestamp_sec": 187.12, "offset_sec": 8, "rgb_path": "assets/cat_08/v_-0NfmccuL9Q_5368_6177/0008s_00240_rgb.jpg", "mask_path": "assets/cat_08/v_-0NfmccuL9Q_5368_6177/0008s_00240_mask.png", "mask_status": "available" }, { "frame": "00268", "timestamp_sec": 188.12, "offset_sec": 9, "rgb_path": "assets/cat_08/v_-0NfmccuL9Q_5368_6177/0009s_00268_rgb.jpg", "mask_path": "assets/cat_08/v_-0NfmccuL9Q_5368_6177/0009s_00268_mask.png", "mask_status": "available" }, { "frame": "00300", "timestamp_sec": 189.12, "offset_sec": 10, "rgb_path": "assets/cat_08/v_-0NfmccuL9Q_5368_6177/0010s_00300_rgb.jpg", "mask_path": "assets/cat_08/v_-0NfmccuL9Q_5368_6177/0010s_00300_mask.png", "mask_status": "available" }, { "frame": "00328", "timestamp_sec": 190.12, "offset_sec": 11, "rgb_path": "assets/cat_08/v_-0NfmccuL9Q_5368_6177/0011s_00328_rgb.jpg", "mask_path": "assets/cat_08/v_-0NfmccuL9Q_5368_6177/0011s_00328_mask.png", "mask_status": "available" }, { "frame": "00360", "timestamp_sec": 191.12, "offset_sec": 12, "rgb_path": "assets/cat_08/v_-0NfmccuL9Q_5368_6177/0012s_00360_rgb.jpg", "mask_path": "assets/cat_08/v_-0NfmccuL9Q_5368_6177/0012s_00360_mask.png", "mask_status": "available" }, { "frame": "00388", "timestamp_sec": 192.12, "offset_sec": 13, "rgb_path": "assets/cat_08/v_-0NfmccuL9Q_5368_6177/0013s_00388_rgb.jpg", "mask_path": "assets/cat_08/v_-0NfmccuL9Q_5368_6177/0013s_00388_mask.png", "mask_status": "available" }, { "frame": "00420", "timestamp_sec": 193.12, "offset_sec": 14, "rgb_path": "assets/cat_08/v_-0NfmccuL9Q_5368_6177/0014s_00420_rgb.jpg", "mask_path": "assets/cat_08/v_-0NfmccuL9Q_5368_6177/0014s_00420_mask.png", "mask_status": "available" }, { "frame": "00448", "timestamp_sec": 194.12, "offset_sec": 15, "rgb_path": "assets/cat_08/v_-0NfmccuL9Q_5368_6177/0015s_00448_rgb.jpg", "mask_path": "assets/cat_08/v_-0NfmccuL9Q_5368_6177/0015s_00448_mask.png", "mask_status": "available" }, { "frame": "00480", "timestamp_sec": 195.12, "offset_sec": 16, "rgb_path": "assets/cat_08/v_-0NfmccuL9Q_5368_6177/0016s_00480_rgb.jpg", "mask_path": "assets/cat_08/v_-0NfmccuL9Q_5368_6177/0016s_00480_mask.png", "mask_status": "available" }, { "frame": "00508", "timestamp_sec": 196.12, "offset_sec": 17, "rgb_path": "assets/cat_08/v_-0NfmccuL9Q_5368_6177/0017s_00508_rgb.jpg", "mask_path": "assets/cat_08/v_-0NfmccuL9Q_5368_6177/0017s_00508_mask.png", "mask_status": "available" }, { "frame": "00540", "timestamp_sec": 197.12, "offset_sec": 18, "rgb_path": "assets/cat_08/v_-0NfmccuL9Q_5368_6177/0018s_00540_rgb.jpg", "mask_path": "assets/cat_08/v_-0NfmccuL9Q_5368_6177/0018s_00540_mask.png", "mask_status": "available" }, { "frame": "00568", "timestamp_sec": 198.12, "offset_sec": 19, "rgb_path": "assets/cat_08/v_-0NfmccuL9Q_5368_6177/0019s_00568_rgb.jpg", "mask_path": "assets/cat_08/v_-0NfmccuL9Q_5368_6177/0019s_00568_mask.png", "mask_status": "available" }, { "frame": "00600", "timestamp_sec": 199.12, "offset_sec": 20, "rgb_path": "assets/cat_08/v_-0NfmccuL9Q_5368_6177/0020s_00600_rgb.jpg", "mask_path": "assets/cat_08/v_-0NfmccuL9Q_5368_6177/0020s_00600_mask.png", "mask_status": "available" }, { "frame": "00628", "timestamp_sec": 200.12, "offset_sec": 21, "rgb_path": "assets/cat_08/v_-0NfmccuL9Q_5368_6177/0021s_00628_rgb.jpg", "mask_path": "assets/cat_08/v_-0NfmccuL9Q_5368_6177/0021s_00628_mask.png", "mask_status": "available" }, { "frame": "00660", "timestamp_sec": 201.12, "offset_sec": 22, "rgb_path": "assets/cat_08/v_-0NfmccuL9Q_5368_6177/0022s_00660_rgb.jpg", "mask_path": "assets/cat_08/v_-0NfmccuL9Q_5368_6177/0022s_00660_mask.png", "mask_status": "available" }, { "frame": "00688", "timestamp_sec": 202.12, "offset_sec": 23, "rgb_path": "assets/cat_08/v_-0NfmccuL9Q_5368_6177/0023s_00688_rgb.jpg", "mask_path": "assets/cat_08/v_-0NfmccuL9Q_5368_6177/0023s_00688_mask.png", "mask_status": "available" }, { "frame": "00720", "timestamp_sec": 203.12, "offset_sec": 24, "rgb_path": "assets/cat_08/v_-0NfmccuL9Q_5368_6177/0024s_00720_rgb.jpg", "mask_path": "assets/cat_08/v_-0NfmccuL9Q_5368_6177/0024s_00720_mask.png", "mask_status": "available" }, { "frame": "00748", "timestamp_sec": 204.12, "offset_sec": 25, "rgb_path": "assets/cat_08/v_-0NfmccuL9Q_5368_6177/0025s_00748_rgb.jpg", "mask_path": "assets/cat_08/v_-0NfmccuL9Q_5368_6177/0025s_00748_mask.png", "mask_status": "available" }, { "frame": "00780", "timestamp_sec": 205.12, "offset_sec": 26, "rgb_path": "assets/cat_08/v_-0NfmccuL9Q_5368_6177/0026s_00780_rgb.jpg", "mask_path": "assets/cat_08/v_-0NfmccuL9Q_5368_6177/0026s_00780_mask.png", "mask_status": "available" } ], "obj_id": [ 0 ], "anno_id": [ 11 ], "fps_inferred_from_clip_interval": 29.963 }, "usage": { "input_tokens": 149007, "input_tokens_details": { "cache_write_tokens": 149004, "cached_tokens": 0 }, "output_tokens": 4523, "output_tokens_details": { "reasoning_tokens": 2950 }, "total_tokens": 153530 }, "model": "gpt-5.6-sol", "effort": "medium", "custom_id": "cat08_cap07_v_-0NfmccuL9Q_5368_6177", "annotation": { "caption": "After placing the impact wrench on the ground, the person crouches to operate the car jack. They use the jack's mechanism to slowly lower the car until the tire is resting firmly on the pavement.", "entity_groups": [ { "id": "entity_1", "label": "impact wrench", "spans": [ { "text": "the impact wrench", "range": [ 14, 31 ] } ], "source_mask_link": null }, { "id": "entity_2", "label": "ground/pavement surface", "spans": [ { "text": "the ground", "range": [ 35, 45 ] }, { "text": "the pavement", "range": [ 182, 194 ] } ], "source_mask_link": null }, { "id": "entity_3", "label": "person", "spans": [ { "text": "the person", "range": [ 47, 57 ] }, { "text": "They", "range": [ 92, 96 ] } ], "source_mask_link": "selected_person" }, { "id": "entity_4", "label": "car jack", "spans": [ { "text": "the car jack", "range": [ 78, 90 ] }, { "text": "the jack's", "range": [ 101, 111 ] } ], "source_mask_link": null }, { "id": "entity_5", "label": "jack mechanism", "spans": [ { "text": "the jack's mechanism", "range": [ 101, 121 ] } ], "source_mask_link": null }, { "id": "entity_6", "label": "car", "spans": [ { "text": "the car", "range": [ 138, 145 ] } ], "source_mask_link": null }, { "id": "entity_7", "label": "tire", "spans": [ { "text": "the tire", "range": [ 152, 160 ] } ], "source_mask_link": null } ], "actions": [ { "id": "action_1", "mentions": [ { "text": "placing", "ranges": [ [ 6, 13 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "type": "simple", "subject_refs": [ "entity_3" ], "object_refs": [ "entity_1", "entity_2" ], "confidence": "medium", "temporal_status": "uncertain", "intervals": [], "evidence": "The mask identifies the person handling the impact wrench, but none of the sampled clean frames clearly shows the wrench being released onto the ground.", "mask_relation": "subject matches the supplied selected-person mask", "source_subject_refs": "entity_3" }, { "id": "action_2", "mentions": [ { "text": "crouches", "ranges": [ [ 58, 66 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "type": "posture", "subject_refs": [ "entity_3" ], "object_refs": [], "confidence": "high", "temporal_status": "localized", "intervals": [ { "start_s": 0.0, "end_s": 27.0, "source_start_s": 179.12, "source_end_s": 206.12 } ], "evidence": "The masked person remains visibly crouched beside the wheel and jack in every sampled clean frame from 0.000 through 26.000 seconds.", "mask_relation": "subject matches the supplied selected-person mask", "source_subject_refs": "entity_3" }, { "id": "action_3", "mentions": [ { "text": "operate", "ranges": [ [ 70, 77 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } }, { "text": "use", "ranges": [ [ 97, 100 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "type": "simple", "subject_refs": [ "entity_3" ], "object_refs": [ "entity_4", "entity_5" ], "confidence": "high", "temporal_status": "localized", "intervals": [ { "start_s": 3.0, "end_s": 27.0, "source_start_s": 182.12, "source_end_s": 206.12 } ], "evidence": "From 3.000 through 26.000 seconds, the masked person repeatedly holds the impact wrench against the jack's drive point, visibly operating the jack mechanism.", "mask_relation": "subject matches the supplied selected-person mask", "source_subject_refs": "entity_3" }, { "id": "action_4", "mentions": [ { "text": "lower", "ranges": [ [ 132, 137 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "type": "simple", "subject_refs": [ "entity_3" ], "object_refs": [ "entity_6" ], "confidence": "medium", "temporal_status": "localized", "intervals": [ { "start_s": 3.0, "end_s": 27.0, "source_start_s": 182.12, "source_end_s": 206.12 } ], "evidence": "While the masked person operates the jack from 3.000 through 26.000 seconds, the jack progressively compresses and the supported car and wheel descend toward the pavement.", "mask_relation": "subject matches the supplied selected-person mask", "source_subject_refs": "entity_3" } ], "notes": [ "The supplied masklet consistently identifies the single person; no source masks were supplied for the wrench, surface, jack, mechanism, car, or tire.", "The source-video interval is [179.120000, 206.120000), corresponding to the 27.000-second clip-relative interval [0.000000, 27.000000).", "\"is resting\" was not annotated as an action because it describes the tire's resulting static contact configuration; its subject also has no supplied mask." ] }, "model_output_raw": { "caption": "After placing the impact wrench on the ground, the person crouches to operate the car jack. They use the jack's mechanism to slowly lower the car until the tire is resting firmly on the pavement.", "entity_groups": [ { "id": "entity_1", "label": "impact wrench", "spans": [ { "start": 14, "end": 31, "text": "the impact wrench" } ], "source_mask": null }, { "id": "entity_2", "label": "ground/pavement surface", "spans": [ { "start": 35, "end": 45, "text": "the ground" }, { "start": 182, "end": 194, "text": "the pavement" } ], "source_mask": null }, { "id": "entity_3", "label": "person", "spans": [ { "start": 47, "end": 57, "text": "the person" }, { "start": 92, "end": 96, "text": "They" } ], "source_mask": "masklet_1" }, { "id": "entity_4", "label": "car jack", "spans": [ { "start": 78, "end": 90, "text": "the car jack" }, { "start": 101, "end": 111, "text": "the jack's" } ], "source_mask": null }, { "id": "entity_5", "label": "jack mechanism", "spans": [ { "start": 101, "end": 121, "text": "the jack's mechanism" } ], "source_mask": null }, { "id": "entity_6", "label": "car", "spans": [ { "start": 138, "end": 145, "text": "the car" } ], "source_mask": null }, { "id": "entity_7", "label": "tire", "spans": [ { "start": 152, "end": 160, "text": "the tire" } ], "source_mask": null } ], "actions": [ { "id": "action_1", "mentions": [ { "text": "placing", "occurrence": 1, "ranges": [ [ 6, 13 ] ], "particle": { "text": null, "occurrence": 1 } } ], "type": "simple", "subject": "entity_3", "objects": [ "entity_1", "entity_2" ], "confidence": "medium", "temporal_status": "uncertain", "intervals": [], "evidence": "The mask identifies the person handling the impact wrench, but none of the sampled clean frames clearly shows the wrench being released onto the ground.", "mask_relation": "subject matches the supplied selected-person mask" }, { "id": "action_2", "mentions": [ { "text": "crouches", "occurrence": 1, "ranges": [ [ 58, 66 ] ], "particle": { "text": null, "occurrence": 1 } } ], "type": "posture", "subject": "entity_3", "objects": [], "confidence": "high", "temporal_status": "localized", "intervals": [ { "start_s": 0.0, "end_s": 27.0 } ], "evidence": "The masked person remains visibly crouched beside the wheel and jack in every sampled clean frame from 0.000 through 26.000 seconds.", "mask_relation": "subject matches the supplied selected-person mask" }, { "id": "action_3", "mentions": [ { "text": "operate", "occurrence": 1, "ranges": [ [ 70, 77 ] ], "particle": { "text": null, "occurrence": 1 } }, { "text": "use", "occurrence": 1, "ranges": [ [ 97, 100 ] ], "particle": { "text": null, "occurrence": 1 } } ], "type": "simple", "subject": "entity_3", "objects": [ "entity_4", "entity_5" ], "confidence": "high", "temporal_status": "localized", "intervals": [ { "start_s": 3.0, "end_s": 27.0 } ], "evidence": "From 3.000 through 26.000 seconds, the masked person repeatedly holds the impact wrench against the jack's drive point, visibly operating the jack mechanism.", "mask_relation": "subject matches the supplied selected-person mask" }, { "id": "action_4", "mentions": [ { "text": "lower", "occurrence": 1, "ranges": [ [ 132, 137 ] ], "particle": { "text": null, "occurrence": 1 } } ], "type": "simple", "subject": "entity_3", "objects": [ "entity_6" ], "confidence": "medium", "temporal_status": "localized", "intervals": [ { "start_s": 3.0, "end_s": 27.0 } ], "evidence": "While the masked person operates the jack from 3.000 through 26.000 seconds, the jack progressively compresses and the supported car and wheel descend toward the pavement.", "mask_relation": "subject matches the supplied selected-person mask" } ], "notes": [ "The supplied masklet consistently identifies the single person; no source masks were supplied for the wrench, surface, jack, mechanism, car, or tire.", "The source-video interval is [179.120000, 206.120000), corresponding to the 27.000-second clip-relative interval [0.000000, 27.000000).", "\"is resting\" was not annotated as an action because it describes the tire's resulting static contact configuration; its subject also has no supplied mask." ] }, "normalization_repairs": 0 }, { "source_video_id": "v_-0NfmccuL9Q", "caption_count_category": 8, "clip": { "caption_index": 8, "caption": "The person pulls the car jack completely out from underneath the car. They then pick up the impact wrench from the ground, stand up, and walk out of view to the right.", "source_interval_sec": [ 206.12, 215.01 ], "loom_clip_id": "v_-0NfmccuL9Q_6177_6444", "clip_frame_interval_4fps": [ 6177, 6444 ], "sample_count_1fps": 9, "sampled_frames": [ { "frame": "00000", "timestamp_sec": 206.12, "offset_sec": 0, "rgb_path": "assets/cat_08/v_-0NfmccuL9Q_6177_6444/0000s_00000_rgb.jpg", "mask_path": "assets/cat_08/v_-0NfmccuL9Q_6177_6444/0000s_00000_mask.png", "mask_status": "available" }, { "frame": "00032", "timestamp_sec": 207.12, "offset_sec": 1, "rgb_path": "assets/cat_08/v_-0NfmccuL9Q_6177_6444/0001s_00032_rgb.jpg", "mask_path": "assets/cat_08/v_-0NfmccuL9Q_6177_6444/0001s_00032_mask.png", "mask_status": "available" }, { "frame": "00060", "timestamp_sec": 208.12, "offset_sec": 2, "rgb_path": "assets/cat_08/v_-0NfmccuL9Q_6177_6444/0002s_00060_rgb.jpg", "mask_path": "assets/cat_08/v_-0NfmccuL9Q_6177_6444/0002s_00060_mask.png", "mask_status": "available" }, { "frame": "00092", "timestamp_sec": 209.12, "offset_sec": 3, "rgb_path": "assets/cat_08/v_-0NfmccuL9Q_6177_6444/0003s_00092_rgb.jpg", "mask_path": "assets/cat_08/v_-0NfmccuL9Q_6177_6444/0003s_00092_mask.png", "mask_status": "available" }, { "frame": "00120", "timestamp_sec": 210.12, "offset_sec": 4, "rgb_path": "assets/cat_08/v_-0NfmccuL9Q_6177_6444/0004s_00120_rgb.jpg", "mask_path": "assets/cat_08/v_-0NfmccuL9Q_6177_6444/0004s_00120_mask.png", "mask_status": "available" }, { "frame": "00152", "timestamp_sec": 211.12, "offset_sec": 5, "rgb_path": "assets/cat_08/v_-0NfmccuL9Q_6177_6444/0005s_00152_rgb.jpg", "mask_path": "assets/cat_08/v_-0NfmccuL9Q_6177_6444/0005s_00152_mask.png", "mask_status": "available" }, { "frame": "00180", "timestamp_sec": 212.12, "offset_sec": 6, "rgb_path": "assets/cat_08/v_-0NfmccuL9Q_6177_6444/0006s_00180_rgb.jpg", "mask_path": "assets/cat_08/v_-0NfmccuL9Q_6177_6444/0006s_00180_mask.png", "mask_status": "available" }, { "frame": "00212", "timestamp_sec": 213.12, "offset_sec": 7, "rgb_path": "assets/cat_08/v_-0NfmccuL9Q_6177_6444/0007s_00212_rgb.jpg", "mask_path": "assets/cat_08/v_-0NfmccuL9Q_6177_6444/0007s_00212_mask.png", "mask_status": "available" }, { "frame": "00240", "timestamp_sec": 214.12, "offset_sec": 8, "rgb_path": "assets/cat_08/v_-0NfmccuL9Q_6177_6444/0008s_00240_rgb.jpg", "mask_path": "assets/cat_08/v_-0NfmccuL9Q_6177_6444/0008s_00240_mask.png", "mask_status": "available" } ], "obj_id": [ 0 ], "anno_id": [ 13 ], "fps_inferred_from_clip_interval": 30.0337 }, "usage": { "input_tokens": 59885, "input_tokens_details": { "cache_write_tokens": 59882, "cached_tokens": 0 }, "output_tokens": 4476, "output_tokens_details": { "reasoning_tokens": 3106 }, "total_tokens": 64361 }, "model": "gpt-5.6-sol", "effort": "medium", "custom_id": "cat08_cap08_v_-0NfmccuL9Q_6177_6444", "annotation": { "caption": "The person pulls the car jack completely out from underneath the car. They then pick up the impact wrench from the ground, stand up, and walk out of view to the right.", "entity_groups": [ { "id": "person_1", "label": "person", "spans": [ { "text": "The person", "range": [ 0, 10 ] }, { "text": "They", "range": [ 70, 74 ] } ], "source_mask_link": "selected_person" }, { "id": "car_jack_1", "label": "car jack", "spans": [ { "text": "the car jack", "range": [ 17, 29 ] } ], "source_mask_link": null }, { "id": "car_1", "label": "car", "spans": [ { "text": "the car", "range": [ 61, 68 ] } ], "source_mask_link": null }, { "id": "impact_wrench_1", "label": "impact wrench", "spans": [ { "text": "the impact wrench", "range": [ 88, 105 ] } ], "source_mask_link": null }, { "id": "ground_1", "label": "ground", "spans": [ { "text": "the ground", "range": [ 111, 121 ] } ], "source_mask_link": null } ], "actions": [ { "id": "action_1", "mentions": [ { "text": "pulls", "ranges": [ [ 11, 16 ] ], "occurrence": 1, "particle": { "text": "out", "occurrence": 1, "range": [ 41, 44 ] } } ], "type": "phrasal", "subject_refs": [ "person_1" ], "object_refs": [ "car_jack_1" ], "confidence": "high", "temporal_status": "supported", "intervals": [ { "start_s": 0.0, "end_s": 2.0, "source_start_s": 206.12, "source_end_s": 208.12 }, { "start_s": 5.0, "end_s": 7.0, "source_start_s": 211.12, "source_end_s": 213.12 } ], "evidence": "At 0–1 seconds the masked person manipulates the jack beneath the car; at 5–6 seconds he reaches toward and draws the jack farther outward.", "mask_relation": "subject matches the supplied selected-person mask", "source_subject_refs": [ "person_1" ] }, { "id": "action_2", "mentions": [ { "text": "pick up", "ranges": [ [ 80, 87 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "type": "phrasal", "subject_refs": [ "person_1" ], "object_refs": [ "impact_wrench_1" ], "confidence": "high", "temporal_status": "supported_at_clip_boundary", "intervals": [ { "start_s": 8.0, "end_s": 8.889999999999986, "source_start_s": 214.12, "source_end_s": 215.01 } ], "evidence": "In the final sampled frame, the masked person grips and lifts the impact wrench from the ground; support continues to the clip boundary.", "mask_relation": "subject matches the supplied selected-person mask", "source_subject_refs": [ "person_1" ] }, { "id": "action_3", "mentions": [ { "text": "stand up", "ranges": [ [ 123, 131 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "type": "phrasal", "subject_refs": [ "person_1" ], "object_refs": [], "confidence": "high", "temporal_status": "uncertain", "intervals": [], "evidence": "No sampled frame after the final wrench pickup clearly shows the captioned stand-up event.", "mask_relation": "subject matches the supplied selected-person mask", "source_subject_refs": [ "person_1" ] }, { "id": "action_4", "mentions": [ { "text": "walk", "ranges": [ [ 137, 141 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "type": "simple", "subject_refs": [ "person_1" ], "object_refs": [], "confidence": "high", "temporal_status": "uncertain", "intervals": [], "evidence": "The sampled frames do not clearly show the masked person walking out of view after picking up the wrench.", "mask_relation": "subject matches the supplied selected-person mask", "source_subject_refs": [ "person_1" ] } ], "notes": [ "The supplied masklet consistently identifies the captioned person.", "No separate source masks were supplied for the car jack, car, impact wrench, or ground.", "The last sampled frame occurs at 8.000 seconds, shortly before the 8.890-second clip boundary, so the later stand-up and walk-out actions cannot be temporally localized from sampled-frame evidence." ] }, "model_output_raw": { "caption": "The person pulls the car jack completely out from underneath the car. They then pick up the impact wrench from the ground, stand up, and walk out of view to the right.", "entity_groups": [ { "id": "person_1", "label": "person", "spans": [ { "text": "The person", "start": 0, "end": 10 }, { "text": "They", "start": 70, "end": 74 } ], "source_mask": "selected_person_masklet" }, { "id": "car_jack_1", "label": "car jack", "spans": [ { "text": "the car jack", "start": 17, "end": 29 } ], "source_mask": null }, { "id": "car_1", "label": "car", "spans": [ { "text": "the car", "start": 61, "end": 68 } ], "source_mask": null }, { "id": "impact_wrench_1", "label": "impact wrench", "spans": [ { "text": "the impact wrench", "start": 88, "end": 105 } ], "source_mask": null }, { "id": "ground_1", "label": "ground", "spans": [ { "text": "the ground", "start": 111, "end": 121 } ], "source_mask": null } ], "actions": [ { "id": "action_1", "mentions": [ { "text": "pulls", "occurrence": 1, "ranges": [ [ 11, 16 ] ], "particle": { "text": "out", "occurrence": 1, "range": [ 41, 44 ] } } ], "type": "phrasal", "subject_refs": [ "person_1" ], "object_refs": [ "car_jack_1" ], "confidence": "high", "mask_relation": "subject matches the supplied selected-person mask", "temporal_status": "supported", "segments": [ { "start_s": 0.0, "end_s": 2.0, "source_start_s": 206.12, "source_end_s": 208.12 }, { "start_s": 5.0, "end_s": 7.0, "source_start_s": 211.12, "source_end_s": 213.12 } ], "evidence": "At 0–1 seconds the masked person manipulates the jack beneath the car; at 5–6 seconds he reaches toward and draws the jack farther outward." }, { "id": "action_2", "mentions": [ { "text": "pick up", "occurrence": 1, "ranges": [ [ 80, 87 ] ], "particle": { "text": null, "occurrence": 1 } } ], "type": "phrasal", "subject_refs": [ "person_1" ], "object_refs": [ "impact_wrench_1" ], "confidence": "high", "mask_relation": "subject matches the supplied selected-person mask", "temporal_status": "supported_at_clip_boundary", "segments": [ { "start_s": 8.0, "end_s": 8.89, "source_start_s": 214.12, "source_end_s": 215.01 } ], "evidence": "In the final sampled frame, the masked person grips and lifts the impact wrench from the ground; support continues to the clip boundary." }, { "id": "action_3", "mentions": [ { "text": "stand up", "occurrence": 1, "ranges": [ [ 123, 131 ] ], "particle": { "text": null, "occurrence": 1 } } ], "type": "phrasal", "subject_refs": [ "person_1" ], "object_refs": [], "confidence": "high", "mask_relation": "subject matches the supplied selected-person mask", "temporal_status": "uncertain", "segments": [], "evidence": "No sampled frame after the final wrench pickup clearly shows the captioned stand-up event." }, { "id": "action_4", "mentions": [ { "text": "walk", "occurrence": 1, "ranges": [ [ 137, 141 ] ], "particle": { "text": null, "occurrence": 1 } } ], "type": "simple", "subject_refs": [ "person_1" ], "object_refs": [], "confidence": "high", "mask_relation": "subject matches the supplied selected-person mask", "temporal_status": "uncertain", "segments": [], "evidence": "The sampled frames do not clearly show the masked person walking out of view after picking up the wrench." } ], "notes": [ "The supplied masklet consistently identifies the captioned person.", "No separate source masks were supplied for the car jack, car, impact wrench, or ground.", "The last sampled frame occurs at 8.000 seconds, shortly before the 8.890-second clip boundary, so the later stand-up and walk-out actions cannot be temporally localized from sampled-frame evidence." ] }, "normalization_repairs": 0 }, { "source_video_id": "v_-ceM4HHukhE", "caption_count_category": 9, "clip": { "caption_index": 1, "caption": "A man wearing a black shirt and pants walks to a rowing machine and sits down on the seat, facing away from the flywheel.", "source_interval_sec": [ 0.0, 6.0 ], "loom_clip_id": "v_-ceM4HHukhE_1_180", "clip_frame_interval_4fps": [ 1, 180 ], "sample_count_1fps": 6, "sampled_frames": [ { "frame": "00000", "timestamp_sec": 0.0, "offset_sec": 0, "rgb_path": "assets/cat_09/v_-ceM4HHukhE_1_180/0000s_00000_rgb.jpg", "mask_path": "assets/cat_09/v_-ceM4HHukhE_1_180/0000s_00000_mask.png", "mask_status": "available" }, { "frame": "00028", "timestamp_sec": 1.0, "offset_sec": 1, "rgb_path": "assets/cat_09/v_-ceM4HHukhE_1_180/0001s_00028_rgb.jpg", "mask_path": "assets/cat_09/v_-ceM4HHukhE_1_180/0001s_00028_mask.png", "mask_status": "available" }, { "frame": "00060", "timestamp_sec": 2.0, "offset_sec": 2, "rgb_path": "assets/cat_09/v_-ceM4HHukhE_1_180/0002s_00060_rgb.jpg", "mask_path": "assets/cat_09/v_-ceM4HHukhE_1_180/0002s_00060_mask.png", "mask_status": "available" }, { "frame": "00088", "timestamp_sec": 3.0, "offset_sec": 3, "rgb_path": "assets/cat_09/v_-ceM4HHukhE_1_180/0003s_00088_rgb.jpg", "mask_path": "assets/cat_09/v_-ceM4HHukhE_1_180/0003s_00088_mask.png", "mask_status": "available" }, { "frame": "00120", "timestamp_sec": 4.0, "offset_sec": 4, "rgb_path": "assets/cat_09/v_-ceM4HHukhE_1_180/0004s_00120_rgb.jpg", "mask_path": "assets/cat_09/v_-ceM4HHukhE_1_180/0004s_00120_mask.png", "mask_status": "available" }, { "frame": "00148", "timestamp_sec": 5.0, "offset_sec": 5, "rgb_path": "assets/cat_09/v_-ceM4HHukhE_1_180/0005s_00148_rgb.jpg", "mask_path": "assets/cat_09/v_-ceM4HHukhE_1_180/0005s_00148_mask.png", "mask_status": "available" } ], "obj_id": [ 0 ], "anno_id": [ 95 ], "fps_inferred_from_clip_interval": 29.8333 }, "usage": { "input_tokens": 21479, "input_tokens_details": { "cache_write_tokens": 21476, "cached_tokens": 0 }, "output_tokens": 3400, "output_tokens_details": { "reasoning_tokens": 2070 }, "total_tokens": 24879 }, "model": "gpt-5.6-sol", "effort": "medium", "custom_id": "cat09_cap01_v_-ceM4HHukhE_1_180", "annotation": { "caption": "A man wearing a black shirt and pants walks to a rowing machine and sits down on the seat, facing away from the flywheel.", "entity_groups": [ { "id": "e1", "label": "man", "spans": [ { "text": "A man", "range": [ 0, 5 ] } ], "source_mask_link": "selected_person" }, { "id": "e2", "label": "shirt", "spans": [ { "text": "a black shirt", "range": [ 14, 27 ] } ], "source_mask_link": null }, { "id": "e3", "label": "pants", "spans": [ { "text": "pants", "range": [ 32, 37 ] } ], "source_mask_link": null }, { "id": "e4", "label": "rowing machine", "spans": [ { "text": "a rowing machine", "range": [ 47, 63 ] } ], "source_mask_link": null }, { "id": "e5", "label": "seat", "spans": [ { "text": "the seat", "range": [ 81, 89 ] } ], "source_mask_link": null }, { "id": "e6", "label": "flywheel", "spans": [ { "text": "the flywheel", "range": [ 108, 120 ] } ], "source_mask_link": null } ], "actions": [ { "id": "action_1", "mentions": [ { "text": "wearing", "ranges": [ [ 6, 13 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "type": "simple", "subject_refs": [ "e1" ], "object_refs": [ "e2", "e3" ], "confidence": "high", "temporal_status": "supported", "intervals": [ { "start_s": 0.0, "end_s": 6.0, "source_start_s": 0.0, "source_end_s": 6.0 } ], "evidence": "The selected man visibly wears a black shirt and black pants in every sampled clean frame.", "mask_relation": "subject_matches_selected_person_mask", "source_subject_refs": "e1" }, { "id": "action_2", "mentions": [ { "text": "walks", "ranges": [ [ 38, 43 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "type": "simple", "subject_refs": [ "e1" ], "object_refs": [], "confidence": "high", "temporal_status": "supported", "intervals": [ { "start_s": 0.0, "end_s": 2.0, "source_start_s": 0.0, "source_end_s": 2.0 } ], "evidence": "At 0 and 1 seconds, the masked man is upright and advancing toward the rowing machine.", "mask_relation": "subject_matches_selected_person_mask", "source_subject_refs": "e1" }, { "id": "action_3", "mentions": [ { "text": "sits down", "ranges": [ [ 68, 77 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "type": "posture", "subject_refs": [ "e1" ], "object_refs": [], "confidence": "high", "temporal_status": "supported", "intervals": [ { "start_s": 2.0, "end_s": 5.0, "source_start_s": 2.0, "source_end_s": 5.0 } ], "evidence": "From 2 through 4 seconds, the masked man bends and lowers himself onto the machine's seat, becoming seated.", "mask_relation": "subject_matches_selected_person_mask", "source_subject_refs": "e1" }, { "id": "action_4", "mentions": [ { "text": "facing", "ranges": [ [ 91, 97 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "type": "posture", "subject_refs": [ "e1" ], "object_refs": [], "confidence": "high", "temporal_status": "supported", "intervals": [ { "start_s": 4.0, "end_s": 6.0, "source_start_s": 4.0, "source_end_s": 6.0 } ], "evidence": "At 4 and 5 seconds, the seated masked man is oriented away from the rowing machine's flywheel, and this posture continues through the final sample.", "mask_relation": "subject_matches_selected_person_mask", "source_subject_refs": "e1" } ], "notes": [ "The supplied mask consistently identifies the foreground man described by the caption.", "Only the man has a supplied source mask. The shirt, pants, rowing machine, seat, and flywheel are text-groundable but have null source-mask links.", "Action intervals are sparse support intervals derived from the sampled clean frames, not exact continuous-video boundaries." ] }, "model_output_raw": { "caption": "A man wearing a black shirt and pants walks to a rowing machine and sits down on the seat, facing away from the flywheel.", "entity_groups": [ { "id": "e1", "entity": "man", "spans": [ { "text": "A man", "range": [ 0, 5 ] } ], "source_mask": "selected_person" }, { "id": "e2", "entity": "shirt", "spans": [ { "text": "a black shirt", "range": [ 14, 27 ] } ], "source_mask": null }, { "id": "e3", "entity": "pants", "spans": [ { "text": "pants", "range": [ 32, 37 ] } ], "source_mask": null }, { "id": "e4", "entity": "rowing machine", "spans": [ { "text": "a rowing machine", "range": [ 47, 63 ] } ], "source_mask": null }, { "id": "e5", "entity": "seat", "spans": [ { "text": "the seat", "range": [ 81, 89 ] } ], "source_mask": null }, { "id": "e6", "entity": "flywheel", "spans": [ { "text": "the flywheel", "range": [ 108, 120 ] } ], "source_mask": null } ], "actions": [ { "mentions": [ { "text": "wearing", "ranges": [ [ 6, 13 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "type": "simple", "subject_entity_group_id": "e1", "object_entity_group_ids": [ "e2", "e3" ], "confidence": "high", "temporal_status": "supported", "intervals": [ { "clip_relative_s": [ 0.0, 6.0 ], "source_video_s": [ 0.0, 6.0 ] } ], "mask_relation": "subject_matches_selected_person_mask", "evidence": "The selected man visibly wears a black shirt and black pants in every sampled clean frame." }, { "mentions": [ { "text": "walks", "ranges": [ [ 38, 43 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "type": "simple", "subject_entity_group_id": "e1", "object_entity_group_ids": [], "arguments": [ { "role": "destination", "entity_group_id": "e4" } ], "confidence": "high", "temporal_status": "supported", "intervals": [ { "clip_relative_s": [ 0.0, 2.0 ], "source_video_s": [ 0.0, 2.0 ] } ], "mask_relation": "subject_matches_selected_person_mask", "evidence": "At 0 and 1 seconds, the masked man is upright and advancing toward the rowing machine." }, { "mentions": [ { "text": "sits down", "ranges": [ [ 68, 77 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "type": "posture", "subject_entity_group_id": "e1", "object_entity_group_ids": [], "arguments": [ { "role": "location", "entity_group_id": "e5" } ], "confidence": "high", "temporal_status": "supported", "intervals": [ { "clip_relative_s": [ 2.0, 5.0 ], "source_video_s": [ 2.0, 5.0 ] } ], "mask_relation": "subject_matches_selected_person_mask", "evidence": "From 2 through 4 seconds, the masked man bends and lowers himself onto the machine's seat, becoming seated." }, { "mentions": [ { "text": "facing", "ranges": [ [ 91, 97 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "type": "posture", "subject_entity_group_id": "e1", "object_entity_group_ids": [], "arguments": [ { "role": "orientation_reference", "entity_group_id": "e6" } ], "confidence": "high", "temporal_status": "supported", "intervals": [ { "clip_relative_s": [ 4.0, 6.0 ], "source_video_s": [ 4.0, 6.0 ] } ], "mask_relation": "subject_matches_selected_person_mask", "evidence": "At 4 and 5 seconds, the seated masked man is oriented away from the rowing machine's flywheel, and this posture continues through the final sample." } ], "notes": [ "The supplied mask consistently identifies the foreground man described by the caption.", "Only the man has a supplied source mask. The shirt, pants, rowing machine, seat, and flywheel are text-groundable but have null source-mask links.", "Action intervals are sparse support intervals derived from the sampled clean frames, not exact continuous-video boundaries." ] }, "normalization_repairs": 0 }, { "source_video_id": "v_-ceM4HHukhE", "caption_count_category": 9, "clip": { "caption_index": 2, "caption": "He sits upright on the machine with his hands on the seat beside him, talking and gesturing as he looks forward.", "source_interval_sec": [ 6.0, 19.5 ], "loom_clip_id": "v_-ceM4HHukhE_180_584", "clip_frame_interval_4fps": [ 180, 584 ], "sample_count_1fps": 14, "sampled_frames": [ { "frame": "00000", "timestamp_sec": 6.0, "offset_sec": 0, "rgb_path": "assets/cat_09/v_-ceM4HHukhE_180_584/0000s_00000_rgb.jpg", "mask_path": "assets/cat_09/v_-ceM4HHukhE_180_584/0000s_00000_mask.png", "mask_status": "available" }, { "frame": "00028", "timestamp_sec": 7.0, "offset_sec": 1, "rgb_path": "assets/cat_09/v_-ceM4HHukhE_180_584/0001s_00028_rgb.jpg", "mask_path": "assets/cat_09/v_-ceM4HHukhE_180_584/0001s_00028_mask.png", "mask_status": "available" }, { "frame": "00060", "timestamp_sec": 8.0, "offset_sec": 2, "rgb_path": "assets/cat_09/v_-ceM4HHukhE_180_584/0002s_00060_rgb.jpg", "mask_path": "assets/cat_09/v_-ceM4HHukhE_180_584/0002s_00060_mask.png", "mask_status": "available" }, { "frame": "00088", "timestamp_sec": 9.0, "offset_sec": 3, "rgb_path": "assets/cat_09/v_-ceM4HHukhE_180_584/0003s_00088_rgb.jpg", "mask_path": "assets/cat_09/v_-ceM4HHukhE_180_584/0003s_00088_mask.png", "mask_status": "available" }, { "frame": "00120", "timestamp_sec": 10.0, "offset_sec": 4, "rgb_path": "assets/cat_09/v_-ceM4HHukhE_180_584/0004s_00120_rgb.jpg", "mask_path": "assets/cat_09/v_-ceM4HHukhE_180_584/0004s_00120_mask.png", "mask_status": "available" }, { "frame": "00148", "timestamp_sec": 11.0, "offset_sec": 5, "rgb_path": "assets/cat_09/v_-ceM4HHukhE_180_584/0005s_00148_rgb.jpg", "mask_path": "assets/cat_09/v_-ceM4HHukhE_180_584/0005s_00148_mask.png", "mask_status": "available" }, { "frame": "00180", "timestamp_sec": 12.0, "offset_sec": 6, "rgb_path": "assets/cat_09/v_-ceM4HHukhE_180_584/0006s_00180_rgb.jpg", "mask_path": "assets/cat_09/v_-ceM4HHukhE_180_584/0006s_00180_mask.png", "mask_status": "available" }, { "frame": "00208", "timestamp_sec": 13.0, "offset_sec": 7, "rgb_path": "assets/cat_09/v_-ceM4HHukhE_180_584/0007s_00208_rgb.jpg", "mask_path": "assets/cat_09/v_-ceM4HHukhE_180_584/0007s_00208_mask.png", "mask_status": "available" }, { "frame": "00240", "timestamp_sec": 14.0, "offset_sec": 8, "rgb_path": "assets/cat_09/v_-ceM4HHukhE_180_584/0008s_00240_rgb.jpg", "mask_path": "assets/cat_09/v_-ceM4HHukhE_180_584/0008s_00240_mask.png", "mask_status": "available" }, { "frame": "00268", "timestamp_sec": 15.0, "offset_sec": 9, "rgb_path": "assets/cat_09/v_-ceM4HHukhE_180_584/0009s_00268_rgb.jpg", "mask_path": "assets/cat_09/v_-ceM4HHukhE_180_584/0009s_00268_mask.png", "mask_status": "available" }, { "frame": "00300", "timestamp_sec": 16.0, "offset_sec": 10, "rgb_path": "assets/cat_09/v_-ceM4HHukhE_180_584/0010s_00300_rgb.jpg", "mask_path": "assets/cat_09/v_-ceM4HHukhE_180_584/0010s_00300_mask.png", "mask_status": "available" }, { "frame": "00328", "timestamp_sec": 17.0, "offset_sec": 11, "rgb_path": "assets/cat_09/v_-ceM4HHukhE_180_584/0011s_00328_rgb.jpg", "mask_path": "assets/cat_09/v_-ceM4HHukhE_180_584/0011s_00328_mask.png", "mask_status": "available" }, { "frame": "00360", "timestamp_sec": 18.0, "offset_sec": 12, "rgb_path": "assets/cat_09/v_-ceM4HHukhE_180_584/0012s_00360_rgb.jpg", "mask_path": "assets/cat_09/v_-ceM4HHukhE_180_584/0012s_00360_mask.png", "mask_status": "available" }, { "frame": "00388", "timestamp_sec": 19.0, "offset_sec": 13, "rgb_path": "assets/cat_09/v_-ceM4HHukhE_180_584/0013s_00388_rgb.jpg", "mask_path": "assets/cat_09/v_-ceM4HHukhE_180_584/0013s_00388_mask.png", "mask_status": "available" } ], "obj_id": [ 0 ], "anno_id": [ 94 ], "fps_inferred_from_clip_interval": 29.9259 }, "usage": { "input_tokens": 29692, "input_tokens_details": { "cache_write_tokens": 29689, "cached_tokens": 0 }, "output_tokens": 3348, "output_tokens_details": { "reasoning_tokens": 2070 }, "total_tokens": 33040 }, "model": "gpt-5.6-sol", "effort": "medium", "custom_id": "cat09_cap02_v_-ceM4HHukhE_180_584", "annotation": { "caption": "He sits upright on the machine with his hands on the seat beside him, talking and gesturing as he looks forward.", "entity_groups": [ { "id": "person_1", "label": "person", "spans": [ { "text": "He", "range": [ 0, 2 ] }, { "text": "his", "range": [ 36, 39 ] }, { "text": "him", "range": [ 65, 68 ] }, { "text": "he", "range": [ 95, 97 ] } ], "source_mask_link": "selected_person" }, { "id": "machine_1", "label": "machine", "spans": [ { "text": "the machine", "range": [ 19, 30 ] } ], "source_mask_link": null }, { "id": "hands_1", "label": "hands", "spans": [ { "text": "his hands", "range": [ 36, 45 ] } ], "source_mask_link": null }, { "id": "seat_1", "label": "seat", "spans": [ { "text": "the seat", "range": [ 49, 57 ] } ], "source_mask_link": null } ], "actions": [ { "id": "action_1", "mentions": [ { "text": "sits", "ranges": [ [ 3, 7 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "type": "posture", "subject_refs": [ "person_1" ], "object_refs": [ "machine_1" ], "confidence": "high", "temporal_status": "localized", "intervals": [ { "start_s": 0.0, "end_s": 13.5, "source_start_s": 6.0, "source_end_s": 19.5 } ], "evidence": "The mask-linked man is visibly seated on the rowing machine in every sampled frame.", "mask_relation": "Action subject matches the supplied selected-person mask.", "source_subject_refs": "person_1" }, { "id": "action_2", "mentions": [ { "text": "talking", "ranges": [ [ 70, 77 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "type": "simple", "subject_refs": [ "person_1" ], "object_refs": [], "confidence": "medium", "temporal_status": "uncertain", "intervals": [], "evidence": "The caption assigns talking to the mask-linked man, but sparse still frames do not reliably establish speech timing.", "mask_relation": "Action subject matches the supplied selected-person mask.", "source_subject_refs": "person_1" }, { "id": "action_3", "mentions": [ { "text": "gesturing", "ranges": [ [ 82, 91 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "type": "simple", "subject_refs": [ "person_1" ], "object_refs": [], "confidence": "high", "temporal_status": "localized", "intervals": [ { "start_s": 8.0, "end_s": 13.5, "source_start_s": 14.0, "source_end_s": 19.5 } ], "evidence": "From the 8-second sample onward, the mask-linked man's hands lift from beside the seat and change position in front of his body; the final sample still supports the gesture.", "mask_relation": "Action subject matches the supplied selected-person mask.", "source_subject_refs": "person_1" }, { "id": "action_4", "mentions": [ { "text": "looks", "ranges": [ [ 98, 103 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "type": "simple", "subject_refs": [ "person_1" ], "object_refs": [], "confidence": "medium", "temporal_status": "localized", "intervals": [ { "start_s": 3.0, "end_s": 4.0, "source_start_s": 9.0, "source_end_s": 10.0 }, { "start_s": 6.0, "end_s": 7.0, "source_start_s": 12.0, "source_end_s": 13.0 }, { "start_s": 9.0, "end_s": 11.0, "source_start_s": 15.0, "source_end_s": 17.0 } ], "evidence": "The mask-linked man's head and face are oriented generally forward in the indicated samples; other samples show him looking down or to the side.", "mask_relation": "Action subject matches the supplied selected-person mask.", "source_subject_refs": "person_1" } ], "notes": [ "The supplied masklet consistently identifies the seated foreground man referred to by He/his/him/he.", "No separate masks were supplied for the rowing machine, hands, or seat.", "The direction modifier 'forward' and the posture modifier 'upright' are excluded from their action spans." ] }, "model_output_raw": { "caption": "He sits upright on the machine with his hands on the seat beside him, talking and gesturing as he looks forward.", "entity_groups": [ { "id": "person_1", "entity_type": "person", "spans": [ { "text": "He", "range": [ 0, 2 ] }, { "text": "his", "range": [ 36, 39 ] }, { "text": "him", "range": [ 65, 68 ] }, { "text": "he", "range": [ 95, 97 ] } ], "source_mask_link": "selected_person_masklet" }, { "id": "machine_1", "entity_type": "machine", "spans": [ { "text": "the machine", "range": [ 19, 30 ] } ], "source_mask_link": null }, { "id": "hands_1", "entity_type": "hands", "spans": [ { "text": "his hands", "range": [ 36, 45 ] } ], "source_mask_link": null }, { "id": "seat_1", "entity_type": "seat", "spans": [ { "text": "the seat", "range": [ 49, 57 ] } ], "source_mask_link": null } ], "actions": [ { "mentions": [ { "text": "sits", "ranges": [ [ 3, 7 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "type": "posture", "subject_ref": "person_1", "object_refs": [ "machine_1" ], "confidence": "high", "temporal_status": "localized", "intervals": [ { "start_s": 0.0, "end_s": 13.5, "source_start_s": 6.0, "source_end_s": 19.5 } ], "evidence": "The mask-linked man is visibly seated on the rowing machine in every sampled frame.", "mask_relation": "Action subject matches the supplied selected-person mask." }, { "mentions": [ { "text": "talking", "ranges": [ [ 70, 77 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "type": "simple", "subject_ref": "person_1", "object_refs": [], "confidence": "medium", "temporal_status": "uncertain", "intervals": [], "evidence": "The caption assigns talking to the mask-linked man, but sparse still frames do not reliably establish speech timing.", "mask_relation": "Action subject matches the supplied selected-person mask." }, { "mentions": [ { "text": "gesturing", "ranges": [ [ 82, 91 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "type": "simple", "subject_ref": "person_1", "object_refs": [], "confidence": "high", "temporal_status": "localized", "intervals": [ { "start_s": 8.0, "end_s": 13.5, "source_start_s": 14.0, "source_end_s": 19.5 } ], "evidence": "From the 8-second sample onward, the mask-linked man's hands lift from beside the seat and change position in front of his body; the final sample still supports the gesture.", "mask_relation": "Action subject matches the supplied selected-person mask." }, { "mentions": [ { "text": "looks", "ranges": [ [ 98, 103 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "type": "simple", "subject_ref": "person_1", "object_refs": [], "confidence": "medium", "temporal_status": "localized", "intervals": [ { "start_s": 3.0, "end_s": 4.0, "source_start_s": 9.0, "source_end_s": 10.0 }, { "start_s": 6.0, "end_s": 7.0, "source_start_s": 12.0, "source_end_s": 13.0 }, { "start_s": 9.0, "end_s": 11.0, "source_start_s": 15.0, "source_end_s": 17.0 } ], "evidence": "The mask-linked man's head and face are oriented generally forward in the indicated samples; other samples show him looking down or to the side.", "mask_relation": "Action subject matches the supplied selected-person mask." } ], "notes": [ "The supplied masklet consistently identifies the seated foreground man referred to by He/his/him/he.", "No separate masks were supplied for the rowing machine, hands, or seat.", "The direction modifier 'forward' and the posture modifier 'upright' are excluded from their action spans." ] }, "normalization_repairs": 0 }, { "source_video_id": "v_-ceM4HHukhE", "caption_count_category": 9, "clip": { "caption_index": 3, "caption": "The man leans forward to adjust the straps on the footrests, using both hands to secure his right foot and then his left foot.", "source_interval_sec": [ 19.5, 35.5 ], "loom_clip_id": "v_-ceM4HHukhE_584_1064", "clip_frame_interval_4fps": [ 584, 1064 ], "sample_count_1fps": 16, "sampled_frames": [ { "frame": "00000", "timestamp_sec": 19.5, "offset_sec": 0, "rgb_path": "assets/cat_09/v_-ceM4HHukhE_584_1064/0000s_00000_rgb.jpg", "mask_path": "assets/cat_09/v_-ceM4HHukhE_584_1064/0000s_00000_mask.png", "mask_status": "available" }, { "frame": "00028", "timestamp_sec": 20.5, "offset_sec": 1, "rgb_path": "assets/cat_09/v_-ceM4HHukhE_584_1064/0001s_00028_rgb.jpg", "mask_path": "assets/cat_09/v_-ceM4HHukhE_584_1064/0001s_00028_mask.png", "mask_status": "available" }, { "frame": "00060", "timestamp_sec": 21.5, "offset_sec": 2, "rgb_path": "assets/cat_09/v_-ceM4HHukhE_584_1064/0002s_00060_rgb.jpg", "mask_path": "assets/cat_09/v_-ceM4HHukhE_584_1064/0002s_00060_mask.png", "mask_status": "available" }, { "frame": "00088", "timestamp_sec": 22.5, "offset_sec": 3, "rgb_path": "assets/cat_09/v_-ceM4HHukhE_584_1064/0003s_00088_rgb.jpg", "mask_path": "assets/cat_09/v_-ceM4HHukhE_584_1064/0003s_00088_mask.png", "mask_status": "available" }, { "frame": "00120", "timestamp_sec": 23.5, "offset_sec": 4, "rgb_path": "assets/cat_09/v_-ceM4HHukhE_584_1064/0004s_00120_rgb.jpg", "mask_path": "assets/cat_09/v_-ceM4HHukhE_584_1064/0004s_00120_mask.png", "mask_status": "available" }, { "frame": "00148", "timestamp_sec": 24.5, "offset_sec": 5, "rgb_path": "assets/cat_09/v_-ceM4HHukhE_584_1064/0005s_00148_rgb.jpg", "mask_path": "assets/cat_09/v_-ceM4HHukhE_584_1064/0005s_00148_mask.png", "mask_status": "available" }, { "frame": "00180", "timestamp_sec": 25.5, "offset_sec": 6, "rgb_path": "assets/cat_09/v_-ceM4HHukhE_584_1064/0006s_00180_rgb.jpg", "mask_path": "assets/cat_09/v_-ceM4HHukhE_584_1064/0006s_00180_mask.png", "mask_status": "available" }, { "frame": "00208", "timestamp_sec": 26.5, "offset_sec": 7, "rgb_path": "assets/cat_09/v_-ceM4HHukhE_584_1064/0007s_00208_rgb.jpg", "mask_path": "assets/cat_09/v_-ceM4HHukhE_584_1064/0007s_00208_mask.png", "mask_status": "available" }, { "frame": "00240", "timestamp_sec": 27.5, "offset_sec": 8, "rgb_path": "assets/cat_09/v_-ceM4HHukhE_584_1064/0008s_00240_rgb.jpg", "mask_path": "assets/cat_09/v_-ceM4HHukhE_584_1064/0008s_00240_mask.png", "mask_status": "available" }, { "frame": "00268", "timestamp_sec": 28.5, "offset_sec": 9, "rgb_path": "assets/cat_09/v_-ceM4HHukhE_584_1064/0009s_00268_rgb.jpg", "mask_path": "assets/cat_09/v_-ceM4HHukhE_584_1064/0009s_00268_mask.png", "mask_status": "available" }, { "frame": "00300", "timestamp_sec": 29.5, "offset_sec": 10, "rgb_path": "assets/cat_09/v_-ceM4HHukhE_584_1064/0010s_00300_rgb.jpg", "mask_path": "assets/cat_09/v_-ceM4HHukhE_584_1064/0010s_00300_mask.png", "mask_status": "available" }, { "frame": "00328", "timestamp_sec": 30.5, "offset_sec": 11, "rgb_path": "assets/cat_09/v_-ceM4HHukhE_584_1064/0011s_00328_rgb.jpg", "mask_path": "assets/cat_09/v_-ceM4HHukhE_584_1064/0011s_00328_mask.png", "mask_status": "available" }, { "frame": "00360", "timestamp_sec": 31.5, "offset_sec": 12, "rgb_path": "assets/cat_09/v_-ceM4HHukhE_584_1064/0012s_00360_rgb.jpg", "mask_path": "assets/cat_09/v_-ceM4HHukhE_584_1064/0012s_00360_mask.png", "mask_status": "available" }, { "frame": "00388", "timestamp_sec": 32.5, "offset_sec": 13, "rgb_path": "assets/cat_09/v_-ceM4HHukhE_584_1064/0013s_00388_rgb.jpg", "mask_path": "assets/cat_09/v_-ceM4HHukhE_584_1064/0013s_00388_mask.png", "mask_status": "available" }, { "frame": "00420", "timestamp_sec": 33.5, "offset_sec": 14, "rgb_path": "assets/cat_09/v_-ceM4HHukhE_584_1064/0014s_00420_rgb.jpg", "mask_path": "assets/cat_09/v_-ceM4HHukhE_584_1064/0014s_00420_mask.png", "mask_status": "available" }, { "frame": "00448", "timestamp_sec": 34.5, "offset_sec": 15, "rgb_path": "assets/cat_09/v_-ceM4HHukhE_584_1064/0015s_00448_rgb.jpg", "mask_path": "assets/cat_09/v_-ceM4HHukhE_584_1064/0015s_00448_mask.png", "mask_status": "available" } ], "obj_id": [ 0 ], "anno_id": [ 100 ], "fps_inferred_from_clip_interval": 30.0 }, "usage": { "input_tokens": 31750, "input_tokens_details": { "cache_write_tokens": 31747, "cached_tokens": 0 }, "output_tokens": 3612, "output_tokens_details": { "reasoning_tokens": 2070 }, "total_tokens": 35362 }, "model": "gpt-5.6-sol", "effort": "medium", "custom_id": "cat09_cap03_v_-ceM4HHukhE_584_1064", "annotation": { "caption": "The man leans forward to adjust the straps on the footrests, using both hands to secure his right foot and then his left foot.", "entity_groups": [ { "id": "person_1", "label": "man", "spans": [ { "text": "The man", "range": [ 0, 7 ] }, { "text": "his", "range": [ 88, 91 ] }, { "text": "his", "range": [ 112, 115 ] } ], "source_mask_link": "selected_person" }, { "id": "straps_1", "label": "straps", "spans": [ { "text": "the straps", "range": [ 32, 42 ] } ], "source_mask_link": null }, { "id": "footrests_1", "label": "footrests", "spans": [ { "text": "the footrests", "range": [ 46, 59 ] } ], "source_mask_link": null }, { "id": "hands_1", "label": "hands", "spans": [ { "text": "both hands", "range": [ 67, 77 ] } ], "source_mask_link": null }, { "id": "right_foot_1", "label": "right foot", "spans": [ { "text": "his right foot", "range": [ 88, 102 ] } ], "source_mask_link": null }, { "id": "left_foot_1", "label": "left foot", "spans": [ { "text": "his left foot", "range": [ 112, 125 ] } ], "source_mask_link": null } ], "actions": [ { "id": "action_1", "mentions": [ { "text": "leans", "ranges": [ [ 8, 13 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "type": "posture", "subject_refs": [ "person_1" ], "object_refs": [], "confidence": "high", "temporal_status": "localized", "intervals": [ { "start_s": 1.0, "end_s": 9.0, "source_start_s": 20.5, "source_end_s": 28.5 } ], "evidence": "At clip-relative 1–8 seconds, the masked man bends and remains leaned forward over the footrests; he is upright again at 9 seconds.", "mask_relation": "subject_matches_selected_person_mask", "source_subject_refs": [ "person_1" ] }, { "id": "action_2", "mentions": [ { "text": "adjust", "ranges": [ [ 25, 31 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "type": "simple", "subject_refs": [ "person_1" ], "object_refs": [ "straps_1" ], "confidence": "high", "temporal_status": "localized", "intervals": [ { "start_s": 2.0, "end_s": 9.0, "source_start_s": 21.5, "source_end_s": 28.5 }, { "start_s": 12.0, "end_s": 13.0, "source_start_s": 31.5, "source_end_s": 32.5 } ], "evidence": "Frames at 2–8 seconds show the masked man manipulating a footrest strap with his hands; at 12 seconds he reaches to the other footrest/strap.", "mask_relation": "subject_matches_selected_person_mask", "source_subject_refs": [ "person_1" ] }, { "id": "action_3", "mentions": [ { "text": "using", "ranges": [ [ 61, 66 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "type": "simple", "subject_refs": [ "person_1" ], "object_refs": [ "hands_1" ], "confidence": "high", "temporal_status": "localized", "intervals": [ { "start_s": 1.0, "end_s": 9.0, "source_start_s": 20.5, "source_end_s": 28.5 } ], "evidence": "At 1–8 seconds, both hands of the masked man reach toward and manipulate the footrest area.", "mask_relation": "subject_matches_selected_person_mask", "source_subject_refs": [ "person_1" ] }, { "id": "action_4", "mentions": [ { "text": "secure", "ranges": [ [ 81, 87 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "type": "simple", "subject_refs": [ "person_1" ], "object_refs": [ "right_foot_1", "left_foot_1" ], "confidence": "medium", "temporal_status": "localized", "intervals": [ { "start_s": 2.0, "end_s": 9.0, "source_start_s": 21.5, "source_end_s": 28.5 }, { "start_s": 10.0, "end_s": 13.0, "source_start_s": 29.5, "source_end_s": 32.5 } ], "evidence": "The masked man manipulates the strap around one foot at 2–8 seconds, then raises, places, and attends to the other foot at 10–12 seconds.", "mask_relation": "subject_matches_selected_person_mask", "source_subject_refs": [ "person_1" ] } ], "notes": [ "The supplied mask consistently identifies the foreground man described by “The man.”", "No separate masks are supplied for the straps, footrests, hands, or feet, so those entity groups have null source-mask links.", "The two securing phases are sparsely supported by the sampled frames; exact continuous-video transition times cannot be determined." ] }, "model_output_raw": { "caption": "The man leans forward to adjust the straps on the footrests, using both hands to secure his right foot and then his left foot.", "entity_groups": [ { "id": "person_1", "label": "man", "spans": [ { "text": "The man", "range": [ 0, 7 ] }, { "text": "his", "range": [ 88, 91 ] }, { "text": "his", "range": [ 112, 115 ] } ], "source_mask": "selected_person_mask" }, { "id": "straps_1", "label": "straps", "spans": [ { "text": "the straps", "range": [ 32, 42 ] } ], "source_mask": null }, { "id": "footrests_1", "label": "footrests", "spans": [ { "text": "the footrests", "range": [ 46, 59 ] } ], "source_mask": null }, { "id": "hands_1", "label": "hands", "spans": [ { "text": "both hands", "range": [ 67, 77 ] } ], "source_mask": null }, { "id": "right_foot_1", "label": "right foot", "spans": [ { "text": "his right foot", "range": [ 88, 102 ] } ], "source_mask": null }, { "id": "left_foot_1", "label": "left foot", "spans": [ { "text": "his left foot", "range": [ 112, 125 ] } ], "source_mask": null } ], "actions": [ { "id": "action_1", "mentions": [ { "text": "leans", "ranges": [ [ 8, 13 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "type": "posture", "subject_refs": [ "person_1" ], "object_refs": [], "confidence": "high", "mask_relation": "subject_matches_selected_person_mask", "temporal_status": "localized", "intervals": [ { "start_s": 1.0, "end_s": 9.0, "source_start_s": 20.5, "source_end_s": 28.5 } ], "evidence": "At clip-relative 1–8 seconds, the masked man bends and remains leaned forward over the footrests; he is upright again at 9 seconds." }, { "id": "action_2", "mentions": [ { "text": "adjust", "ranges": [ [ 25, 31 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "type": "simple", "subject_refs": [ "person_1" ], "object_refs": [ "straps_1" ], "confidence": "high", "mask_relation": "subject_matches_selected_person_mask", "temporal_status": "localized", "intervals": [ { "start_s": 2.0, "end_s": 9.0, "source_start_s": 21.5, "source_end_s": 28.5 }, { "start_s": 12.0, "end_s": 13.0, "source_start_s": 31.5, "source_end_s": 32.5 } ], "evidence": "Frames at 2–8 seconds show the masked man manipulating a footrest strap with his hands; at 12 seconds he reaches to the other footrest/strap." }, { "id": "action_3", "mentions": [ { "text": "using", "ranges": [ [ 61, 66 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "type": "simple", "subject_refs": [ "person_1" ], "object_refs": [ "hands_1" ], "confidence": "high", "mask_relation": "subject_matches_selected_person_mask", "temporal_status": "localized", "intervals": [ { "start_s": 1.0, "end_s": 9.0, "source_start_s": 20.5, "source_end_s": 28.5 } ], "evidence": "At 1–8 seconds, both hands of the masked man reach toward and manipulate the footrest area." }, { "id": "action_4", "mentions": [ { "text": "secure", "ranges": [ [ 81, 87 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "type": "simple", "subject_refs": [ "person_1" ], "object_refs": [ "right_foot_1", "left_foot_1" ], "confidence": "medium", "mask_relation": "subject_matches_selected_person_mask", "temporal_status": "localized", "intervals": [ { "start_s": 2.0, "end_s": 9.0, "source_start_s": 21.5, "source_end_s": 28.5 }, { "start_s": 10.0, "end_s": 13.0, "source_start_s": 29.5, "source_end_s": 32.5 } ], "evidence": "The masked man manipulates the strap around one foot at 2–8 seconds, then raises, places, and attends to the other foot at 10–12 seconds." } ], "notes": [ "The supplied mask consistently identifies the foreground man described by “The man.”", "No separate masks are supplied for the straps, footrests, hands, or feet, so those entity groups have null source-mask links.", "The two securing phases are sparsely supported by the sampled frames; exact continuous-video transition times cannot be determined." ] }, "normalization_repairs": 0 }, { "source_video_id": "v_-ceM4HHukhE", "caption_count_category": 9, "clip": { "caption_index": 4, "caption": "Sitting upright with his feet now secured, he continues to talk and gesture, pointing towards the machine's monitor and clasping his hands.", "source_interval_sec": [ 35.5, 76.0 ], "loom_clip_id": "v_-ceM4HHukhE_1064_2278", "clip_frame_interval_4fps": [ 1064, 2278 ], "sample_count_1fps": 41, "sampled_frames": [ { "frame": "00000", "timestamp_sec": 35.5, "offset_sec": 0, "rgb_path": "assets/cat_09/v_-ceM4HHukhE_1064_2278/0000s_00000_rgb.jpg", "mask_path": "assets/cat_09/v_-ceM4HHukhE_1064_2278/0000s_00000_mask.png", "mask_status": "available" }, { "frame": "00028", "timestamp_sec": 36.5, "offset_sec": 1, "rgb_path": "assets/cat_09/v_-ceM4HHukhE_1064_2278/0001s_00028_rgb.jpg", "mask_path": "assets/cat_09/v_-ceM4HHukhE_1064_2278/0001s_00028_mask.png", "mask_status": "available" }, { "frame": "00060", "timestamp_sec": 37.5, "offset_sec": 2, "rgb_path": "assets/cat_09/v_-ceM4HHukhE_1064_2278/0002s_00060_rgb.jpg", "mask_path": "assets/cat_09/v_-ceM4HHukhE_1064_2278/0002s_00060_mask.png", "mask_status": "available" }, { "frame": "00088", "timestamp_sec": 38.5, "offset_sec": 3, "rgb_path": "assets/cat_09/v_-ceM4HHukhE_1064_2278/0003s_00088_rgb.jpg", "mask_path": "assets/cat_09/v_-ceM4HHukhE_1064_2278/0003s_00088_mask.png", "mask_status": "available" }, { "frame": "00120", "timestamp_sec": 39.5, "offset_sec": 4, "rgb_path": "assets/cat_09/v_-ceM4HHukhE_1064_2278/0004s_00120_rgb.jpg", "mask_path": "assets/cat_09/v_-ceM4HHukhE_1064_2278/0004s_00120_mask.png", "mask_status": "available" }, { "frame": "00148", "timestamp_sec": 40.5, "offset_sec": 5, "rgb_path": "assets/cat_09/v_-ceM4HHukhE_1064_2278/0005s_00148_rgb.jpg", "mask_path": "assets/cat_09/v_-ceM4HHukhE_1064_2278/0005s_00148_mask.png", "mask_status": "available" }, { "frame": "00180", "timestamp_sec": 41.5, "offset_sec": 6, "rgb_path": "assets/cat_09/v_-ceM4HHukhE_1064_2278/0006s_00180_rgb.jpg", "mask_path": "assets/cat_09/v_-ceM4HHukhE_1064_2278/0006s_00180_mask.png", "mask_status": "available" }, { "frame": "00208", "timestamp_sec": 42.5, "offset_sec": 7, "rgb_path": "assets/cat_09/v_-ceM4HHukhE_1064_2278/0007s_00208_rgb.jpg", "mask_path": "assets/cat_09/v_-ceM4HHukhE_1064_2278/0007s_00208_mask.png", "mask_status": "available" }, { "frame": "00240", "timestamp_sec": 43.5, "offset_sec": 8, "rgb_path": "assets/cat_09/v_-ceM4HHukhE_1064_2278/0008s_00240_rgb.jpg", "mask_path": "assets/cat_09/v_-ceM4HHukhE_1064_2278/0008s_00240_mask.png", "mask_status": "available" }, { "frame": "00268", "timestamp_sec": 44.5, "offset_sec": 9, "rgb_path": "assets/cat_09/v_-ceM4HHukhE_1064_2278/0009s_00268_rgb.jpg", "mask_path": "assets/cat_09/v_-ceM4HHukhE_1064_2278/0009s_00268_mask.png", "mask_status": "available" }, { "frame": "00300", "timestamp_sec": 45.5, "offset_sec": 10, "rgb_path": "assets/cat_09/v_-ceM4HHukhE_1064_2278/0010s_00300_rgb.jpg", "mask_path": "assets/cat_09/v_-ceM4HHukhE_1064_2278/0010s_00300_mask.png", "mask_status": "available" }, { "frame": "00328", "timestamp_sec": 46.5, "offset_sec": 11, "rgb_path": "assets/cat_09/v_-ceM4HHukhE_1064_2278/0011s_00328_rgb.jpg", "mask_path": "assets/cat_09/v_-ceM4HHukhE_1064_2278/0011s_00328_mask.png", "mask_status": "available" }, { "frame": "00360", "timestamp_sec": 47.5, "offset_sec": 12, "rgb_path": "assets/cat_09/v_-ceM4HHukhE_1064_2278/0012s_00360_rgb.jpg", "mask_path": "assets/cat_09/v_-ceM4HHukhE_1064_2278/0012s_00360_mask.png", "mask_status": "available" }, { "frame": "00388", "timestamp_sec": 48.5, "offset_sec": 13, "rgb_path": "assets/cat_09/v_-ceM4HHukhE_1064_2278/0013s_00388_rgb.jpg", "mask_path": "assets/cat_09/v_-ceM4HHukhE_1064_2278/0013s_00388_mask.png", "mask_status": "available" }, { "frame": "00420", "timestamp_sec": 49.5, "offset_sec": 14, "rgb_path": "assets/cat_09/v_-ceM4HHukhE_1064_2278/0014s_00420_rgb.jpg", "mask_path": "assets/cat_09/v_-ceM4HHukhE_1064_2278/0014s_00420_mask.png", "mask_status": "available" }, { "frame": "00448", "timestamp_sec": 50.5, "offset_sec": 15, "rgb_path": "assets/cat_09/v_-ceM4HHukhE_1064_2278/0015s_00448_rgb.jpg", "mask_path": "assets/cat_09/v_-ceM4HHukhE_1064_2278/0015s_00448_mask.png", "mask_status": "available" }, { "frame": "00480", "timestamp_sec": 51.5, "offset_sec": 16, "rgb_path": "assets/cat_09/v_-ceM4HHukhE_1064_2278/0016s_00480_rgb.jpg", "mask_path": "assets/cat_09/v_-ceM4HHukhE_1064_2278/0016s_00480_mask.png", "mask_status": "available" }, { "frame": "00508", "timestamp_sec": 52.5, "offset_sec": 17, "rgb_path": "assets/cat_09/v_-ceM4HHukhE_1064_2278/0017s_00508_rgb.jpg", "mask_path": "assets/cat_09/v_-ceM4HHukhE_1064_2278/0017s_00508_mask.png", "mask_status": "available" }, { "frame": "00540", "timestamp_sec": 53.5, "offset_sec": 18, "rgb_path": "assets/cat_09/v_-ceM4HHukhE_1064_2278/0018s_00540_rgb.jpg", "mask_path": "assets/cat_09/v_-ceM4HHukhE_1064_2278/0018s_00540_mask.png", "mask_status": "available" }, { "frame": "00568", "timestamp_sec": 54.5, "offset_sec": 19, "rgb_path": "assets/cat_09/v_-ceM4HHukhE_1064_2278/0019s_00568_rgb.jpg", "mask_path": "assets/cat_09/v_-ceM4HHukhE_1064_2278/0019s_00568_mask.png", "mask_status": "available" }, { "frame": "00600", "timestamp_sec": 55.5, "offset_sec": 20, "rgb_path": "assets/cat_09/v_-ceM4HHukhE_1064_2278/0020s_00600_rgb.jpg", "mask_path": "assets/cat_09/v_-ceM4HHukhE_1064_2278/0020s_00600_mask.png", "mask_status": "available" }, { "frame": "00628", "timestamp_sec": 56.5, "offset_sec": 21, "rgb_path": "assets/cat_09/v_-ceM4HHukhE_1064_2278/0021s_00628_rgb.jpg", "mask_path": "assets/cat_09/v_-ceM4HHukhE_1064_2278/0021s_00628_mask.png", "mask_status": "available" }, { "frame": "00660", "timestamp_sec": 57.5, "offset_sec": 22, "rgb_path": "assets/cat_09/v_-ceM4HHukhE_1064_2278/0022s_00660_rgb.jpg", "mask_path": "assets/cat_09/v_-ceM4HHukhE_1064_2278/0022s_00660_mask.png", "mask_status": "available" }, { "frame": "00688", "timestamp_sec": 58.5, "offset_sec": 23, "rgb_path": "assets/cat_09/v_-ceM4HHukhE_1064_2278/0023s_00688_rgb.jpg", "mask_path": "assets/cat_09/v_-ceM4HHukhE_1064_2278/0023s_00688_mask.png", "mask_status": "available" }, { "frame": "00720", "timestamp_sec": 59.5, "offset_sec": 24, "rgb_path": "assets/cat_09/v_-ceM4HHukhE_1064_2278/0024s_00720_rgb.jpg", "mask_path": "assets/cat_09/v_-ceM4HHukhE_1064_2278/0024s_00720_mask.png", "mask_status": "available" }, { "frame": "00748", "timestamp_sec": 60.5, "offset_sec": 25, "rgb_path": "assets/cat_09/v_-ceM4HHukhE_1064_2278/0025s_00748_rgb.jpg", "mask_path": "assets/cat_09/v_-ceM4HHukhE_1064_2278/0025s_00748_mask.png", "mask_status": "available" }, { "frame": "00780", "timestamp_sec": 61.5, "offset_sec": 26, "rgb_path": "assets/cat_09/v_-ceM4HHukhE_1064_2278/0026s_00780_rgb.jpg", "mask_path": "assets/cat_09/v_-ceM4HHukhE_1064_2278/0026s_00780_mask.png", "mask_status": "available" }, { "frame": "00808", "timestamp_sec": 62.5, "offset_sec": 27, "rgb_path": "assets/cat_09/v_-ceM4HHukhE_1064_2278/0027s_00808_rgb.jpg", "mask_path": "assets/cat_09/v_-ceM4HHukhE_1064_2278/0027s_00808_mask.png", "mask_status": "available" }, { "frame": "00840", "timestamp_sec": 63.5, "offset_sec": 28, "rgb_path": "assets/cat_09/v_-ceM4HHukhE_1064_2278/0028s_00840_rgb.jpg", "mask_path": "assets/cat_09/v_-ceM4HHukhE_1064_2278/0028s_00840_mask.png", "mask_status": "available" }, { "frame": "00868", "timestamp_sec": 64.5, "offset_sec": 29, "rgb_path": "assets/cat_09/v_-ceM4HHukhE_1064_2278/0029s_00868_rgb.jpg", "mask_path": "assets/cat_09/v_-ceM4HHukhE_1064_2278/0029s_00868_mask.png", "mask_status": "available" }, { "frame": "00900", "timestamp_sec": 65.5, "offset_sec": 30, "rgb_path": "assets/cat_09/v_-ceM4HHukhE_1064_2278/0030s_00900_rgb.jpg", "mask_path": "assets/cat_09/v_-ceM4HHukhE_1064_2278/0030s_00900_mask.png", "mask_status": "available" }, { "frame": "00928", "timestamp_sec": 66.5, "offset_sec": 31, "rgb_path": "assets/cat_09/v_-ceM4HHukhE_1064_2278/0031s_00928_rgb.jpg", "mask_path": "assets/cat_09/v_-ceM4HHukhE_1064_2278/0031s_00928_mask.png", "mask_status": "available" }, { "frame": "00960", "timestamp_sec": 67.5, "offset_sec": 32, "rgb_path": "assets/cat_09/v_-ceM4HHukhE_1064_2278/0032s_00960_rgb.jpg", "mask_path": "assets/cat_09/v_-ceM4HHukhE_1064_2278/0032s_00960_mask.png", "mask_status": "available" }, { "frame": "00988", "timestamp_sec": 68.5, "offset_sec": 33, "rgb_path": "assets/cat_09/v_-ceM4HHukhE_1064_2278/0033s_00988_rgb.jpg", "mask_path": "assets/cat_09/v_-ceM4HHukhE_1064_2278/0033s_00988_mask.png", "mask_status": "available" }, { "frame": "01020", "timestamp_sec": 69.5, "offset_sec": 34, "rgb_path": "assets/cat_09/v_-ceM4HHukhE_1064_2278/0034s_01020_rgb.jpg", "mask_path": "assets/cat_09/v_-ceM4HHukhE_1064_2278/0034s_01020_mask.png", "mask_status": "available" }, { "frame": "01048", "timestamp_sec": 70.5, "offset_sec": 35, "rgb_path": "assets/cat_09/v_-ceM4HHukhE_1064_2278/0035s_01048_rgb.jpg", "mask_path": "assets/cat_09/v_-ceM4HHukhE_1064_2278/0035s_01048_mask.png", "mask_status": "available" }, { "frame": "01080", "timestamp_sec": 71.5, "offset_sec": 36, "rgb_path": "assets/cat_09/v_-ceM4HHukhE_1064_2278/0036s_01080_rgb.jpg", "mask_path": "assets/cat_09/v_-ceM4HHukhE_1064_2278/0036s_01080_mask.png", "mask_status": "available" }, { "frame": "01108", "timestamp_sec": 72.5, "offset_sec": 37, "rgb_path": "assets/cat_09/v_-ceM4HHukhE_1064_2278/0037s_01108_rgb.jpg", "mask_path": "assets/cat_09/v_-ceM4HHukhE_1064_2278/0037s_01108_mask.png", "mask_status": "available" }, { "frame": "01140", "timestamp_sec": 73.5, "offset_sec": 38, "rgb_path": "assets/cat_09/v_-ceM4HHukhE_1064_2278/0038s_01140_rgb.jpg", "mask_path": "assets/cat_09/v_-ceM4HHukhE_1064_2278/0038s_01140_mask.png", "mask_status": "available" }, { "frame": "01168", "timestamp_sec": 74.5, "offset_sec": 39, "rgb_path": "assets/cat_09/v_-ceM4HHukhE_1064_2278/0039s_01168_rgb.jpg", "mask_path": "assets/cat_09/v_-ceM4HHukhE_1064_2278/0039s_01168_mask.png", "mask_status": "available" }, { "frame": "01200", "timestamp_sec": 75.5, "offset_sec": 40, "rgb_path": "assets/cat_09/v_-ceM4HHukhE_1064_2278/0040s_01200_rgb.jpg", "mask_path": "assets/cat_09/v_-ceM4HHukhE_1064_2278/0040s_01200_mask.png", "mask_status": "available" } ], "obj_id": [ 0 ], "anno_id": [ 93 ], "fps_inferred_from_clip_interval": 29.9753 }, "usage": { "input_tokens": 57425, "input_tokens_details": { "cache_write_tokens": 57422, "cached_tokens": 0 }, "output_tokens": 3557, "output_tokens_details": { "reasoning_tokens": 2066 }, "total_tokens": 60982 }, "model": "gpt-5.6-sol", "effort": "medium", "custom_id": "cat09_cap04_v_-ceM4HHukhE_1064_2278", "annotation": { "caption": "Sitting upright with his feet now secured, he continues to talk and gesture, pointing towards the machine's monitor and clasping his hands.", "entity_groups": [ { "id": "entity_1", "label": "person", "spans": [ { "text": "his", "range": [ 21, 24 ] }, { "text": "he", "range": [ 43, 45 ] }, { "text": "his", "range": [ 129, 132 ] } ], "source_mask_link": "selected_person" }, { "id": "entity_2", "label": "feet", "spans": [ { "text": "his feet", "range": [ 21, 29 ] } ], "source_mask_link": null }, { "id": "entity_3", "label": "machine", "spans": [ { "text": "the machine's", "range": [ 94, 107 ] } ], "source_mask_link": null }, { "id": "entity_4", "label": "monitor", "spans": [ { "text": "the machine's monitor", "range": [ 94, 115 ] } ], "source_mask_link": null }, { "id": "entity_5", "label": "hands", "spans": [ { "text": "his hands", "range": [ 129, 138 ] } ], "source_mask_link": null } ], "actions": [ { "id": "action_1", "mentions": [ { "text": "Sitting", "ranges": [ [ 0, 7 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "type": "posture", "subject_refs": [ "entity_1" ], "object_refs": [], "confidence": "high", "temporal_status": "supported", "intervals": [ { "start_s": 0.0, "end_s": 40.5, "source_start_s": 35.5, "source_end_s": 76.0 } ], "evidence": "The masked person remains seated on the rowing machine throughout all sampled frames.", "mask_relation": "subject matches the supplied selected-person mask", "source_subject_refs": "entity_1" }, { "id": "action_2", "mentions": [ { "text": "talk", "ranges": [ [ 59, 63 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "type": "simple", "subject_refs": [ "entity_1" ], "object_refs": [], "confidence": "medium", "temporal_status": "supported", "intervals": [ { "start_s": 0.0, "end_s": 27.0, "source_start_s": 35.5, "source_end_s": 62.5 }, { "start_s": 34.0, "end_s": 36.0, "source_start_s": 69.5, "source_end_s": 71.5 }, { "start_s": 39.0, "end_s": 40.0, "source_start_s": 74.5, "source_end_s": 75.5 } ], "evidence": "Across these sampled frames, the masked person's mouth and facial positions visibly vary in a manner consistent with ongoing speech.", "mask_relation": "subject matches the supplied selected-person mask", "source_subject_refs": "entity_1" }, { "id": "action_3", "mentions": [ { "text": "gesture", "ranges": [ [ 68, 75 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } }, { "text": "pointing", "ranges": [ [ 77, 85 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "type": "simple", "subject_refs": [ "entity_1" ], "object_refs": [ "entity_4" ], "confidence": "high", "temporal_status": "supported", "intervals": [ { "start_s": 11.0, "end_s": 18.0, "source_start_s": 46.5, "source_end_s": 53.5 }, { "start_s": 22.0, "end_s": 25.0, "source_start_s": 57.5, "source_end_s": 60.5 }, { "start_s": 31.0, "end_s": 32.0, "source_start_s": 66.5, "source_end_s": 67.5 }, { "start_s": 35.0, "end_s": 36.0, "source_start_s": 70.5, "source_end_s": 71.5 }, { "start_s": 39.0, "end_s": 40.5, "source_start_s": 74.5, "source_end_s": 76.0 } ], "evidence": "The masked person repeatedly moves his hands while speaking; at 15–17 seconds he clearly extends an arm and points toward the machine/monitor area. Later samples show additional explanatory hand gestures.", "mask_relation": "subject matches the supplied selected-person mask", "source_subject_refs": "entity_1" }, { "id": "action_4", "mentions": [ { "text": "clasping", "ranges": [ [ 120, 128 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "type": "simple", "subject_refs": [ "entity_1" ], "object_refs": [ "entity_5" ], "confidence": "high", "temporal_status": "supported", "intervals": [ { "start_s": 2.0, "end_s": 10.0, "source_start_s": 37.5, "source_end_s": 45.5 } ], "evidence": "In samples from 2 through 9 seconds, the masked person's hands are visibly held together over or between his knees.", "mask_relation": "subject matches the supplied selected-person mask", "source_subject_refs": "entity_1" } ], "notes": [ "The supplied mask consistently identifies the seated foreground person referenced by “his” and “he.”", "No separate masks were supplied for the feet, rowing machine, monitor, or hands.", "“secured” is omitted because the caption presents it as the resulting state of the feet rather than an asserted securing event.", "“continues” is an aspectual predicate and is excluded from the action spans.", "“gesture” and “pointing” are grouped as two mentions of one gesturing event because the participial phrase elaborates the gesture." ] }, "model_output_raw": { "caption": "Sitting upright with his feet now secured, he continues to talk and gesture, pointing towards the machine's monitor and clasping his hands.", "entity_groups": [ { "id": "entity_1", "label": "person", "spans": [ { "text": "his", "range": [ 21, 24 ] }, { "text": "he", "range": [ 43, 45 ] }, { "text": "his", "range": [ 129, 132 ] } ], "source_mask": "selected_person_masklet" }, { "id": "entity_2", "label": "feet", "spans": [ { "text": "his feet", "range": [ 21, 29 ] } ], "source_mask": null }, { "id": "entity_3", "label": "machine", "spans": [ { "text": "the machine's", "range": [ 94, 107 ] } ], "source_mask": null }, { "id": "entity_4", "label": "monitor", "spans": [ { "text": "the machine's monitor", "range": [ 94, 115 ] } ], "source_mask": null }, { "id": "entity_5", "label": "hands", "spans": [ { "text": "his hands", "range": [ 129, 138 ] } ], "source_mask": null } ], "actions": [ { "id": "action_1", "mentions": [ { "text": "Sitting", "ranges": [ [ 0, 7 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "type": "posture", "subject_ref": "entity_1", "object_refs": [], "confidence": "high", "temporal_status": "supported", "intervals": [ { "start_s": 0.0, "end_s": 40.5 } ], "mask_relation": "subject matches the supplied selected-person mask", "evidence": "The masked person remains seated on the rowing machine throughout all sampled frames." }, { "id": "action_2", "mentions": [ { "text": "talk", "ranges": [ [ 59, 63 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "type": "simple", "subject_ref": "entity_1", "object_refs": [], "confidence": "medium", "temporal_status": "supported", "intervals": [ { "start_s": 0.0, "end_s": 27.0 }, { "start_s": 34.0, "end_s": 36.0 }, { "start_s": 39.0, "end_s": 40.0 } ], "mask_relation": "subject matches the supplied selected-person mask", "evidence": "Across these sampled frames, the masked person's mouth and facial positions visibly vary in a manner consistent with ongoing speech." }, { "id": "action_3", "mentions": [ { "text": "gesture", "ranges": [ [ 68, 75 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } }, { "text": "pointing", "ranges": [ [ 77, 85 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "type": "simple", "subject_ref": "entity_1", "object_refs": [ "entity_4" ], "confidence": "high", "temporal_status": "supported", "intervals": [ { "start_s": 11.0, "end_s": 18.0 }, { "start_s": 22.0, "end_s": 25.0 }, { "start_s": 31.0, "end_s": 32.0 }, { "start_s": 35.0, "end_s": 36.0 }, { "start_s": 39.0, "end_s": 40.5 } ], "mask_relation": "subject matches the supplied selected-person mask", "evidence": "The masked person repeatedly moves his hands while speaking; at 15–17 seconds he clearly extends an arm and points toward the machine/monitor area. Later samples show additional explanatory hand gestures." }, { "id": "action_4", "mentions": [ { "text": "clasping", "ranges": [ [ 120, 128 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "type": "simple", "subject_ref": "entity_1", "object_refs": [ "entity_5" ], "confidence": "high", "temporal_status": "supported", "intervals": [ { "start_s": 2.0, "end_s": 10.0 } ], "mask_relation": "subject matches the supplied selected-person mask", "evidence": "In samples from 2 through 9 seconds, the masked person's hands are visibly held together over or between his knees." } ], "notes": [ "The supplied mask consistently identifies the seated foreground person referenced by “his” and “he.”", "No separate masks were supplied for the feet, rowing machine, monitor, or hands.", "“secured” is omitted because the caption presents it as the resulting state of the feet rather than an asserted securing event.", "“continues” is an aspectual predicate and is excluded from the action spans.", "“gesture” and “pointing” are grouped as two mentions of one gesturing event because the participial phrase elaborates the gesture." ] }, "normalization_repairs": 0 }, { "source_video_id": "v_-ceM4HHukhE", "caption_count_category": 9, "clip": { "caption_index": 5, "caption": "He leans his body to the left and repeatedly points at the flywheel damper while explaining its function, then straightens up.", "source_interval_sec": [ 76.0, 121.5 ], "loom_clip_id": "v_-ceM4HHukhE_2278_3641", "clip_frame_interval_4fps": [ 2278, 3641 ], "sample_count_1fps": 46, "sampled_frames": [ { "frame": "00000", "timestamp_sec": 76.0, "offset_sec": 0, "rgb_path": "assets/cat_09/v_-ceM4HHukhE_2278_3641/0000s_00000_rgb.jpg", "mask_path": "assets/cat_09/v_-ceM4HHukhE_2278_3641/0000s_00000_mask.png", "mask_status": "available" }, { "frame": "00028", "timestamp_sec": 77.0, "offset_sec": 1, "rgb_path": "assets/cat_09/v_-ceM4HHukhE_2278_3641/0001s_00028_rgb.jpg", "mask_path": "assets/cat_09/v_-ceM4HHukhE_2278_3641/0001s_00028_mask.png", "mask_status": "available" }, { "frame": "00060", "timestamp_sec": 78.0, "offset_sec": 2, "rgb_path": "assets/cat_09/v_-ceM4HHukhE_2278_3641/0002s_00060_rgb.jpg", "mask_path": "assets/cat_09/v_-ceM4HHukhE_2278_3641/0002s_00060_mask.png", "mask_status": "available" }, { "frame": "00088", "timestamp_sec": 79.0, "offset_sec": 3, "rgb_path": "assets/cat_09/v_-ceM4HHukhE_2278_3641/0003s_00088_rgb.jpg", "mask_path": "assets/cat_09/v_-ceM4HHukhE_2278_3641/0003s_00088_mask.png", "mask_status": "available" }, { "frame": "00120", "timestamp_sec": 80.0, "offset_sec": 4, "rgb_path": "assets/cat_09/v_-ceM4HHukhE_2278_3641/0004s_00120_rgb.jpg", "mask_path": "assets/cat_09/v_-ceM4HHukhE_2278_3641/0004s_00120_mask.png", "mask_status": "available" }, { "frame": "00148", "timestamp_sec": 81.0, "offset_sec": 5, "rgb_path": "assets/cat_09/v_-ceM4HHukhE_2278_3641/0005s_00148_rgb.jpg", "mask_path": "assets/cat_09/v_-ceM4HHukhE_2278_3641/0005s_00148_mask.png", "mask_status": "available" }, { "frame": "00180", "timestamp_sec": 82.0, "offset_sec": 6, "rgb_path": "assets/cat_09/v_-ceM4HHukhE_2278_3641/0006s_00180_rgb.jpg", "mask_path": "assets/cat_09/v_-ceM4HHukhE_2278_3641/0006s_00180_mask.png", "mask_status": "available" }, { "frame": "00208", "timestamp_sec": 83.0, "offset_sec": 7, "rgb_path": "assets/cat_09/v_-ceM4HHukhE_2278_3641/0007s_00208_rgb.jpg", "mask_path": "assets/cat_09/v_-ceM4HHukhE_2278_3641/0007s_00208_mask.png", "mask_status": "available" }, { "frame": "00240", "timestamp_sec": 84.0, "offset_sec": 8, "rgb_path": "assets/cat_09/v_-ceM4HHukhE_2278_3641/0008s_00240_rgb.jpg", "mask_path": "assets/cat_09/v_-ceM4HHukhE_2278_3641/0008s_00240_mask.png", "mask_status": "available" }, { "frame": "00268", "timestamp_sec": 85.0, "offset_sec": 9, "rgb_path": "assets/cat_09/v_-ceM4HHukhE_2278_3641/0009s_00268_rgb.jpg", "mask_path": "assets/cat_09/v_-ceM4HHukhE_2278_3641/0009s_00268_mask.png", "mask_status": "available" }, { "frame": "00300", "timestamp_sec": 86.0, "offset_sec": 10, "rgb_path": "assets/cat_09/v_-ceM4HHukhE_2278_3641/0010s_00300_rgb.jpg", "mask_path": "assets/cat_09/v_-ceM4HHukhE_2278_3641/0010s_00300_mask.png", "mask_status": "available" }, { "frame": "00328", "timestamp_sec": 87.0, "offset_sec": 11, "rgb_path": "assets/cat_09/v_-ceM4HHukhE_2278_3641/0011s_00328_rgb.jpg", "mask_path": "assets/cat_09/v_-ceM4HHukhE_2278_3641/0011s_00328_mask.png", "mask_status": "available" }, { "frame": "00360", "timestamp_sec": 88.0, "offset_sec": 12, "rgb_path": "assets/cat_09/v_-ceM4HHukhE_2278_3641/0012s_00360_rgb.jpg", "mask_path": "assets/cat_09/v_-ceM4HHukhE_2278_3641/0012s_00360_mask.png", "mask_status": "available" }, { "frame": "00388", "timestamp_sec": 89.0, "offset_sec": 13, "rgb_path": "assets/cat_09/v_-ceM4HHukhE_2278_3641/0013s_00388_rgb.jpg", "mask_path": "assets/cat_09/v_-ceM4HHukhE_2278_3641/0013s_00388_mask.png", "mask_status": "available" }, { "frame": "00420", "timestamp_sec": 90.0, "offset_sec": 14, "rgb_path": "assets/cat_09/v_-ceM4HHukhE_2278_3641/0014s_00420_rgb.jpg", "mask_path": "assets/cat_09/v_-ceM4HHukhE_2278_3641/0014s_00420_mask.png", "mask_status": "available" }, { "frame": "00448", "timestamp_sec": 91.0, "offset_sec": 15, "rgb_path": "assets/cat_09/v_-ceM4HHukhE_2278_3641/0015s_00448_rgb.jpg", "mask_path": "assets/cat_09/v_-ceM4HHukhE_2278_3641/0015s_00448_mask.png", "mask_status": "available" }, { "frame": "00480", "timestamp_sec": 92.0, "offset_sec": 16, "rgb_path": "assets/cat_09/v_-ceM4HHukhE_2278_3641/0016s_00480_rgb.jpg", "mask_path": "assets/cat_09/v_-ceM4HHukhE_2278_3641/0016s_00480_mask.png", "mask_status": "available" }, { "frame": "00508", "timestamp_sec": 93.0, "offset_sec": 17, "rgb_path": "assets/cat_09/v_-ceM4HHukhE_2278_3641/0017s_00508_rgb.jpg", "mask_path": "assets/cat_09/v_-ceM4HHukhE_2278_3641/0017s_00508_mask.png", "mask_status": "available" }, { "frame": "00540", "timestamp_sec": 94.0, "offset_sec": 18, "rgb_path": "assets/cat_09/v_-ceM4HHukhE_2278_3641/0018s_00540_rgb.jpg", "mask_path": "assets/cat_09/v_-ceM4HHukhE_2278_3641/0018s_00540_mask.png", "mask_status": "available" }, { "frame": "00568", "timestamp_sec": 95.0, "offset_sec": 19, "rgb_path": "assets/cat_09/v_-ceM4HHukhE_2278_3641/0019s_00568_rgb.jpg", "mask_path": "assets/cat_09/v_-ceM4HHukhE_2278_3641/0019s_00568_mask.png", "mask_status": "available" }, { "frame": "00600", "timestamp_sec": 96.0, "offset_sec": 20, "rgb_path": "assets/cat_09/v_-ceM4HHukhE_2278_3641/0020s_00600_rgb.jpg", "mask_path": "assets/cat_09/v_-ceM4HHukhE_2278_3641/0020s_00600_mask.png", "mask_status": "available" }, { "frame": "00628", "timestamp_sec": 97.0, "offset_sec": 21, "rgb_path": "assets/cat_09/v_-ceM4HHukhE_2278_3641/0021s_00628_rgb.jpg", "mask_path": "assets/cat_09/v_-ceM4HHukhE_2278_3641/0021s_00628_mask.png", "mask_status": "available" }, { "frame": "00660", "timestamp_sec": 98.0, "offset_sec": 22, "rgb_path": "assets/cat_09/v_-ceM4HHukhE_2278_3641/0022s_00660_rgb.jpg", "mask_path": "assets/cat_09/v_-ceM4HHukhE_2278_3641/0022s_00660_mask.png", "mask_status": "available" }, { "frame": "00688", "timestamp_sec": 99.0, "offset_sec": 23, "rgb_path": "assets/cat_09/v_-ceM4HHukhE_2278_3641/0023s_00688_rgb.jpg", "mask_path": "assets/cat_09/v_-ceM4HHukhE_2278_3641/0023s_00688_mask.png", "mask_status": "available" }, { "frame": "00720", "timestamp_sec": 100.0, "offset_sec": 24, "rgb_path": "assets/cat_09/v_-ceM4HHukhE_2278_3641/0024s_00720_rgb.jpg", "mask_path": "assets/cat_09/v_-ceM4HHukhE_2278_3641/0024s_00720_mask.png", "mask_status": "available" }, { "frame": "00748", "timestamp_sec": 101.0, "offset_sec": 25, "rgb_path": "assets/cat_09/v_-ceM4HHukhE_2278_3641/0025s_00748_rgb.jpg", "mask_path": "assets/cat_09/v_-ceM4HHukhE_2278_3641/0025s_00748_mask.png", "mask_status": "available" }, { "frame": "00780", "timestamp_sec": 102.0, "offset_sec": 26, "rgb_path": "assets/cat_09/v_-ceM4HHukhE_2278_3641/0026s_00780_rgb.jpg", "mask_path": "assets/cat_09/v_-ceM4HHukhE_2278_3641/0026s_00780_mask.png", "mask_status": "available" }, { "frame": "00808", "timestamp_sec": 103.0, "offset_sec": 27, "rgb_path": "assets/cat_09/v_-ceM4HHukhE_2278_3641/0027s_00808_rgb.jpg", "mask_path": "assets/cat_09/v_-ceM4HHukhE_2278_3641/0027s_00808_mask.png", "mask_status": "available" }, { "frame": "00840", "timestamp_sec": 104.0, "offset_sec": 28, "rgb_path": "assets/cat_09/v_-ceM4HHukhE_2278_3641/0028s_00840_rgb.jpg", "mask_path": "assets/cat_09/v_-ceM4HHukhE_2278_3641/0028s_00840_mask.png", "mask_status": "available" }, { "frame": "00868", "timestamp_sec": 105.0, "offset_sec": 29, "rgb_path": "assets/cat_09/v_-ceM4HHukhE_2278_3641/0029s_00868_rgb.jpg", "mask_path": "assets/cat_09/v_-ceM4HHukhE_2278_3641/0029s_00868_mask.png", "mask_status": "available" }, { "frame": "00900", "timestamp_sec": 106.0, "offset_sec": 30, "rgb_path": "assets/cat_09/v_-ceM4HHukhE_2278_3641/0030s_00900_rgb.jpg", "mask_path": "assets/cat_09/v_-ceM4HHukhE_2278_3641/0030s_00900_mask.png", "mask_status": "available" }, { "frame": "00928", "timestamp_sec": 107.0, "offset_sec": 31, "rgb_path": "assets/cat_09/v_-ceM4HHukhE_2278_3641/0031s_00928_rgb.jpg", "mask_path": "assets/cat_09/v_-ceM4HHukhE_2278_3641/0031s_00928_mask.png", "mask_status": "available" }, { "frame": "00960", "timestamp_sec": 108.0, "offset_sec": 32, "rgb_path": "assets/cat_09/v_-ceM4HHukhE_2278_3641/0032s_00960_rgb.jpg", "mask_path": "assets/cat_09/v_-ceM4HHukhE_2278_3641/0032s_00960_mask.png", "mask_status": "available" }, { "frame": "00988", "timestamp_sec": 109.0, "offset_sec": 33, "rgb_path": "assets/cat_09/v_-ceM4HHukhE_2278_3641/0033s_00988_rgb.jpg", "mask_path": "assets/cat_09/v_-ceM4HHukhE_2278_3641/0033s_00988_mask.png", "mask_status": "available" }, { "frame": "01020", "timestamp_sec": 110.0, "offset_sec": 34, "rgb_path": "assets/cat_09/v_-ceM4HHukhE_2278_3641/0034s_01020_rgb.jpg", "mask_path": "assets/cat_09/v_-ceM4HHukhE_2278_3641/0034s_01020_mask.png", "mask_status": "available" }, { "frame": "01048", "timestamp_sec": 111.0, "offset_sec": 35, "rgb_path": "assets/cat_09/v_-ceM4HHukhE_2278_3641/0035s_01048_rgb.jpg", "mask_path": "assets/cat_09/v_-ceM4HHukhE_2278_3641/0035s_01048_mask.png", "mask_status": "available" }, { "frame": "01080", "timestamp_sec": 112.0, "offset_sec": 36, "rgb_path": "assets/cat_09/v_-ceM4HHukhE_2278_3641/0036s_01080_rgb.jpg", "mask_path": "assets/cat_09/v_-ceM4HHukhE_2278_3641/0036s_01080_mask.png", "mask_status": "available" }, { "frame": "01108", "timestamp_sec": 113.0, "offset_sec": 37, "rgb_path": "assets/cat_09/v_-ceM4HHukhE_2278_3641/0037s_01108_rgb.jpg", "mask_path": "assets/cat_09/v_-ceM4HHukhE_2278_3641/0037s_01108_mask.png", "mask_status": "available" }, { "frame": "01140", "timestamp_sec": 114.0, "offset_sec": 38, "rgb_path": "assets/cat_09/v_-ceM4HHukhE_2278_3641/0038s_01140_rgb.jpg", "mask_path": "assets/cat_09/v_-ceM4HHukhE_2278_3641/0038s_01140_mask.png", "mask_status": "available" }, { "frame": "01168", "timestamp_sec": 115.0, "offset_sec": 39, "rgb_path": "assets/cat_09/v_-ceM4HHukhE_2278_3641/0039s_01168_rgb.jpg", "mask_path": "assets/cat_09/v_-ceM4HHukhE_2278_3641/0039s_01168_mask.png", "mask_status": "available" }, { "frame": "01200", "timestamp_sec": 116.0, "offset_sec": 40, "rgb_path": "assets/cat_09/v_-ceM4HHukhE_2278_3641/0040s_01200_rgb.jpg", "mask_path": "assets/cat_09/v_-ceM4HHukhE_2278_3641/0040s_01200_mask.png", "mask_status": "available" }, { "frame": "01228", "timestamp_sec": 117.0, "offset_sec": 41, "rgb_path": "assets/cat_09/v_-ceM4HHukhE_2278_3641/0041s_01228_rgb.jpg", "mask_path": "assets/cat_09/v_-ceM4HHukhE_2278_3641/0041s_01228_mask.png", "mask_status": "available" }, { "frame": "01260", "timestamp_sec": 118.0, "offset_sec": 42, "rgb_path": "assets/cat_09/v_-ceM4HHukhE_2278_3641/0042s_01260_rgb.jpg", "mask_path": "assets/cat_09/v_-ceM4HHukhE_2278_3641/0042s_01260_mask.png", "mask_status": "available" }, { "frame": "01288", "timestamp_sec": 119.0, "offset_sec": 43, "rgb_path": "assets/cat_09/v_-ceM4HHukhE_2278_3641/0043s_01288_rgb.jpg", "mask_path": "assets/cat_09/v_-ceM4HHukhE_2278_3641/0043s_01288_mask.png", "mask_status": "available" }, { "frame": "01320", "timestamp_sec": 120.0, "offset_sec": 44, "rgb_path": "assets/cat_09/v_-ceM4HHukhE_2278_3641/0044s_01320_rgb.jpg", "mask_path": "assets/cat_09/v_-ceM4HHukhE_2278_3641/0044s_01320_mask.png", "mask_status": "available" }, { "frame": "01348", "timestamp_sec": 121.0, "offset_sec": 45, "rgb_path": "assets/cat_09/v_-ceM4HHukhE_2278_3641/0045s_01348_rgb.jpg", "mask_path": "assets/cat_09/v_-ceM4HHukhE_2278_3641/0045s_01348_mask.png", "mask_status": "available" } ], "obj_id": [ 0 ], "anno_id": [ 96 ], "fps_inferred_from_clip_interval": 29.956 }, "usage": { "input_tokens": 62559, "input_tokens_details": { "cache_write_tokens": 62556, "cached_tokens": 0 }, "output_tokens": 4076, "output_tokens_details": { "reasoning_tokens": 2839 }, "total_tokens": 66635 }, "model": "gpt-5.6-sol", "effort": "medium", "custom_id": "cat09_cap05_v_-ceM4HHukhE_2278_3641", "annotation": { "caption": "He leans his body to the left and repeatedly points at the flywheel damper while explaining its function, then straightens up.", "entity_groups": [ { "id": "person_1", "label": "person", "spans": [ { "text": "He", "range": [ 0, 2 ] }, { "text": "his", "range": [ 9, 12 ] } ], "source_mask_link": "selected_person" }, { "id": "body_1", "label": "body", "spans": [ { "text": "his body", "range": [ 9, 17 ] } ], "source_mask_link": null }, { "id": "flywheel_damper_1", "label": "flywheel damper", "spans": [ { "text": "the flywheel damper", "range": [ 55, 74 ] }, { "text": "its", "range": [ 92, 95 ] } ], "source_mask_link": null } ], "actions": [ { "id": "action_1", "mentions": [ { "text": "leans", "ranges": [ [ 3, 8 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "type": "posture", "subject_refs": [ "person_1" ], "object_refs": [ "body_1" ], "confidence": "high", "temporal_status": "supported", "intervals": [ { "start_s": 0.0, "end_s": 6.0, "source_start_s": 76.0, "source_end_s": 82.0 } ], "evidence": "At 0–5 seconds, the masked seated person holds his torso noticeably inclined toward his left while reaching toward the machine.", "mask_relation": "subject_matches_selected_person_mask", "source_subject_refs": "person_1" }, { "id": "action_2", "mentions": [ { "text": "points", "ranges": [ [ 45, 51 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "type": "prepositional", "subject_refs": [ "person_1" ], "object_refs": [ "flywheel_damper_1" ], "confidence": "high", "temporal_status": "supported", "intervals": [ { "start_s": 1.0, "end_s": 5.0, "source_start_s": 77.0, "source_end_s": 81.0 }, { "start_s": 21.0, "end_s": 24.0, "source_start_s": 97.0, "source_end_s": 100.0 }, { "start_s": 32.0, "end_s": 35.0, "source_start_s": 108.0, "source_end_s": 111.0 } ], "evidence": "The masked person repeatedly extends a hand or index finger toward the flywheel housing and nearby damper area in several separated frame sequences.", "mask_relation": "subject_matches_selected_person_mask", "source_subject_refs": "person_1" }, { "id": "action_3", "mentions": [ { "text": "explaining", "ranges": [ [ 81, 91 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "type": "simple", "subject_refs": [ "person_1" ], "object_refs": [], "confidence": "medium", "temporal_status": "supported", "intervals": [ { "start_s": 0.0, "end_s": 42.0, "source_start_s": 76.0, "source_end_s": 118.0 } ], "evidence": "Across most sampled frames, the masked person appears to speak while making sustained explanatory hand gestures; the specific verbal content cannot be verified visually.", "mask_relation": "subject_matches_selected_person_mask", "source_subject_refs": "person_1" }, { "id": "action_4", "mentions": [ { "text": "straightens up", "ranges": [ [ 111, 125 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "type": "phrasal", "subject_refs": [ "person_1" ], "object_refs": [], "confidence": "medium", "temporal_status": "supported", "intervals": [ { "start_s": 5.0, "end_s": 7.0, "source_start_s": 81.0, "source_end_s": 83.0 } ], "evidence": "Between the 5- and 7-second samples, the masked person shifts from the pronounced lean to a more upright seated torso.", "mask_relation": "subject_matches_selected_person_mask", "source_subject_refs": "person_1" } ], "notes": [ "The supplied masklet consistently identifies the seated man in the foreground, corresponding to “He.”", "No separate masks are supplied for his body or the flywheel damper; these entity groups therefore have null source-mask links.", "“the left” is a spatial direction and “function” is abstract, so neither is annotated as an entity.", "The semantic content of “explaining its function” is not directly recoverable from silent frames; its timing is based on visible speaking-like facial activity and explanatory gestures." ] }, "model_output_raw": { "caption": "He leans his body to the left and repeatedly points at the flywheel damper while explaining its function, then straightens up.", "entity_groups": [ { "id": "person_1", "entity": "person", "spans": [ { "text": "He", "range": [ 0, 2 ] }, { "text": "his", "range": [ 9, 12 ] } ], "source_mask": "selected_person_masklet" }, { "id": "body_1", "entity": "body", "spans": [ { "text": "his body", "range": [ 9, 17 ] } ], "source_mask": null }, { "id": "flywheel_damper_1", "entity": "flywheel damper", "spans": [ { "text": "the flywheel damper", "range": [ 55, 74 ] }, { "text": "its", "range": [ 92, 95 ] } ], "source_mask": null } ], "actions": [ { "mentions": [ { "text": "leans", "ranges": [ [ 3, 8 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "type": "posture", "subject": "person_1", "object": "body_1", "confidence": "high", "temporal_status": "supported", "intervals": [ { "clip_relative_s": [ 0.0, 6.0 ], "source_video_s": [ 76.0, 82.0 ] } ], "evidence": "At 0–5 seconds, the masked seated person holds his torso noticeably inclined toward his left while reaching toward the machine.", "mask_relation": "subject_matches_selected_person_mask" }, { "mentions": [ { "text": "points", "ranges": [ [ 45, 51 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "type": "prepositional", "subject": "person_1", "object": "flywheel_damper_1", "confidence": "high", "temporal_status": "supported", "intervals": [ { "clip_relative_s": [ 1.0, 5.0 ], "source_video_s": [ 77.0, 81.0 ] }, { "clip_relative_s": [ 21.0, 24.0 ], "source_video_s": [ 97.0, 100.0 ] }, { "clip_relative_s": [ 32.0, 35.0 ], "source_video_s": [ 108.0, 111.0 ] } ], "evidence": "The masked person repeatedly extends a hand or index finger toward the flywheel housing and nearby damper area in several separated frame sequences.", "mask_relation": "subject_matches_selected_person_mask" }, { "mentions": [ { "text": "explaining", "ranges": [ [ 81, 91 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "type": "simple", "subject": "person_1", "object": null, "confidence": "medium", "temporal_status": "supported", "intervals": [ { "clip_relative_s": [ 0.0, 42.0 ], "source_video_s": [ 76.0, 118.0 ] } ], "evidence": "Across most sampled frames, the masked person appears to speak while making sustained explanatory hand gestures; the specific verbal content cannot be verified visually.", "mask_relation": "subject_matches_selected_person_mask" }, { "mentions": [ { "text": "straightens up", "ranges": [ [ 111, 125 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "type": "phrasal", "subject": "person_1", "object": null, "confidence": "medium", "temporal_status": "supported", "intervals": [ { "clip_relative_s": [ 5.0, 7.0 ], "source_video_s": [ 81.0, 83.0 ] } ], "evidence": "Between the 5- and 7-second samples, the masked person shifts from the pronounced lean to a more upright seated torso.", "mask_relation": "subject_matches_selected_person_mask" } ], "notes": [ "The supplied masklet consistently identifies the seated man in the foreground, corresponding to “He.”", "No separate masks are supplied for his body or the flywheel damper; these entity groups therefore have null source-mask links.", "“the left” is a spatial direction and “function” is abstract, so neither is annotated as an entity.", "The semantic content of “explaining its function” is not directly recoverable from silent frames; its timing is based on visible speaking-like facial activity and explanatory gestures." ] }, "normalization_repairs": 0 }, { "source_video_id": "v_-ceM4HHukhE", "caption_count_category": 9, "clip": { "caption_index": 6, "caption": "He grabs the handle with both hands, extends his arms, and leans his torso forward to demonstrate the starting \"catch\" position.", "source_interval_sec": [ 121.5, 151.5 ], "loom_clip_id": "v_-ceM4HHukhE_3641_4540", "clip_frame_interval_4fps": [ 3641, 4540 ], "sample_count_1fps": 30, "sampled_frames": [ { "frame": "00000", "timestamp_sec": 121.5, "offset_sec": 0, "rgb_path": "assets/cat_09/v_-ceM4HHukhE_3641_4540/0000s_00000_rgb.jpg", "mask_path": "assets/cat_09/v_-ceM4HHukhE_3641_4540/0000s_00000_mask.png", "mask_status": "available" }, { "frame": "00028", "timestamp_sec": 122.5, "offset_sec": 1, "rgb_path": "assets/cat_09/v_-ceM4HHukhE_3641_4540/0001s_00028_rgb.jpg", "mask_path": "assets/cat_09/v_-ceM4HHukhE_3641_4540/0001s_00028_mask.png", "mask_status": "available" }, { "frame": "00060", "timestamp_sec": 123.5, "offset_sec": 2, "rgb_path": "assets/cat_09/v_-ceM4HHukhE_3641_4540/0002s_00060_rgb.jpg", "mask_path": "assets/cat_09/v_-ceM4HHukhE_3641_4540/0002s_00060_mask.png", "mask_status": "available" }, { "frame": "00088", "timestamp_sec": 124.5, "offset_sec": 3, "rgb_path": "assets/cat_09/v_-ceM4HHukhE_3641_4540/0003s_00088_rgb.jpg", "mask_path": "assets/cat_09/v_-ceM4HHukhE_3641_4540/0003s_00088_mask.png", "mask_status": "available" }, { "frame": "00120", "timestamp_sec": 125.5, "offset_sec": 4, "rgb_path": "assets/cat_09/v_-ceM4HHukhE_3641_4540/0004s_00120_rgb.jpg", "mask_path": "assets/cat_09/v_-ceM4HHukhE_3641_4540/0004s_00120_mask.png", "mask_status": "available" }, { "frame": "00148", "timestamp_sec": 126.5, "offset_sec": 5, "rgb_path": "assets/cat_09/v_-ceM4HHukhE_3641_4540/0005s_00148_rgb.jpg", "mask_path": "assets/cat_09/v_-ceM4HHukhE_3641_4540/0005s_00148_mask.png", "mask_status": "available" }, { "frame": "00180", "timestamp_sec": 127.5, "offset_sec": 6, "rgb_path": "assets/cat_09/v_-ceM4HHukhE_3641_4540/0006s_00180_rgb.jpg", "mask_path": "assets/cat_09/v_-ceM4HHukhE_3641_4540/0006s_00180_mask.png", "mask_status": "available" }, { "frame": "00208", "timestamp_sec": 128.5, "offset_sec": 7, "rgb_path": "assets/cat_09/v_-ceM4HHukhE_3641_4540/0007s_00208_rgb.jpg", "mask_path": "assets/cat_09/v_-ceM4HHukhE_3641_4540/0007s_00208_mask.png", "mask_status": "available" }, { "frame": "00240", "timestamp_sec": 129.5, "offset_sec": 8, "rgb_path": "assets/cat_09/v_-ceM4HHukhE_3641_4540/0008s_00240_rgb.jpg", "mask_path": "assets/cat_09/v_-ceM4HHukhE_3641_4540/0008s_00240_mask.png", "mask_status": "available" }, { "frame": "00268", "timestamp_sec": 130.5, "offset_sec": 9, "rgb_path": "assets/cat_09/v_-ceM4HHukhE_3641_4540/0009s_00268_rgb.jpg", "mask_path": "assets/cat_09/v_-ceM4HHukhE_3641_4540/0009s_00268_mask.png", "mask_status": "available" }, { "frame": "00300", "timestamp_sec": 131.5, "offset_sec": 10, "rgb_path": "assets/cat_09/v_-ceM4HHukhE_3641_4540/0010s_00300_rgb.jpg", "mask_path": "assets/cat_09/v_-ceM4HHukhE_3641_4540/0010s_00300_mask.png", "mask_status": "available" }, { "frame": "00328", "timestamp_sec": 132.5, "offset_sec": 11, "rgb_path": "assets/cat_09/v_-ceM4HHukhE_3641_4540/0011s_00328_rgb.jpg", "mask_path": "assets/cat_09/v_-ceM4HHukhE_3641_4540/0011s_00328_mask.png", "mask_status": "available" }, { "frame": "00360", "timestamp_sec": 133.5, "offset_sec": 12, "rgb_path": "assets/cat_09/v_-ceM4HHukhE_3641_4540/0012s_00360_rgb.jpg", "mask_path": "assets/cat_09/v_-ceM4HHukhE_3641_4540/0012s_00360_mask.png", "mask_status": "available" }, { "frame": "00388", "timestamp_sec": 134.5, "offset_sec": 13, "rgb_path": "assets/cat_09/v_-ceM4HHukhE_3641_4540/0013s_00388_rgb.jpg", "mask_path": "assets/cat_09/v_-ceM4HHukhE_3641_4540/0013s_00388_mask.png", "mask_status": "available" }, { "frame": "00420", "timestamp_sec": 135.5, "offset_sec": 14, "rgb_path": "assets/cat_09/v_-ceM4HHukhE_3641_4540/0014s_00420_rgb.jpg", "mask_path": "assets/cat_09/v_-ceM4HHukhE_3641_4540/0014s_00420_mask.png", "mask_status": "available" }, { "frame": "00448", "timestamp_sec": 136.5, "offset_sec": 15, "rgb_path": "assets/cat_09/v_-ceM4HHukhE_3641_4540/0015s_00448_rgb.jpg", "mask_path": "assets/cat_09/v_-ceM4HHukhE_3641_4540/0015s_00448_mask.png", "mask_status": "available" }, { "frame": "00480", "timestamp_sec": 137.5, "offset_sec": 16, "rgb_path": "assets/cat_09/v_-ceM4HHukhE_3641_4540/0016s_00480_rgb.jpg", "mask_path": "assets/cat_09/v_-ceM4HHukhE_3641_4540/0016s_00480_mask.png", "mask_status": "available" }, { "frame": "00508", "timestamp_sec": 138.5, "offset_sec": 17, "rgb_path": "assets/cat_09/v_-ceM4HHukhE_3641_4540/0017s_00508_rgb.jpg", "mask_path": "assets/cat_09/v_-ceM4HHukhE_3641_4540/0017s_00508_mask.png", "mask_status": "available" }, { "frame": "00540", "timestamp_sec": 139.5, "offset_sec": 18, "rgb_path": "assets/cat_09/v_-ceM4HHukhE_3641_4540/0018s_00540_rgb.jpg", "mask_path": "assets/cat_09/v_-ceM4HHukhE_3641_4540/0018s_00540_mask.png", "mask_status": "available" }, { "frame": "00568", "timestamp_sec": 140.5, "offset_sec": 19, "rgb_path": "assets/cat_09/v_-ceM4HHukhE_3641_4540/0019s_00568_rgb.jpg", "mask_path": "assets/cat_09/v_-ceM4HHukhE_3641_4540/0019s_00568_mask.png", "mask_status": "available" }, { "frame": "00600", "timestamp_sec": 141.5, "offset_sec": 20, "rgb_path": "assets/cat_09/v_-ceM4HHukhE_3641_4540/0020s_00600_rgb.jpg", "mask_path": "assets/cat_09/v_-ceM4HHukhE_3641_4540/0020s_00600_mask.png", "mask_status": "available" }, { "frame": "00628", "timestamp_sec": 142.5, "offset_sec": 21, "rgb_path": "assets/cat_09/v_-ceM4HHukhE_3641_4540/0021s_00628_rgb.jpg", "mask_path": "assets/cat_09/v_-ceM4HHukhE_3641_4540/0021s_00628_mask.png", "mask_status": "available" }, { "frame": "00660", "timestamp_sec": 143.5, "offset_sec": 22, "rgb_path": "assets/cat_09/v_-ceM4HHukhE_3641_4540/0022s_00660_rgb.jpg", "mask_path": "assets/cat_09/v_-ceM4HHukhE_3641_4540/0022s_00660_mask.png", "mask_status": "available" }, { "frame": "00688", "timestamp_sec": 144.5, "offset_sec": 23, "rgb_path": "assets/cat_09/v_-ceM4HHukhE_3641_4540/0023s_00688_rgb.jpg", "mask_path": "assets/cat_09/v_-ceM4HHukhE_3641_4540/0023s_00688_mask.png", "mask_status": "available" }, { "frame": "00720", "timestamp_sec": 145.5, "offset_sec": 24, "rgb_path": "assets/cat_09/v_-ceM4HHukhE_3641_4540/0024s_00720_rgb.jpg", "mask_path": "assets/cat_09/v_-ceM4HHukhE_3641_4540/0024s_00720_mask.png", "mask_status": "available" }, { "frame": "00748", "timestamp_sec": 146.5, "offset_sec": 25, "rgb_path": "assets/cat_09/v_-ceM4HHukhE_3641_4540/0025s_00748_rgb.jpg", "mask_path": "assets/cat_09/v_-ceM4HHukhE_3641_4540/0025s_00748_mask.png", "mask_status": "available" }, { "frame": "00780", "timestamp_sec": 147.5, "offset_sec": 26, "rgb_path": "assets/cat_09/v_-ceM4HHukhE_3641_4540/0026s_00780_rgb.jpg", "mask_path": "assets/cat_09/v_-ceM4HHukhE_3641_4540/0026s_00780_mask.png", "mask_status": "available" }, { "frame": "00808", "timestamp_sec": 148.5, "offset_sec": 27, "rgb_path": "assets/cat_09/v_-ceM4HHukhE_3641_4540/0027s_00808_rgb.jpg", "mask_path": "assets/cat_09/v_-ceM4HHukhE_3641_4540/0027s_00808_mask.png", "mask_status": "available" }, { "frame": "00840", "timestamp_sec": 149.5, "offset_sec": 28, "rgb_path": "assets/cat_09/v_-ceM4HHukhE_3641_4540/0028s_00840_rgb.jpg", "mask_path": "assets/cat_09/v_-ceM4HHukhE_3641_4540/0028s_00840_mask.png", "mask_status": "available" }, { "frame": "00868", "timestamp_sec": 150.5, "offset_sec": 29, "rgb_path": "assets/cat_09/v_-ceM4HHukhE_3641_4540/0029s_00868_rgb.jpg", "mask_path": "assets/cat_09/v_-ceM4HHukhE_3641_4540/0029s_00868_mask.png", "mask_status": "available" } ], "obj_id": [ 0 ], "anno_id": [ 97 ], "fps_inferred_from_clip_interval": 29.9667 }, "usage": { "input_tokens": 46127, "input_tokens_details": { "cache_write_tokens": 46124, "cached_tokens": 0 }, "output_tokens": 3491, "output_tokens_details": { "reasoning_tokens": 2070 }, "total_tokens": 49618 }, "model": "gpt-5.6-sol", "effort": "medium", "custom_id": "cat09_cap06_v_-ceM4HHukhE_3641_4540", "annotation": { "caption": "He grabs the handle with both hands, extends his arms, and leans his torso forward to demonstrate the starting \"catch\" position.", "entity_groups": [ { "id": "person_1", "label": "person", "spans": [ { "text": "He", "range": [ 0, 2 ] }, { "text": "his", "range": [ 45, 48 ] }, { "text": "his", "range": [ 65, 68 ] } ], "source_mask_link": "selected_person" }, { "id": "handle_1", "label": "handle", "spans": [ { "text": "the handle", "range": [ 9, 19 ] } ], "source_mask_link": null }, { "id": "hands_1", "label": "hands", "spans": [ { "text": "both hands", "range": [ 25, 35 ] } ], "source_mask_link": null }, { "id": "arms_1", "label": "arms", "spans": [ { "text": "his arms", "range": [ 45, 53 ] } ], "source_mask_link": null }, { "id": "torso_1", "label": "torso", "spans": [ { "text": "his torso", "range": [ 65, 74 ] } ], "source_mask_link": null } ], "actions": [ { "id": "action_1", "mentions": [ { "text": "grabs", "ranges": [ [ 3, 8 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "type": "simple", "subject_refs": [ "person_1" ], "object_refs": [ "handle_1" ], "confidence": "high", "temporal_status": "localized", "intervals": [ { "start_s": 0.0, "end_s": 2.0, "source_start_s": 121.5, "source_end_s": 123.5 } ], "evidence": "At 0.0 seconds the selected person reaches onto the handle; by 1.0 seconds both hands visibly grasp it.", "mask_relation": "subject_matches_selected_person_mask", "source_subject_refs": [ "person_1" ] }, { "id": "action_2", "mentions": [ { "text": "extends", "ranges": [ [ 37, 44 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "type": "simple", "subject_refs": [ "person_1" ], "object_refs": [ "arms_1" ], "confidence": "high", "temporal_status": "localized", "intervals": [ { "start_s": 0.0, "end_s": 3.0, "source_start_s": 121.5, "source_end_s": 124.5 } ], "evidence": "Across the 0.0–2.0-second samples, the selected person's arms move from a lowered, bent reach to a forward extended configuration.", "mask_relation": "subject_matches_selected_person_mask", "source_subject_refs": [ "person_1" ] }, { "id": "action_3", "mentions": [ { "text": "leans", "ranges": [ [ 59, 64 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "type": "simple", "subject_refs": [ "person_1" ], "object_refs": [ "torso_1" ], "confidence": "high", "temporal_status": "localized", "intervals": [ { "start_s": 14.0, "end_s": 18.0, "source_start_s": 135.5, "source_end_s": 139.5 }, { "start_s": 20.0, "end_s": 23.0, "source_start_s": 141.5, "source_end_s": 144.5 } ], "evidence": "The selected person visibly shifts and holds his torso forward in the 14.0–17.0-second and 20.0–22.0-second samples.", "mask_relation": "subject_matches_selected_person_mask", "source_subject_refs": [ "person_1" ] }, { "id": "action_4", "mentions": [ { "text": "demonstrate", "ranges": [ [ 86, 97 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "type": "simple", "subject_refs": [ "person_1" ], "object_refs": [], "confidence": "high", "temporal_status": "localized", "intervals": [ { "start_s": 0.0, "end_s": 2.0, "source_start_s": 121.5, "source_end_s": 123.5 }, { "start_s": 8.0, "end_s": 10.0, "source_start_s": 129.5, "source_end_s": 131.5 }, { "start_s": 16.0, "end_s": 20.0, "source_start_s": 137.5, "source_end_s": 141.5 }, { "start_s": 22.0, "end_s": 30.0, "source_start_s": 143.5, "source_end_s": 151.5 } ], "evidence": "In these samples the selected person repeatedly displays the catch configuration: seated with knees bent, hands on the handle, arms forward, and torso inclined toward the machine. The supported final sequence continues through the clip boundary.", "mask_relation": "subject_matches_selected_person_mask", "source_subject_refs": [ "person_1" ] } ], "notes": "The supplied mask consistently identifies the seated foreground person referred to by “He” and both occurrences of “his.” No separate masks are supplied for the handle, hands, arms, or torso. The nominal phrase “the starting \"catch\" position” denotes a posture label rather than a concrete segmentable entity, so it has no entity group." }, "model_output_raw": { "caption": "He grabs the handle with both hands, extends his arms, and leans his torso forward to demonstrate the starting \"catch\" position.", "entity_groups": [ { "id": "person_1", "label": "person", "spans": [ { "text": "He", "start": 0, "end": 2 }, { "text": "his", "start": 45, "end": 48 }, { "text": "his", "start": 65, "end": 68 } ], "source_mask": "selected_person_mask" }, { "id": "handle_1", "label": "handle", "spans": [ { "text": "the handle", "start": 9, "end": 19 } ], "source_mask": null }, { "id": "hands_1", "label": "hands", "spans": [ { "text": "both hands", "start": 25, "end": 35 } ], "source_mask": null }, { "id": "arms_1", "label": "arms", "spans": [ { "text": "his arms", "start": 45, "end": 53 } ], "source_mask": null }, { "id": "torso_1", "label": "torso", "spans": [ { "text": "his torso", "start": 65, "end": 74 } ], "source_mask": null } ], "actions": [ { "mentions": [ { "text": "grabs", "ranges": [ { "start": 3, "end": 8 } ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "type": "simple", "subject_refs": [ "person_1" ], "object_refs": [ "handle_1" ], "confidence": "high", "temporal_status": "localized", "intervals": [ { "start_s": 0.0, "end_s": 2.0 } ], "evidence": "At 0.0 seconds the selected person reaches onto the handle; by 1.0 seconds both hands visibly grasp it.", "mask_relation": "subject_matches_selected_person_mask" }, { "mentions": [ { "text": "extends", "ranges": [ { "start": 37, "end": 44 } ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "type": "simple", "subject_refs": [ "person_1" ], "object_refs": [ "arms_1" ], "confidence": "high", "temporal_status": "localized", "intervals": [ { "start_s": 0.0, "end_s": 3.0 } ], "evidence": "Across the 0.0–2.0-second samples, the selected person's arms move from a lowered, bent reach to a forward extended configuration.", "mask_relation": "subject_matches_selected_person_mask" }, { "mentions": [ { "text": "leans", "ranges": [ { "start": 59, "end": 64 } ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "type": "simple", "subject_refs": [ "person_1" ], "object_refs": [ "torso_1" ], "confidence": "high", "temporal_status": "localized", "intervals": [ { "start_s": 14.0, "end_s": 18.0 }, { "start_s": 20.0, "end_s": 23.0 } ], "evidence": "The selected person visibly shifts and holds his torso forward in the 14.0–17.0-second and 20.0–22.0-second samples.", "mask_relation": "subject_matches_selected_person_mask" }, { "mentions": [ { "text": "demonstrate", "ranges": [ { "start": 86, "end": 97 } ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "type": "simple", "subject_refs": [ "person_1" ], "object_refs": [], "confidence": "high", "temporal_status": "localized", "intervals": [ { "start_s": 0.0, "end_s": 2.0 }, { "start_s": 8.0, "end_s": 10.0 }, { "start_s": 16.0, "end_s": 20.0 }, { "start_s": 22.0, "end_s": 30.0 } ], "evidence": "In these samples the selected person repeatedly displays the catch configuration: seated with knees bent, hands on the handle, arms forward, and torso inclined toward the machine. The supported final sequence continues through the clip boundary.", "mask_relation": "subject_matches_selected_person_mask" } ], "notes": "The supplied mask consistently identifies the seated foreground person referred to by “He” and both occurrences of “his.” No separate masks are supplied for the handle, hands, arms, or torso. The nominal phrase “the starting \"catch\" position” denotes a posture label rather than a concrete segmentable entity, so it has no entity group." }, "normalization_repairs": 0 }, { "source_video_id": "v_-ceM4HHukhE", "caption_count_category": 9, "clip": { "caption_index": 7, "caption": "While holding the handle, he demonstrates the initial phase of the rowing motion by pushing with his legs to slide the seat back, keeping his arms straight.", "source_interval_sec": [ 151.5, 175.0 ], "loom_clip_id": "v_-ceM4HHukhE_4540_5245", "clip_frame_interval_4fps": [ 4540, 5245 ], "sample_count_1fps": 24, "sampled_frames": [ { "frame": "00000", "timestamp_sec": 151.5, "offset_sec": 0, "rgb_path": "assets/cat_09/v_-ceM4HHukhE_4540_5245/0000s_00000_rgb.jpg", "mask_path": "assets/cat_09/v_-ceM4HHukhE_4540_5245/0000s_00000_mask.png", "mask_status": "available" }, { "frame": "00028", "timestamp_sec": 152.5, "offset_sec": 1, "rgb_path": "assets/cat_09/v_-ceM4HHukhE_4540_5245/0001s_00028_rgb.jpg", "mask_path": "assets/cat_09/v_-ceM4HHukhE_4540_5245/0001s_00028_mask.png", "mask_status": "available" }, { "frame": "00060", "timestamp_sec": 153.5, "offset_sec": 2, "rgb_path": "assets/cat_09/v_-ceM4HHukhE_4540_5245/0002s_00060_rgb.jpg", "mask_path": "assets/cat_09/v_-ceM4HHukhE_4540_5245/0002s_00060_mask.png", "mask_status": "available" }, { "frame": "00088", "timestamp_sec": 154.5, "offset_sec": 3, "rgb_path": "assets/cat_09/v_-ceM4HHukhE_4540_5245/0003s_00088_rgb.jpg", "mask_path": "assets/cat_09/v_-ceM4HHukhE_4540_5245/0003s_00088_mask.png", "mask_status": "available" }, { "frame": "00120", "timestamp_sec": 155.5, "offset_sec": 4, "rgb_path": "assets/cat_09/v_-ceM4HHukhE_4540_5245/0004s_00120_rgb.jpg", "mask_path": "assets/cat_09/v_-ceM4HHukhE_4540_5245/0004s_00120_mask.png", "mask_status": "available" }, { "frame": "00148", "timestamp_sec": 156.5, "offset_sec": 5, "rgb_path": "assets/cat_09/v_-ceM4HHukhE_4540_5245/0005s_00148_rgb.jpg", "mask_path": "assets/cat_09/v_-ceM4HHukhE_4540_5245/0005s_00148_mask.png", "mask_status": "available" }, { "frame": "00180", "timestamp_sec": 157.5, "offset_sec": 6, "rgb_path": "assets/cat_09/v_-ceM4HHukhE_4540_5245/0006s_00180_rgb.jpg", "mask_path": "assets/cat_09/v_-ceM4HHukhE_4540_5245/0006s_00180_mask.png", "mask_status": "available" }, { "frame": "00208", "timestamp_sec": 158.5, "offset_sec": 7, "rgb_path": "assets/cat_09/v_-ceM4HHukhE_4540_5245/0007s_00208_rgb.jpg", "mask_path": "assets/cat_09/v_-ceM4HHukhE_4540_5245/0007s_00208_mask.png", "mask_status": "available" }, { "frame": "00240", "timestamp_sec": 159.5, "offset_sec": 8, "rgb_path": "assets/cat_09/v_-ceM4HHukhE_4540_5245/0008s_00240_rgb.jpg", "mask_path": "assets/cat_09/v_-ceM4HHukhE_4540_5245/0008s_00240_mask.png", "mask_status": "available" }, { "frame": "00268", "timestamp_sec": 160.5, "offset_sec": 9, "rgb_path": "assets/cat_09/v_-ceM4HHukhE_4540_5245/0009s_00268_rgb.jpg", "mask_path": "assets/cat_09/v_-ceM4HHukhE_4540_5245/0009s_00268_mask.png", "mask_status": "available" }, { "frame": "00300", "timestamp_sec": 161.5, "offset_sec": 10, "rgb_path": "assets/cat_09/v_-ceM4HHukhE_4540_5245/0010s_00300_rgb.jpg", "mask_path": "assets/cat_09/v_-ceM4HHukhE_4540_5245/0010s_00300_mask.png", "mask_status": "available" }, { "frame": "00328", "timestamp_sec": 162.5, "offset_sec": 11, "rgb_path": "assets/cat_09/v_-ceM4HHukhE_4540_5245/0011s_00328_rgb.jpg", "mask_path": "assets/cat_09/v_-ceM4HHukhE_4540_5245/0011s_00328_mask.png", "mask_status": "available" }, { "frame": "00360", "timestamp_sec": 163.5, "offset_sec": 12, "rgb_path": "assets/cat_09/v_-ceM4HHukhE_4540_5245/0012s_00360_rgb.jpg", "mask_path": "assets/cat_09/v_-ceM4HHukhE_4540_5245/0012s_00360_mask.png", "mask_status": "available" }, { "frame": "00388", "timestamp_sec": 164.5, "offset_sec": 13, "rgb_path": "assets/cat_09/v_-ceM4HHukhE_4540_5245/0013s_00388_rgb.jpg", "mask_path": "assets/cat_09/v_-ceM4HHukhE_4540_5245/0013s_00388_mask.png", "mask_status": "available" }, { "frame": "00420", "timestamp_sec": 165.5, "offset_sec": 14, "rgb_path": "assets/cat_09/v_-ceM4HHukhE_4540_5245/0014s_00420_rgb.jpg", "mask_path": "assets/cat_09/v_-ceM4HHukhE_4540_5245/0014s_00420_mask.png", "mask_status": "available" }, { "frame": "00448", "timestamp_sec": 166.5, "offset_sec": 15, "rgb_path": "assets/cat_09/v_-ceM4HHukhE_4540_5245/0015s_00448_rgb.jpg", "mask_path": "assets/cat_09/v_-ceM4HHukhE_4540_5245/0015s_00448_mask.png", "mask_status": "available" }, { "frame": "00480", "timestamp_sec": 167.5, "offset_sec": 16, "rgb_path": "assets/cat_09/v_-ceM4HHukhE_4540_5245/0016s_00480_rgb.jpg", "mask_path": "assets/cat_09/v_-ceM4HHukhE_4540_5245/0016s_00480_mask.png", "mask_status": "available" }, { "frame": "00508", "timestamp_sec": 168.5, "offset_sec": 17, "rgb_path": "assets/cat_09/v_-ceM4HHukhE_4540_5245/0017s_00508_rgb.jpg", "mask_path": "assets/cat_09/v_-ceM4HHukhE_4540_5245/0017s_00508_mask.png", "mask_status": "available" }, { "frame": "00540", "timestamp_sec": 169.5, "offset_sec": 18, "rgb_path": "assets/cat_09/v_-ceM4HHukhE_4540_5245/0018s_00540_rgb.jpg", "mask_path": "assets/cat_09/v_-ceM4HHukhE_4540_5245/0018s_00540_mask.png", "mask_status": "available" }, { "frame": "00568", "timestamp_sec": 170.5, "offset_sec": 19, "rgb_path": "assets/cat_09/v_-ceM4HHukhE_4540_5245/0019s_00568_rgb.jpg", "mask_path": "assets/cat_09/v_-ceM4HHukhE_4540_5245/0019s_00568_mask.png", "mask_status": "available" }, { "frame": "00600", "timestamp_sec": 171.5, "offset_sec": 20, "rgb_path": "assets/cat_09/v_-ceM4HHukhE_4540_5245/0020s_00600_rgb.jpg", "mask_path": "assets/cat_09/v_-ceM4HHukhE_4540_5245/0020s_00600_mask.png", "mask_status": "available" }, { "frame": "00628", "timestamp_sec": 172.5, "offset_sec": 21, "rgb_path": "assets/cat_09/v_-ceM4HHukhE_4540_5245/0021s_00628_rgb.jpg", "mask_path": "assets/cat_09/v_-ceM4HHukhE_4540_5245/0021s_00628_mask.png", "mask_status": "available" }, { "frame": "00660", "timestamp_sec": 173.5, "offset_sec": 22, "rgb_path": "assets/cat_09/v_-ceM4HHukhE_4540_5245/0022s_00660_rgb.jpg", "mask_path": "assets/cat_09/v_-ceM4HHukhE_4540_5245/0022s_00660_mask.png", "mask_status": "available" }, { "frame": "00688", "timestamp_sec": 174.5, "offset_sec": 23, "rgb_path": "assets/cat_09/v_-ceM4HHukhE_4540_5245/0023s_00688_rgb.jpg", "mask_path": "assets/cat_09/v_-ceM4HHukhE_4540_5245/0023s_00688_mask.png", "mask_status": "available" } ], "obj_id": [ 0 ], "anno_id": [ 98 ], "fps_inferred_from_clip_interval": 30.0 }, "usage": { "input_tokens": 39968, "input_tokens_details": { "cache_write_tokens": 39965, "cached_tokens": 0 }, "output_tokens": 3663, "output_tokens_details": { "reasoning_tokens": 2170 }, "total_tokens": 43631 }, "model": "gpt-5.6-sol", "effort": "medium", "custom_id": "cat09_cap07_v_-ceM4HHukhE_4540_5245", "annotation": { "caption": "While holding the handle, he demonstrates the initial phase of the rowing motion by pushing with his legs to slide the seat back, keeping his arms straight.", "entity_groups": [ { "id": "E1", "label": "person", "spans": [ { "text": "he", "range": [ 26, 28 ] }, { "text": "his", "range": [ 97, 100 ] }, { "text": "his", "range": [ 138, 141 ] } ], "source_mask_link": "selected_person" }, { "id": "E2", "label": "handle", "spans": [ { "text": "the handle", "range": [ 14, 24 ] } ], "source_mask_link": null }, { "id": "E3", "label": "legs", "spans": [ { "text": "his legs", "range": [ 97, 105 ] } ], "source_mask_link": null }, { "id": "E4", "label": "seat", "spans": [ { "text": "the seat", "range": [ 115, 123 ] } ], "source_mask_link": null }, { "id": "E5", "label": "arms", "spans": [ { "text": "his arms", "range": [ 138, 146 ] } ], "source_mask_link": null } ], "actions": [ { "id": "action_1", "mentions": [ { "text": "holding", "ranges": [ [ 6, 13 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "type": "simple", "subject_refs": [ "E1" ], "object_refs": [ "E2" ], "confidence": "high", "temporal_status": "localized", "intervals": [ { "start_s": 0.0, "end_s": 23.5, "source_start_s": 151.5, "source_end_s": 175.0 } ], "evidence": "Across all sampled frames, the mask-linked man maintains at least one hand on the rowing handle.", "mask_relation": "The action subject matches the supplied selected-person mask.", "source_subject_refs": [ "E1" ] }, { "id": "action_2", "mentions": [ { "text": "demonstrates", "ranges": [ [ 29, 41 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "type": "simple", "subject_refs": [ "E1" ], "object_refs": [], "confidence": "high", "temporal_status": "localized", "intervals": [ { "start_s": 0.0, "end_s": 5.0, "source_start_s": 151.5, "source_end_s": 156.5 } ], "evidence": "Samples at 0–4 seconds show the mask-linked man visibly performing the initial drive phase on the rowing machine.", "mask_relation": "The action subject matches the supplied selected-person mask.", "source_subject_refs": [ "E1" ] }, { "id": "action_3", "mentions": [ { "text": "pushing", "ranges": [ [ 84, 91 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "type": "simple", "subject_refs": [ "E1" ], "object_refs": [], "confidence": "high", "temporal_status": "localized", "intervals": [ { "start_s": 0.0, "end_s": 5.0, "source_start_s": 151.5, "source_end_s": 156.5 } ], "evidence": "From samples 0–4 seconds, his knees progressively extend while his body and seat move rearward.", "mask_relation": "The action subject matches the supplied selected-person mask.", "source_subject_refs": [ "E1" ] }, { "id": "action_4", "mentions": [ { "text": "slide", "ranges": [ [ 109, 114 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "type": "simple", "subject_refs": [ "E1" ], "object_refs": [ "E4" ], "confidence": "high", "temporal_status": "localized", "intervals": [ { "start_s": 0.0, "end_s": 5.0, "source_start_s": 151.5, "source_end_s": 156.5 } ], "evidence": "Samples 0–4 seconds show the man and rowing-machine seat translating backward as his legs straighten.", "mask_relation": "The action subject matches the supplied selected-person mask.", "source_subject_refs": [ "E1" ] }, { "id": "action_5", "mentions": [ { "text": "keeping", "ranges": [ [ 130, 137 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "type": "posture", "subject_refs": [ "E1" ], "object_refs": [ "E5" ], "confidence": "high", "temporal_status": "localized", "intervals": [ { "start_s": 0.0, "end_s": 5.0, "source_start_s": 151.5, "source_end_s": 156.5 } ], "evidence": "During samples 0–4 seconds, the mask-linked man's arms remain extended while the leg-driven rearward slide occurs.", "mask_relation": "The action subject matches the supplied selected-person mask.", "source_subject_refs": [ "E1" ] } ], "notes": [ "The supplied mask consistently identifies the foreground man seated on the rowing machine, matching “he” and both possessive mentions “his.”", "No separate source masks were supplied for the handle, legs, seat, or arms; those entity groups therefore have null source-mask links.", "Other visible people are not referred to by the caption and are not added as text entity groups.", "The directional word “back” is excluded from the “slide” action span." ] }, "model_output_raw": { "caption": "While holding the handle, he demonstrates the initial phase of the rowing motion by pushing with his legs to slide the seat back, keeping his arms straight.", "entity_groups": [ { "id": "E1", "entity": "person", "spans": [ { "text": "he", "range": [ 26, 28 ] }, { "text": "his", "range": [ 97, 100 ] }, { "text": "his", "range": [ 138, 141 ] } ], "source_mask": "selected_person_mask" }, { "id": "E2", "entity": "handle", "spans": [ { "text": "the handle", "range": [ 14, 24 ] } ], "source_mask": null }, { "id": "E3", "entity": "legs", "spans": [ { "text": "his legs", "range": [ 97, 105 ] } ], "source_mask": null }, { "id": "E4", "entity": "seat", "spans": [ { "text": "the seat", "range": [ 115, 123 ] } ], "source_mask": null }, { "id": "E5", "entity": "arms", "spans": [ { "text": "his arms", "range": [ 138, 146 ] } ], "source_mask": null } ], "actions": [ { "mentions": [ { "text": "holding", "occurrence": 1, "ranges": [ [ 6, 13 ] ], "particle": { "text": null, "occurrence": 1 } } ], "type": "simple", "subject_refs": [ "E1" ], "object_refs": [ "E2" ], "confidence": "high", "temporal_status": "localized", "intervals": [ { "start_s": 0.0, "end_s": 23.5, "source_start_s": 151.5, "source_end_s": 175.0 } ], "evidence": "Across all sampled frames, the mask-linked man maintains at least one hand on the rowing handle.", "mask_relation": "The action subject matches the supplied selected-person mask." }, { "mentions": [ { "text": "demonstrates", "occurrence": 1, "ranges": [ [ 29, 41 ] ], "particle": { "text": null, "occurrence": 1 } } ], "type": "simple", "subject_refs": [ "E1" ], "object_refs": [], "confidence": "high", "temporal_status": "localized", "intervals": [ { "start_s": 0.0, "end_s": 5.0, "source_start_s": 151.5, "source_end_s": 156.5 } ], "evidence": "Samples at 0–4 seconds show the mask-linked man visibly performing the initial drive phase on the rowing machine.", "mask_relation": "The action subject matches the supplied selected-person mask." }, { "mentions": [ { "text": "pushing", "occurrence": 1, "ranges": [ [ 84, 91 ] ], "particle": { "text": null, "occurrence": 1 } } ], "type": "simple", "subject_refs": [ "E1" ], "object_refs": [], "instrument_refs": [ "E3" ], "confidence": "high", "temporal_status": "localized", "intervals": [ { "start_s": 0.0, "end_s": 5.0, "source_start_s": 151.5, "source_end_s": 156.5 } ], "evidence": "From samples 0–4 seconds, his knees progressively extend while his body and seat move rearward.", "mask_relation": "The action subject matches the supplied selected-person mask." }, { "mentions": [ { "text": "slide", "occurrence": 1, "ranges": [ [ 109, 114 ] ], "particle": { "text": null, "occurrence": 1 } } ], "type": "simple", "subject_refs": [ "E1" ], "object_refs": [ "E4" ], "confidence": "high", "temporal_status": "localized", "intervals": [ { "start_s": 0.0, "end_s": 5.0, "source_start_s": 151.5, "source_end_s": 156.5 } ], "evidence": "Samples 0–4 seconds show the man and rowing-machine seat translating backward as his legs straighten.", "mask_relation": "The action subject matches the supplied selected-person mask." }, { "mentions": [ { "text": "keeping", "occurrence": 1, "ranges": [ [ 130, 137 ] ], "particle": { "text": null, "occurrence": 1 } } ], "type": "posture", "subject_refs": [ "E1" ], "object_refs": [ "E5" ], "confidence": "high", "temporal_status": "localized", "intervals": [ { "start_s": 0.0, "end_s": 5.0, "source_start_s": 151.5, "source_end_s": 156.5 } ], "evidence": "During samples 0–4 seconds, the mask-linked man's arms remain extended while the leg-driven rearward slide occurs.", "mask_relation": "The action subject matches the supplied selected-person mask." } ], "notes": [ "The supplied mask consistently identifies the foreground man seated on the rowing machine, matching “he” and both possessive mentions “his.”", "No separate source masks were supplied for the handle, legs, seat, or arms; those entity groups therefore have null source-mask links.", "Other visible people are not referred to by the caption and are not added as text entity groups.", "The directional word “back” is excluded from the “slide” action span." ] }, "normalization_repairs": 0 }, { "source_video_id": "v_-ceM4HHukhE", "caption_count_category": 9, "clip": { "caption_index": 8, "caption": "He performs several complete and continuous rowing strokes, pushing with his legs, leaning back, and pulling the handle to his chest before returning forward.", "source_interval_sec": [ 175.0, 202.0 ], "loom_clip_id": "v_-ceM4HHukhE_5245_6054", "clip_frame_interval_4fps": [ 5245, 6054 ], "sample_count_1fps": 27, "sampled_frames": [ { "frame": "00000", "timestamp_sec": 175.0, "offset_sec": 0, "rgb_path": "assets/cat_09/v_-ceM4HHukhE_5245_6054/0000s_00000_rgb.jpg", "mask_path": "assets/cat_09/v_-ceM4HHukhE_5245_6054/0000s_00000_mask.png", "mask_status": "available" }, { "frame": "00028", "timestamp_sec": 176.0, "offset_sec": 1, "rgb_path": "assets/cat_09/v_-ceM4HHukhE_5245_6054/0001s_00028_rgb.jpg", "mask_path": "assets/cat_09/v_-ceM4HHukhE_5245_6054/0001s_00028_mask.png", "mask_status": "available" }, { "frame": "00060", "timestamp_sec": 177.0, "offset_sec": 2, "rgb_path": "assets/cat_09/v_-ceM4HHukhE_5245_6054/0002s_00060_rgb.jpg", "mask_path": "assets/cat_09/v_-ceM4HHukhE_5245_6054/0002s_00060_mask.png", "mask_status": "available" }, { "frame": "00088", "timestamp_sec": 178.0, "offset_sec": 3, "rgb_path": "assets/cat_09/v_-ceM4HHukhE_5245_6054/0003s_00088_rgb.jpg", "mask_path": "assets/cat_09/v_-ceM4HHukhE_5245_6054/0003s_00088_mask.png", "mask_status": "available" }, { "frame": "00120", "timestamp_sec": 179.0, "offset_sec": 4, "rgb_path": "assets/cat_09/v_-ceM4HHukhE_5245_6054/0004s_00120_rgb.jpg", "mask_path": "assets/cat_09/v_-ceM4HHukhE_5245_6054/0004s_00120_mask.png", "mask_status": "available" }, { "frame": "00148", "timestamp_sec": 180.0, "offset_sec": 5, "rgb_path": "assets/cat_09/v_-ceM4HHukhE_5245_6054/0005s_00148_rgb.jpg", "mask_path": "assets/cat_09/v_-ceM4HHukhE_5245_6054/0005s_00148_mask.png", "mask_status": "available" }, { "frame": "00180", "timestamp_sec": 181.0, "offset_sec": 6, "rgb_path": "assets/cat_09/v_-ceM4HHukhE_5245_6054/0006s_00180_rgb.jpg", "mask_path": "assets/cat_09/v_-ceM4HHukhE_5245_6054/0006s_00180_mask.png", "mask_status": "available" }, { "frame": "00208", "timestamp_sec": 182.0, "offset_sec": 7, "rgb_path": "assets/cat_09/v_-ceM4HHukhE_5245_6054/0007s_00208_rgb.jpg", "mask_path": "assets/cat_09/v_-ceM4HHukhE_5245_6054/0007s_00208_mask.png", "mask_status": "available" }, { "frame": "00240", "timestamp_sec": 183.0, "offset_sec": 8, "rgb_path": "assets/cat_09/v_-ceM4HHukhE_5245_6054/0008s_00240_rgb.jpg", "mask_path": "assets/cat_09/v_-ceM4HHukhE_5245_6054/0008s_00240_mask.png", "mask_status": "available" }, { "frame": "00268", "timestamp_sec": 184.0, "offset_sec": 9, "rgb_path": "assets/cat_09/v_-ceM4HHukhE_5245_6054/0009s_00268_rgb.jpg", "mask_path": "assets/cat_09/v_-ceM4HHukhE_5245_6054/0009s_00268_mask.png", "mask_status": "available" }, { "frame": "00300", "timestamp_sec": 185.0, "offset_sec": 10, "rgb_path": "assets/cat_09/v_-ceM4HHukhE_5245_6054/0010s_00300_rgb.jpg", "mask_path": "assets/cat_09/v_-ceM4HHukhE_5245_6054/0010s_00300_mask.png", "mask_status": "available" }, { "frame": "00328", "timestamp_sec": 186.0, "offset_sec": 11, "rgb_path": "assets/cat_09/v_-ceM4HHukhE_5245_6054/0011s_00328_rgb.jpg", "mask_path": "assets/cat_09/v_-ceM4HHukhE_5245_6054/0011s_00328_mask.png", "mask_status": "available" }, { "frame": "00360", "timestamp_sec": 187.0, "offset_sec": 12, "rgb_path": "assets/cat_09/v_-ceM4HHukhE_5245_6054/0012s_00360_rgb.jpg", "mask_path": "assets/cat_09/v_-ceM4HHukhE_5245_6054/0012s_00360_mask.png", "mask_status": "available" }, { "frame": "00388", "timestamp_sec": 188.0, "offset_sec": 13, "rgb_path": "assets/cat_09/v_-ceM4HHukhE_5245_6054/0013s_00388_rgb.jpg", "mask_path": "assets/cat_09/v_-ceM4HHukhE_5245_6054/0013s_00388_mask.png", "mask_status": "available" }, { "frame": "00420", "timestamp_sec": 189.0, "offset_sec": 14, "rgb_path": "assets/cat_09/v_-ceM4HHukhE_5245_6054/0014s_00420_rgb.jpg", "mask_path": "assets/cat_09/v_-ceM4HHukhE_5245_6054/0014s_00420_mask.png", "mask_status": "available" }, { "frame": "00448", "timestamp_sec": 190.0, "offset_sec": 15, "rgb_path": "assets/cat_09/v_-ceM4HHukhE_5245_6054/0015s_00448_rgb.jpg", "mask_path": "assets/cat_09/v_-ceM4HHukhE_5245_6054/0015s_00448_mask.png", "mask_status": "available" }, { "frame": "00480", "timestamp_sec": 191.0, "offset_sec": 16, "rgb_path": "assets/cat_09/v_-ceM4HHukhE_5245_6054/0016s_00480_rgb.jpg", "mask_path": "assets/cat_09/v_-ceM4HHukhE_5245_6054/0016s_00480_mask.png", "mask_status": "available" }, { "frame": "00508", "timestamp_sec": 192.0, "offset_sec": 17, "rgb_path": "assets/cat_09/v_-ceM4HHukhE_5245_6054/0017s_00508_rgb.jpg", "mask_path": "assets/cat_09/v_-ceM4HHukhE_5245_6054/0017s_00508_mask.png", "mask_status": "available" }, { "frame": "00540", "timestamp_sec": 193.0, "offset_sec": 18, "rgb_path": "assets/cat_09/v_-ceM4HHukhE_5245_6054/0018s_00540_rgb.jpg", "mask_path": "assets/cat_09/v_-ceM4HHukhE_5245_6054/0018s_00540_mask.png", "mask_status": "available" }, { "frame": "00568", "timestamp_sec": 194.0, "offset_sec": 19, "rgb_path": "assets/cat_09/v_-ceM4HHukhE_5245_6054/0019s_00568_rgb.jpg", "mask_path": "assets/cat_09/v_-ceM4HHukhE_5245_6054/0019s_00568_mask.png", "mask_status": "available" }, { "frame": "00600", "timestamp_sec": 195.0, "offset_sec": 20, "rgb_path": "assets/cat_09/v_-ceM4HHukhE_5245_6054/0020s_00600_rgb.jpg", "mask_path": "assets/cat_09/v_-ceM4HHukhE_5245_6054/0020s_00600_mask.png", "mask_status": "available" }, { "frame": "00628", "timestamp_sec": 196.0, "offset_sec": 21, "rgb_path": "assets/cat_09/v_-ceM4HHukhE_5245_6054/0021s_00628_rgb.jpg", "mask_path": "assets/cat_09/v_-ceM4HHukhE_5245_6054/0021s_00628_mask.png", "mask_status": "available" }, { "frame": "00660", "timestamp_sec": 197.0, "offset_sec": 22, "rgb_path": "assets/cat_09/v_-ceM4HHukhE_5245_6054/0022s_00660_rgb.jpg", "mask_path": "assets/cat_09/v_-ceM4HHukhE_5245_6054/0022s_00660_mask.png", "mask_status": "available" }, { "frame": "00688", "timestamp_sec": 198.0, "offset_sec": 23, "rgb_path": "assets/cat_09/v_-ceM4HHukhE_5245_6054/0023s_00688_rgb.jpg", "mask_path": "assets/cat_09/v_-ceM4HHukhE_5245_6054/0023s_00688_mask.png", "mask_status": "available" }, { "frame": "00720", "timestamp_sec": 199.0, "offset_sec": 24, "rgb_path": "assets/cat_09/v_-ceM4HHukhE_5245_6054/0024s_00720_rgb.jpg", "mask_path": "assets/cat_09/v_-ceM4HHukhE_5245_6054/0024s_00720_mask.png", "mask_status": "available" }, { "frame": "00748", "timestamp_sec": 200.0, "offset_sec": 25, "rgb_path": "assets/cat_09/v_-ceM4HHukhE_5245_6054/0025s_00748_rgb.jpg", "mask_path": "assets/cat_09/v_-ceM4HHukhE_5245_6054/0025s_00748_mask.png", "mask_status": "available" }, { "frame": "00780", "timestamp_sec": 201.0, "offset_sec": 26, "rgb_path": "assets/cat_09/v_-ceM4HHukhE_5245_6054/0026s_00780_rgb.jpg", "mask_path": "assets/cat_09/v_-ceM4HHukhE_5245_6054/0026s_00780_mask.png", "mask_status": "available" } ], "obj_id": [ 0 ], "anno_id": [ 99 ], "fps_inferred_from_clip_interval": 29.963 }, "usage": { "input_tokens": 43047, "input_tokens_details": { "cache_write_tokens": 43044, "cached_tokens": 0 }, "output_tokens": 4357, "output_tokens_details": { "reasoning_tokens": 2070 }, "total_tokens": 47404 }, "model": "gpt-5.6-sol", "effort": "medium", "custom_id": "cat09_cap08_v_-ceM4HHukhE_5245_6054", "annotation": { "caption": "He performs several complete and continuous rowing strokes, pushing with his legs, leaning back, and pulling the handle to his chest before returning forward.", "entity_groups": [ { "id": "entity_person_1", "label": "person", "spans": [ { "text": "He", "range": [ 0, 2 ] }, { "text": "his", "range": [ 73, 76 ] }, { "text": "his", "range": [ 123, 126 ] } ], "source_mask_link": "selected_person" }, { "id": "entity_legs_1", "label": "legs", "spans": [ { "text": "his legs", "range": [ 73, 81 ] } ], "source_mask_link": null }, { "id": "entity_handle_1", "label": "handle", "spans": [ { "text": "the handle", "range": [ 109, 119 ] } ], "source_mask_link": null }, { "id": "entity_chest_1", "label": "chest", "spans": [ { "text": "his chest", "range": [ 123, 132 ] } ], "source_mask_link": null } ], "actions": [ { "id": "action_rowing_1", "mentions": [ { "text": "performs several complete and continuous rowing strokes", "ranges": [ [ 3, 58 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "type": "light_verb", "subject_refs": [ "entity_person_1" ], "object_refs": [], "confidence": "high", "temporal_status": "localized_sparse", "intervals": [ { "start_s": 0.0, "end_s": 27.0, "source_start_s": 175.0, "source_end_s": 202.0 } ], "evidence": "Across all sampled frames, the masked person repeatedly cycles between a compressed catch position and a legs-extended finish while operating the rowing-machine handle.", "mask_relation": "subject matches the supplied selected-person mask", "source_subject_refs": "entity_person_1" }, { "id": "action_push_1", "mentions": [ { "text": "pushing", "ranges": [ [ 60, 67 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "type": "simple", "subject_refs": [ "entity_person_1" ], "object_refs": [], "confidence": "high", "temporal_status": "localized_sparse", "intervals": [ { "start_s": 3.0, "end_s": 7.0, "source_start_s": 178.0, "source_end_s": 182.0 }, { "start_s": 9.0, "end_s": 11.0, "source_start_s": 184.0, "source_end_s": 186.0 }, { "start_s": 11.0, "end_s": 13.0, "source_start_s": 186.0, "source_end_s": 188.0 }, { "start_s": 13.0, "end_s": 15.0, "source_start_s": 188.0, "source_end_s": 190.0 }, { "start_s": 15.0, "end_s": 17.0, "source_start_s": 190.0, "source_end_s": 192.0 }, { "start_s": 17.0, "end_s": 19.0, "source_start_s": 192.0, "source_end_s": 194.0 }, { "start_s": 19.0, "end_s": 21.0, "source_start_s": 194.0, "source_end_s": 196.0 }, { "start_s": 22.0, "end_s": 24.0, "source_start_s": 197.0, "source_end_s": 199.0 }, { "start_s": 24.0, "end_s": 27.0, "source_start_s": 199.0, "source_end_s": 202.0 } ], "evidence": "These sampled phases show the masked rower driving from bent knees toward leg extension.", "mask_relation": "subject matches the supplied selected-person mask", "source_subject_refs": "entity_person_1" }, { "id": "action_lean_1", "mentions": [ { "text": "leaning", "ranges": [ [ 83, 90 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "type": "simple", "subject_refs": [ "entity_person_1" ], "object_refs": [], "confidence": "high", "temporal_status": "localized_sparse", "intervals": [ { "start_s": 5.0, "end_s": 7.0, "source_start_s": 180.0, "source_end_s": 182.0 }, { "start_s": 10.0, "end_s": 11.0, "source_start_s": 185.0, "source_end_s": 186.0 }, { "start_s": 12.0, "end_s": 13.0, "source_start_s": 187.0, "source_end_s": 188.0 }, { "start_s": 14.0, "end_s": 15.0, "source_start_s": 189.0, "source_end_s": 190.0 }, { "start_s": 16.0, "end_s": 17.0, "source_start_s": 191.0, "source_end_s": 192.0 }, { "start_s": 18.0, "end_s": 19.0, "source_start_s": 193.0, "source_end_s": 194.0 }, { "start_s": 20.0, "end_s": 21.0, "source_start_s": 195.0, "source_end_s": 196.0 }, { "start_s": 23.0, "end_s": 24.0, "source_start_s": 198.0, "source_end_s": 199.0 }, { "start_s": 25.0, "end_s": 27.0, "source_start_s": 200.0, "source_end_s": 202.0 } ], "evidence": "At successive finish phases, the masked person's torso is visibly reclined relative to the forward catch posture.", "mask_relation": "subject matches the supplied selected-person mask", "source_subject_refs": "entity_person_1" }, { "id": "action_pull_1", "mentions": [ { "text": "pulling", "ranges": [ [ 101, 108 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "type": "simple", "subject_refs": [ "entity_person_1" ], "object_refs": [ "entity_handle_1" ], "confidence": "high", "temporal_status": "localized_sparse", "intervals": [ { "start_s": 4.0, "end_s": 7.0, "source_start_s": 179.0, "source_end_s": 182.0 }, { "start_s": 9.0, "end_s": 11.0, "source_start_s": 184.0, "source_end_s": 186.0 }, { "start_s": 11.0, "end_s": 13.0, "source_start_s": 186.0, "source_end_s": 188.0 }, { "start_s": 13.0, "end_s": 15.0, "source_start_s": 188.0, "source_end_s": 190.0 }, { "start_s": 15.0, "end_s": 17.0, "source_start_s": 190.0, "source_end_s": 192.0 }, { "start_s": 17.0, "end_s": 19.0, "source_start_s": 192.0, "source_end_s": 194.0 }, { "start_s": 19.0, "end_s": 21.0, "source_start_s": 194.0, "source_end_s": 196.0 }, { "start_s": 22.0, "end_s": 24.0, "source_start_s": 197.0, "source_end_s": 199.0 }, { "start_s": 24.0, "end_s": 27.0, "source_start_s": 199.0, "source_end_s": 202.0 } ], "evidence": "During drive and finish phases, the masked person bends his arms and brings the rowing handle toward his torso.", "mask_relation": "subject matches the supplied selected-person mask", "source_subject_refs": "entity_person_1" }, { "id": "action_return_1", "mentions": [ { "text": "returning", "ranges": [ [ 140, 149 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "type": "simple", "subject_refs": [ "entity_person_1" ], "object_refs": [], "confidence": "high", "temporal_status": "localized_sparse", "intervals": [ { "start_s": 7.0, "end_s": 10.0, "source_start_s": 182.0, "source_end_s": 185.0 }, { "start_s": 11.0, "end_s": 12.0, "source_start_s": 186.0, "source_end_s": 187.0 }, { "start_s": 13.0, "end_s": 14.0, "source_start_s": 188.0, "source_end_s": 189.0 }, { "start_s": 15.0, "end_s": 16.0, "source_start_s": 190.0, "source_end_s": 191.0 }, { "start_s": 17.0, "end_s": 18.0, "source_start_s": 192.0, "source_end_s": 193.0 }, { "start_s": 19.0, "end_s": 20.0, "source_start_s": 194.0, "source_end_s": 195.0 }, { "start_s": 21.0, "end_s": 23.0, "source_start_s": 196.0, "source_end_s": 198.0 }, { "start_s": 24.0, "end_s": 25.0, "source_start_s": 199.0, "source_end_s": 200.0 } ], "evidence": "Between finish and catch phases, the masked rower extends the arms, brings the torso forward, and bends the knees to slide toward the machine.", "mask_relation": "subject matches the supplied selected-person mask", "source_subject_refs": "entity_person_1" } ], "notes": [ "The supplied mask consistently identifies the foreground person using the rowing machine; all annotated action subjects refer to that person.", "The legs and chest are text-groundable body-part entities but have no independent supplied masks.", "The rowing handle is visible but has no supplied object mask.", "Action intervals are sparse clip-relative support derived from the one-second sampled clean frames, not exact continuous-video onset and offset boundaries." ] }, "model_output_raw": { "caption": "He performs several complete and continuous rowing strokes, pushing with his legs, leaning back, and pulling the handle to his chest before returning forward.", "entity_groups": [ { "id": "entity_person_1", "label": "person", "spans": [ { "text": "He", "start": 0, "end": 2 }, { "text": "his", "start": 73, "end": 76 }, { "text": "his", "start": 123, "end": 126 } ], "source_mask": "selected_person_mask" }, { "id": "entity_legs_1", "label": "legs", "spans": [ { "text": "his legs", "start": 73, "end": 81 } ], "source_mask": null }, { "id": "entity_handle_1", "label": "handle", "spans": [ { "text": "the handle", "start": 109, "end": 119 } ], "source_mask": null }, { "id": "entity_chest_1", "label": "chest", "spans": [ { "text": "his chest", "start": 123, "end": 132 } ], "source_mask": null } ], "actions": [ { "id": "action_rowing_1", "mentions": [ { "text": "performs several complete and continuous rowing strokes", "ranges": [ { "start": 3, "end": 58 } ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "type": "light_verb", "subject_entity_group_id": "entity_person_1", "object_entity_group_ids": [], "confidence": "high", "mask_relation": "subject matches the supplied selected-person mask", "temporal_status": "localized_sparse", "intervals": [ { "start_s": 0.0, "end_s": 27.0 } ], "evidence": "Across all sampled frames, the masked person repeatedly cycles between a compressed catch position and a legs-extended finish while operating the rowing-machine handle." }, { "id": "action_push_1", "mentions": [ { "text": "pushing", "ranges": [ { "start": 60, "end": 67 } ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "type": "simple", "subject_entity_group_id": "entity_person_1", "object_entity_group_ids": [], "related_entity_group_ids": [ "entity_legs_1" ], "confidence": "high", "mask_relation": "subject matches the supplied selected-person mask", "temporal_status": "localized_sparse", "intervals": [ { "start_s": 3.0, "end_s": 7.0 }, { "start_s": 9.0, "end_s": 11.0 }, { "start_s": 11.0, "end_s": 13.0 }, { "start_s": 13.0, "end_s": 15.0 }, { "start_s": 15.0, "end_s": 17.0 }, { "start_s": 17.0, "end_s": 19.0 }, { "start_s": 19.0, "end_s": 21.0 }, { "start_s": 22.0, "end_s": 24.0 }, { "start_s": 24.0, "end_s": 27.0 } ], "evidence": "These sampled phases show the masked rower driving from bent knees toward leg extension." }, { "id": "action_lean_1", "mentions": [ { "text": "leaning", "ranges": [ { "start": 83, "end": 90 } ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "type": "simple", "subject_entity_group_id": "entity_person_1", "object_entity_group_ids": [], "confidence": "high", "mask_relation": "subject matches the supplied selected-person mask", "temporal_status": "localized_sparse", "intervals": [ { "start_s": 5.0, "end_s": 7.0 }, { "start_s": 10.0, "end_s": 11.0 }, { "start_s": 12.0, "end_s": 13.0 }, { "start_s": 14.0, "end_s": 15.0 }, { "start_s": 16.0, "end_s": 17.0 }, { "start_s": 18.0, "end_s": 19.0 }, { "start_s": 20.0, "end_s": 21.0 }, { "start_s": 23.0, "end_s": 24.0 }, { "start_s": 25.0, "end_s": 27.0 } ], "evidence": "At successive finish phases, the masked person's torso is visibly reclined relative to the forward catch posture." }, { "id": "action_pull_1", "mentions": [ { "text": "pulling", "ranges": [ { "start": 101, "end": 108 } ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "type": "simple", "subject_entity_group_id": "entity_person_1", "object_entity_group_ids": [ "entity_handle_1" ], "related_entity_group_ids": [ "entity_chest_1" ], "confidence": "high", "mask_relation": "subject matches the supplied selected-person mask", "temporal_status": "localized_sparse", "intervals": [ { "start_s": 4.0, "end_s": 7.0 }, { "start_s": 9.0, "end_s": 11.0 }, { "start_s": 11.0, "end_s": 13.0 }, { "start_s": 13.0, "end_s": 15.0 }, { "start_s": 15.0, "end_s": 17.0 }, { "start_s": 17.0, "end_s": 19.0 }, { "start_s": 19.0, "end_s": 21.0 }, { "start_s": 22.0, "end_s": 24.0 }, { "start_s": 24.0, "end_s": 27.0 } ], "evidence": "During drive and finish phases, the masked person bends his arms and brings the rowing handle toward his torso." }, { "id": "action_return_1", "mentions": [ { "text": "returning", "ranges": [ { "start": 140, "end": 149 } ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "type": "simple", "subject_entity_group_id": "entity_person_1", "object_entity_group_ids": [], "confidence": "high", "mask_relation": "subject matches the supplied selected-person mask", "temporal_status": "localized_sparse", "intervals": [ { "start_s": 7.0, "end_s": 10.0 }, { "start_s": 11.0, "end_s": 12.0 }, { "start_s": 13.0, "end_s": 14.0 }, { "start_s": 15.0, "end_s": 16.0 }, { "start_s": 17.0, "end_s": 18.0 }, { "start_s": 19.0, "end_s": 20.0 }, { "start_s": 21.0, "end_s": 23.0 }, { "start_s": 24.0, "end_s": 25.0 } ], "evidence": "Between finish and catch phases, the masked rower extends the arms, brings the torso forward, and bends the knees to slide toward the machine." } ], "notes": [ "The supplied mask consistently identifies the foreground person using the rowing machine; all annotated action subjects refer to that person.", "The legs and chest are text-groundable body-part entities but have no independent supplied masks.", "The rowing handle is visible but has no supplied object mask.", "Action intervals are sparse clip-relative support derived from the one-second sampled clean frames, not exact continuous-video onset and offset boundaries." ] }, "normalization_repairs": 0 }, { "source_video_id": "v_-ceM4HHukhE", "caption_count_category": 9, "clip": { "caption_index": 9, "caption": "After completing a final stroke, he places the handle back in its holder and leans forward to unstrap his feet from the footrests.", "source_interval_sec": [ 202.0, 210.84 ], "loom_clip_id": "v_-ceM4HHukhE_6054_6319", "clip_frame_interval_4fps": [ 6054, 6319 ], "sample_count_1fps": 9, "sampled_frames": [ { "frame": "00000", "timestamp_sec": 202.0, "offset_sec": 0, "rgb_path": "assets/cat_09/v_-ceM4HHukhE_6054_6319/0000s_00000_rgb.jpg", "mask_path": "assets/cat_09/v_-ceM4HHukhE_6054_6319/0000s_00000_mask.png", "mask_status": "available" }, { "frame": "00028", "timestamp_sec": 203.0, "offset_sec": 1, "rgb_path": "assets/cat_09/v_-ceM4HHukhE_6054_6319/0001s_00028_rgb.jpg", "mask_path": "assets/cat_09/v_-ceM4HHukhE_6054_6319/0001s_00028_mask.png", "mask_status": "available" }, { "frame": "00060", "timestamp_sec": 204.0, "offset_sec": 2, "rgb_path": "assets/cat_09/v_-ceM4HHukhE_6054_6319/0002s_00060_rgb.jpg", "mask_path": "assets/cat_09/v_-ceM4HHukhE_6054_6319/0002s_00060_mask.png", "mask_status": "available" }, { "frame": "00088", "timestamp_sec": 205.0, "offset_sec": 3, "rgb_path": "assets/cat_09/v_-ceM4HHukhE_6054_6319/0003s_00088_rgb.jpg", "mask_path": "assets/cat_09/v_-ceM4HHukhE_6054_6319/0003s_00088_mask.png", "mask_status": "available" }, { "frame": "00120", "timestamp_sec": 206.0, "offset_sec": 4, "rgb_path": "assets/cat_09/v_-ceM4HHukhE_6054_6319/0004s_00120_rgb.jpg", "mask_path": "assets/cat_09/v_-ceM4HHukhE_6054_6319/0004s_00120_mask.png", "mask_status": "available" }, { "frame": "00148", "timestamp_sec": 207.0, "offset_sec": 5, "rgb_path": "assets/cat_09/v_-ceM4HHukhE_6054_6319/0005s_00148_rgb.jpg", "mask_path": "assets/cat_09/v_-ceM4HHukhE_6054_6319/0005s_00148_mask.png", "mask_status": "available" }, { "frame": "00180", "timestamp_sec": 208.0, "offset_sec": 6, "rgb_path": "assets/cat_09/v_-ceM4HHukhE_6054_6319/0006s_00180_rgb.jpg", "mask_path": "assets/cat_09/v_-ceM4HHukhE_6054_6319/0006s_00180_mask.png", "mask_status": "available" }, { "frame": "00208", "timestamp_sec": 209.0, "offset_sec": 7, "rgb_path": "assets/cat_09/v_-ceM4HHukhE_6054_6319/0007s_00208_rgb.jpg", "mask_path": "assets/cat_09/v_-ceM4HHukhE_6054_6319/0007s_00208_mask.png", "mask_status": "available" }, { "frame": "00240", "timestamp_sec": 210.0, "offset_sec": 8, "rgb_path": "assets/cat_09/v_-ceM4HHukhE_6054_6319/0008s_00240_rgb.jpg", "mask_path": "assets/cat_09/v_-ceM4HHukhE_6054_6319/0008s_00240_mask.png", "mask_status": "available" } ], "obj_id": [ 0 ], "anno_id": [ 101 ], "fps_inferred_from_clip_interval": 29.9774 }, "usage": { "input_tokens": 24561, "input_tokens_details": { "cache_write_tokens": 24558, "cached_tokens": 0 }, "output_tokens": 3994, "output_tokens_details": { "reasoning_tokens": 2689 }, "total_tokens": 28555 }, "model": "gpt-5.6-sol", "effort": "medium", "custom_id": "cat09_cap09_v_-ceM4HHukhE_6054_6319", "annotation": { "caption": "After completing a final stroke, he places the handle back in its holder and leans forward to unstrap his feet from the footrests.", "entity_groups": [ { "id": "person_1", "label": "person", "spans": [ { "text": "he", "range": [ 33, 35 ] }, { "text": "his", "range": [ 102, 105 ] } ], "source_mask_link": "selected_person" }, { "id": "handle_1", "label": "handle", "spans": [ { "text": "the handle", "range": [ 43, 53 ] }, { "text": "its", "range": [ 62, 65 ] } ], "source_mask_link": null }, { "id": "holder_1", "label": "holder", "spans": [ { "text": "its holder", "range": [ 62, 72 ] } ], "source_mask_link": null }, { "id": "feet_1", "label": "feet", "spans": [ { "text": "his feet", "range": [ 102, 110 ] } ], "source_mask_link": null }, { "id": "footrests_1", "label": "footrests", "spans": [ { "text": "the footrests", "range": [ 116, 129 ] } ], "source_mask_link": null } ], "actions": [ { "id": "action_1", "mentions": [ { "text": "completing a final stroke", "ranges": [ [ 6, 31 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "type": "light_verb", "subject_refs": [ "person_1" ], "object_refs": [], "confidence": "high", "temporal_status": "sparse_localized", "intervals": [ { "start_s": 0.0, "end_s": 2.0, "source_start_s": 202.0, "source_end_s": 204.0 } ], "evidence": "At local 0.000 s the selected man is at the forward position holding the rowing handle; at 1.000 s he is leaned back with the handle drawn toward him, visually supporting completion of a rowing stroke.", "mask_relation": "subject_matches_selected_person_mask", "source_subject_refs": "person_1" }, { "id": "action_2", "mentions": [ { "text": "places", "ranges": [ [ 36, 42 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "type": "simple", "subject_refs": [ "person_1" ], "object_refs": [ "handle_1" ], "confidence": "medium", "temporal_status": "sparse_localized", "intervals": [ { "start_s": 6.0, "end_s": 8.0, "source_start_s": 208.0, "source_end_s": 210.0 } ], "evidence": "At local 6.000 s the selected man still holds the handle; by 7.000 s he lowers it toward the rowing machine, and at 8.000 s his hands have moved away toward his feet.", "mask_relation": "subject_matches_selected_person_mask", "source_subject_refs": "person_1" }, { "id": "action_3", "mentions": [ { "text": "leans", "ranges": [ [ 77, 82 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "type": "simple", "subject_refs": [ "person_1" ], "object_refs": [], "confidence": "high", "temporal_status": "sparse_localized", "intervals": [ { "start_s": 7.0, "end_s": 8.840000000000003, "source_start_s": 209.0, "source_end_s": 210.84 } ], "evidence": "The selected man bends his torso progressively toward his feet between local 7.000 and 8.000 s.", "mask_relation": "subject_matches_selected_person_mask", "source_subject_refs": "person_1" }, { "id": "action_4", "mentions": [ { "text": "unstrap", "ranges": [ [ 94, 101 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "type": "simple", "subject_refs": [ "person_1" ], "object_refs": [ "feet_1" ], "confidence": "medium", "temporal_status": "sparse_localized", "intervals": [ { "start_s": 8.0, "end_s": 8.840000000000003, "source_start_s": 210.0, "source_end_s": 210.84 } ], "evidence": "At local 8.000 s the selected man is bent over with both hands at the shoe and footrest area, supporting manipulation of the foot straps through the clip boundary.", "mask_relation": "subject_matches_selected_person_mask", "source_subject_refs": "person_1" } ], "notes": [ "The supplied masklet consistently identifies the seated man referred to by “he” and “his.”", "No separate source masks were supplied for the handle, holder, feet, or footrests.", "The stroke itself is an action named by an event noun and is therefore not included as a concrete entity group.", "The exact completion of handle placement and release of the foot straps are only sparsely visible, so those actions have medium confidence." ] }, "model_output_raw": { "caption": "After completing a final stroke, he places the handle back in its holder and leans forward to unstrap his feet from the footrests.", "entity_groups": [ { "id": "person_1", "entity_type": "person", "spans": [ { "text": "he", "range": [ 33, 35 ] }, { "text": "his", "range": [ 102, 105 ] } ], "source_mask": "selected_person_masklet" }, { "id": "handle_1", "entity_type": "handle", "spans": [ { "text": "the handle", "range": [ 43, 53 ] }, { "text": "its", "range": [ 62, 65 ] } ], "source_mask": null }, { "id": "holder_1", "entity_type": "holder", "spans": [ { "text": "its holder", "range": [ 62, 72 ] } ], "source_mask": null }, { "id": "feet_1", "entity_type": "feet", "spans": [ { "text": "his feet", "range": [ 102, 110 ] } ], "source_mask": null }, { "id": "footrests_1", "entity_type": "footrests", "spans": [ { "text": "the footrests", "range": [ 116, 129 ] } ], "source_mask": null } ], "actions": [ { "mentions": [ { "text": "completing a final stroke", "ranges": [ [ 6, 31 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "type": "light_verb", "subject_entity_group": "person_1", "object_entity_groups": [], "confidence": "high", "temporal_status": "sparse_localized", "intervals": [ { "start_s": 0.0, "end_s": 2.0 } ], "evidence": "At local 0.000 s the selected man is at the forward position holding the rowing handle; at 1.000 s he is leaned back with the handle drawn toward him, visually supporting completion of a rowing stroke.", "mask_relation": "subject_matches_selected_person_mask" }, { "mentions": [ { "text": "places", "ranges": [ [ 36, 42 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "type": "simple", "subject_entity_group": "person_1", "object_entity_groups": [ "handle_1" ], "confidence": "medium", "temporal_status": "sparse_localized", "intervals": [ { "start_s": 6.0, "end_s": 8.0 } ], "evidence": "At local 6.000 s the selected man still holds the handle; by 7.000 s he lowers it toward the rowing machine, and at 8.000 s his hands have moved away toward his feet.", "mask_relation": "subject_matches_selected_person_mask" }, { "mentions": [ { "text": "leans", "ranges": [ [ 77, 82 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "type": "simple", "subject_entity_group": "person_1", "object_entity_groups": [], "confidence": "high", "temporal_status": "sparse_localized", "intervals": [ { "start_s": 7.0, "end_s": 8.84 } ], "evidence": "The selected man bends his torso progressively toward his feet between local 7.000 and 8.000 s.", "mask_relation": "subject_matches_selected_person_mask" }, { "mentions": [ { "text": "unstrap", "ranges": [ [ 94, 101 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "type": "simple", "subject_entity_group": "person_1", "object_entity_groups": [ "feet_1" ], "confidence": "medium", "temporal_status": "sparse_localized", "intervals": [ { "start_s": 8.0, "end_s": 8.84 } ], "evidence": "At local 8.000 s the selected man is bent over with both hands at the shoe and footrest area, supporting manipulation of the foot straps through the clip boundary.", "mask_relation": "subject_matches_selected_person_mask" } ], "notes": [ "The supplied masklet consistently identifies the seated man referred to by “he” and “his.”", "No separate source masks were supplied for the handle, holder, feet, or footrests.", "The stroke itself is an action named by an event noun and is therefore not included as a concrete entity group.", "The exact completion of handle placement and release of the foot straps are only sparsely visible, so those actions have medium confidence." ] }, "normalization_repairs": 0 }, { "source_video_id": "v_-ypKYhjrlXw", "caption_count_category": 10, "clip": { "caption_index": 1, "caption": "A man in a white lab coat and red tie sits at a table and speaks, occasionally gesturing with his hands. In front of him are several white bottles and containers, and a woman's arm is resting on the table to his left.", "source_interval_sec": [ 0.0, 35.5 ], "loom_clip_id": "v_-ypKYhjrlXw_1_1064", "clip_frame_interval_4fps": [ 1, 1064 ], "sample_count_1fps": 36, "sampled_frames": [ { "frame": "00000", "timestamp_sec": 0.0, "offset_sec": 0, "rgb_path": "assets/cat_10/v_-ypKYhjrlXw_1_1064/0000s_00000_rgb.jpg", "mask_path": "assets/cat_10/v_-ypKYhjrlXw_1_1064/0000s_00000_mask.png", "mask_status": "available" }, { "frame": "00028", "timestamp_sec": 1.0, "offset_sec": 1, "rgb_path": "assets/cat_10/v_-ypKYhjrlXw_1_1064/0001s_00028_rgb.jpg", "mask_path": "assets/cat_10/v_-ypKYhjrlXw_1_1064/0001s_00028_mask.png", "mask_status": "available" }, { "frame": "00060", "timestamp_sec": 2.0, "offset_sec": 2, "rgb_path": "assets/cat_10/v_-ypKYhjrlXw_1_1064/0002s_00060_rgb.jpg", "mask_path": "assets/cat_10/v_-ypKYhjrlXw_1_1064/0002s_00060_mask.png", "mask_status": "available" }, { "frame": "00088", "timestamp_sec": 3.0, "offset_sec": 3, "rgb_path": "assets/cat_10/v_-ypKYhjrlXw_1_1064/0003s_00088_rgb.jpg", "mask_path": "assets/cat_10/v_-ypKYhjrlXw_1_1064/0003s_00088_mask.png", "mask_status": "available" }, { "frame": "00120", "timestamp_sec": 4.0, "offset_sec": 4, "rgb_path": "assets/cat_10/v_-ypKYhjrlXw_1_1064/0004s_00120_rgb.jpg", "mask_path": "assets/cat_10/v_-ypKYhjrlXw_1_1064/0004s_00120_mask.png", "mask_status": "available" }, { "frame": "00148", "timestamp_sec": 5.0, "offset_sec": 5, "rgb_path": "assets/cat_10/v_-ypKYhjrlXw_1_1064/0005s_00148_rgb.jpg", "mask_path": "assets/cat_10/v_-ypKYhjrlXw_1_1064/0005s_00148_mask.png", "mask_status": "available" }, { "frame": "00180", "timestamp_sec": 6.0, "offset_sec": 6, "rgb_path": "assets/cat_10/v_-ypKYhjrlXw_1_1064/0006s_00180_rgb.jpg", "mask_path": "assets/cat_10/v_-ypKYhjrlXw_1_1064/0006s_00180_mask.png", "mask_status": "available" }, { "frame": "00208", "timestamp_sec": 7.0, "offset_sec": 7, "rgb_path": "assets/cat_10/v_-ypKYhjrlXw_1_1064/0007s_00208_rgb.jpg", "mask_path": "assets/cat_10/v_-ypKYhjrlXw_1_1064/0007s_00208_mask.png", "mask_status": "available" }, { "frame": "00240", "timestamp_sec": 8.0, "offset_sec": 8, "rgb_path": "assets/cat_10/v_-ypKYhjrlXw_1_1064/0008s_00240_rgb.jpg", "mask_path": "assets/cat_10/v_-ypKYhjrlXw_1_1064/0008s_00240_mask.png", "mask_status": "available" }, { "frame": "00268", "timestamp_sec": 9.0, "offset_sec": 9, "rgb_path": "assets/cat_10/v_-ypKYhjrlXw_1_1064/0009s_00268_rgb.jpg", "mask_path": "assets/cat_10/v_-ypKYhjrlXw_1_1064/0009s_00268_mask.png", "mask_status": "available" }, { "frame": "00300", "timestamp_sec": 10.0, "offset_sec": 10, "rgb_path": "assets/cat_10/v_-ypKYhjrlXw_1_1064/0010s_00300_rgb.jpg", "mask_path": "assets/cat_10/v_-ypKYhjrlXw_1_1064/0010s_00300_mask.png", "mask_status": "available" }, { "frame": "00328", "timestamp_sec": 11.0, "offset_sec": 11, "rgb_path": "assets/cat_10/v_-ypKYhjrlXw_1_1064/0011s_00328_rgb.jpg", "mask_path": "assets/cat_10/v_-ypKYhjrlXw_1_1064/0011s_00328_mask.png", "mask_status": "available" }, { "frame": "00360", "timestamp_sec": 12.0, "offset_sec": 12, "rgb_path": "assets/cat_10/v_-ypKYhjrlXw_1_1064/0012s_00360_rgb.jpg", "mask_path": "assets/cat_10/v_-ypKYhjrlXw_1_1064/0012s_00360_mask.png", "mask_status": "available" }, { "frame": "00388", "timestamp_sec": 13.0, "offset_sec": 13, "rgb_path": "assets/cat_10/v_-ypKYhjrlXw_1_1064/0013s_00388_rgb.jpg", "mask_path": "assets/cat_10/v_-ypKYhjrlXw_1_1064/0013s_00388_mask.png", "mask_status": "available" }, { "frame": "00420", "timestamp_sec": 14.0, "offset_sec": 14, "rgb_path": "assets/cat_10/v_-ypKYhjrlXw_1_1064/0014s_00420_rgb.jpg", "mask_path": "assets/cat_10/v_-ypKYhjrlXw_1_1064/0014s_00420_mask.png", "mask_status": "available" }, { "frame": "00448", "timestamp_sec": 15.0, "offset_sec": 15, "rgb_path": "assets/cat_10/v_-ypKYhjrlXw_1_1064/0015s_00448_rgb.jpg", "mask_path": "assets/cat_10/v_-ypKYhjrlXw_1_1064/0015s_00448_mask.png", "mask_status": "available" }, { "frame": "00480", "timestamp_sec": 16.0, "offset_sec": 16, "rgb_path": "assets/cat_10/v_-ypKYhjrlXw_1_1064/0016s_00480_rgb.jpg", "mask_path": "assets/cat_10/v_-ypKYhjrlXw_1_1064/0016s_00480_mask.png", "mask_status": "available" }, { "frame": "00508", "timestamp_sec": 17.0, "offset_sec": 17, "rgb_path": "assets/cat_10/v_-ypKYhjrlXw_1_1064/0017s_00508_rgb.jpg", "mask_path": "assets/cat_10/v_-ypKYhjrlXw_1_1064/0017s_00508_mask.png", "mask_status": "available" }, { "frame": "00540", "timestamp_sec": 18.0, "offset_sec": 18, "rgb_path": "assets/cat_10/v_-ypKYhjrlXw_1_1064/0018s_00540_rgb.jpg", "mask_path": "assets/cat_10/v_-ypKYhjrlXw_1_1064/0018s_00540_mask.png", "mask_status": "available" }, { "frame": "00568", "timestamp_sec": 19.0, "offset_sec": 19, "rgb_path": "assets/cat_10/v_-ypKYhjrlXw_1_1064/0019s_00568_rgb.jpg", "mask_path": "assets/cat_10/v_-ypKYhjrlXw_1_1064/0019s_00568_mask.png", "mask_status": "available" }, { "frame": "00600", "timestamp_sec": 20.0, "offset_sec": 20, "rgb_path": "assets/cat_10/v_-ypKYhjrlXw_1_1064/0020s_00600_rgb.jpg", "mask_path": "assets/cat_10/v_-ypKYhjrlXw_1_1064/0020s_00600_mask.png", "mask_status": "available" }, { "frame": "00628", "timestamp_sec": 21.0, "offset_sec": 21, "rgb_path": "assets/cat_10/v_-ypKYhjrlXw_1_1064/0021s_00628_rgb.jpg", "mask_path": "assets/cat_10/v_-ypKYhjrlXw_1_1064/0021s_00628_mask.png", "mask_status": "available" }, { "frame": "00660", "timestamp_sec": 22.0, "offset_sec": 22, "rgb_path": "assets/cat_10/v_-ypKYhjrlXw_1_1064/0022s_00660_rgb.jpg", "mask_path": "assets/cat_10/v_-ypKYhjrlXw_1_1064/0022s_00660_mask.png", "mask_status": "available" }, { "frame": "00688", "timestamp_sec": 23.0, "offset_sec": 23, "rgb_path": "assets/cat_10/v_-ypKYhjrlXw_1_1064/0023s_00688_rgb.jpg", "mask_path": "assets/cat_10/v_-ypKYhjrlXw_1_1064/0023s_00688_mask.png", "mask_status": "available" }, { "frame": "00720", "timestamp_sec": 24.0, "offset_sec": 24, "rgb_path": "assets/cat_10/v_-ypKYhjrlXw_1_1064/0024s_00720_rgb.jpg", "mask_path": "assets/cat_10/v_-ypKYhjrlXw_1_1064/0024s_00720_mask.png", "mask_status": "available" }, { "frame": "00748", "timestamp_sec": 25.0, "offset_sec": 25, "rgb_path": "assets/cat_10/v_-ypKYhjrlXw_1_1064/0025s_00748_rgb.jpg", "mask_path": "assets/cat_10/v_-ypKYhjrlXw_1_1064/0025s_00748_mask.png", "mask_status": "available" }, { "frame": "00780", "timestamp_sec": 26.0, "offset_sec": 26, "rgb_path": "assets/cat_10/v_-ypKYhjrlXw_1_1064/0026s_00780_rgb.jpg", "mask_path": "assets/cat_10/v_-ypKYhjrlXw_1_1064/0026s_00780_mask.png", "mask_status": "available" }, { "frame": "00808", "timestamp_sec": 27.0, "offset_sec": 27, "rgb_path": "assets/cat_10/v_-ypKYhjrlXw_1_1064/0027s_00808_rgb.jpg", "mask_path": "assets/cat_10/v_-ypKYhjrlXw_1_1064/0027s_00808_mask.png", "mask_status": "available" }, { "frame": "00840", "timestamp_sec": 28.0, "offset_sec": 28, "rgb_path": "assets/cat_10/v_-ypKYhjrlXw_1_1064/0028s_00840_rgb.jpg", "mask_path": "assets/cat_10/v_-ypKYhjrlXw_1_1064/0028s_00840_mask.png", "mask_status": "available" }, { "frame": "00868", "timestamp_sec": 29.0, "offset_sec": 29, "rgb_path": "assets/cat_10/v_-ypKYhjrlXw_1_1064/0029s_00868_rgb.jpg", "mask_path": "assets/cat_10/v_-ypKYhjrlXw_1_1064/0029s_00868_mask.png", "mask_status": "available" }, { "frame": "00900", "timestamp_sec": 30.0, "offset_sec": 30, "rgb_path": "assets/cat_10/v_-ypKYhjrlXw_1_1064/0030s_00900_rgb.jpg", "mask_path": "assets/cat_10/v_-ypKYhjrlXw_1_1064/0030s_00900_mask.png", "mask_status": "available" }, { "frame": "00928", "timestamp_sec": 31.0, "offset_sec": 31, "rgb_path": "assets/cat_10/v_-ypKYhjrlXw_1_1064/0031s_00928_rgb.jpg", "mask_path": "assets/cat_10/v_-ypKYhjrlXw_1_1064/0031s_00928_mask.png", "mask_status": "available" }, { "frame": "00960", "timestamp_sec": 32.0, "offset_sec": 32, "rgb_path": "assets/cat_10/v_-ypKYhjrlXw_1_1064/0032s_00960_rgb.jpg", "mask_path": "assets/cat_10/v_-ypKYhjrlXw_1_1064/0032s_00960_mask.png", "mask_status": "available" }, { "frame": "00988", "timestamp_sec": 33.0, "offset_sec": 33, "rgb_path": "assets/cat_10/v_-ypKYhjrlXw_1_1064/0033s_00988_rgb.jpg", "mask_path": "assets/cat_10/v_-ypKYhjrlXw_1_1064/0033s_00988_mask.png", "mask_status": "available" }, { "frame": "01020", "timestamp_sec": 34.0, "offset_sec": 34, "rgb_path": "assets/cat_10/v_-ypKYhjrlXw_1_1064/0034s_01020_rgb.jpg", "mask_path": "assets/cat_10/v_-ypKYhjrlXw_1_1064/0034s_01020_mask.png", "mask_status": "available" }, { "frame": "01048", "timestamp_sec": 35.0, "offset_sec": 35, "rgb_path": "assets/cat_10/v_-ypKYhjrlXw_1_1064/0035s_01048_rgb.jpg", "mask_path": "assets/cat_10/v_-ypKYhjrlXw_1_1064/0035s_01048_mask.png", "mask_status": "available" } ], "obj_id": [ 0 ], "anno_id": [ 136 ], "fps_inferred_from_clip_interval": 29.9437 }, "usage": { "input_tokens": 193574, "input_tokens_details": { "cache_write_tokens": 193571, "cached_tokens": 0 }, "output_tokens": 3651, "output_tokens_details": { "reasoning_tokens": 2262 }, "total_tokens": 197225 }, "model": "gpt-5.6-sol", "effort": "medium", "custom_id": "cat10_cap01_v_-ypKYhjrlXw_1_1064", "annotation": { "caption": "A man in a white lab coat and red tie sits at a table and speaks, occasionally gesturing with his hands. In front of him are several white bottles and containers, and a woman's arm is resting on the table to his left.", "entity_groups": [ { "id": "entity_1", "label": "man", "spans": [ { "text": "A man", "range": [ 0, 5 ] }, { "text": "his", "range": [ 94, 97 ] }, { "text": "him", "range": [ 117, 120 ] }, { "text": "his", "range": [ 208, 211 ] } ], "source_mask_link": "selected_person" }, { "id": "entity_2", "label": "lab coat", "spans": [ { "text": "a white lab coat", "range": [ 9, 25 ] } ], "source_mask_link": null }, { "id": "entity_3", "label": "tie", "spans": [ { "text": "red tie", "range": [ 30, 37 ] } ], "source_mask_link": null }, { "id": "entity_4", "label": "table", "spans": [ { "text": "a table", "range": [ 46, 53 ] }, { "text": "the table", "range": [ 195, 204 ] } ], "source_mask_link": null }, { "id": "entity_5", "label": "hands", "spans": [ { "text": "his hands", "range": [ 94, 103 ] } ], "source_mask_link": null }, { "id": "entity_6", "label": "bottles", "spans": [ { "text": "several white bottles", "range": [ 125, 146 ] } ], "source_mask_link": null }, { "id": "entity_7", "label": "containers", "spans": [ { "text": "containers", "range": [ 151, 161 ] } ], "source_mask_link": null }, { "id": "entity_8", "label": "woman", "spans": [ { "text": "woman's", "range": [ 169, 176 ] } ], "source_mask_link": null }, { "id": "entity_9", "label": "arm", "spans": [ { "text": "a woman's arm", "range": [ 167, 180 ] } ], "source_mask_link": null } ], "actions": [ { "id": "action_1", "mentions": [ { "text": "sits", "ranges": [ [ 38, 42 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "type": "posture", "subject_refs": [ "entity_1" ], "object_refs": [ "entity_4" ], "confidence": "high", "temporal_status": "localized", "intervals": [ { "start_s": 0.0, "end_s": 35.5, "source_start_s": 0.0, "source_end_s": 35.5 } ], "evidence": "The mask-selected man remains in a seated posture at the table throughout all sampled frames.", "mask_relation": "The action subject matches the supplied selected-person mask.", "source_subject_refs": [ "entity_1" ] }, { "id": "action_2", "mentions": [ { "text": "speaks", "ranges": [ [ 58, 64 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "type": "simple", "subject_refs": [ "entity_1" ], "object_refs": [], "confidence": "high", "temporal_status": "localized", "intervals": [ { "start_s": 1.0, "end_s": 35.5, "source_start_s": 1.0, "source_end_s": 35.5 } ], "evidence": "Across the sampled sequence, the mask-selected man's mouth repeatedly changes between open and closed speech configurations, continuing through the final frame.", "mask_relation": "The action subject matches the supplied selected-person mask.", "source_subject_refs": [ "entity_1" ] }, { "id": "action_3", "mentions": [ { "text": "gesturing", "ranges": [ [ 79, 88 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "type": "simple", "subject_refs": [ "entity_1" ], "object_refs": [ "entity_5" ], "confidence": "high", "temporal_status": "localized", "intervals": [ { "start_s": 29.0, "end_s": 30.0, "source_start_s": 29.0, "source_end_s": 30.0 } ], "evidence": "At 29.0 seconds the mask-selected man raises an open hand in a clear conversational gesture; it is lowered again by 30.0 seconds.", "mask_relation": "The action subject matches the supplied selected-person mask.", "source_subject_refs": [ "entity_1" ] } ], "notes": [ "The supplied masklet consistently identifies the seated man and no other person or object.", "The coat, tie, table, hands, bottles, containers, woman, and woman's arm have no separate supplied source masks, so their source-mask links are null.", "The captioned action \"is resting\" has the woman's arm as its subject. It is omitted from actions because neither the woman nor her arm has a corresponding supplied mask." ] }, "model_output_raw": { "caption": "A man in a white lab coat and red tie sits at a table and speaks, occasionally gesturing with his hands. In front of him are several white bottles and containers, and a woman's arm is resting on the table to his left.", "entity_groups": [ { "id": "entity_1", "label": "man", "spans": [ { "text": "A man", "range": [ 0, 5 ] }, { "text": "his", "range": [ 94, 97 ] }, { "text": "him", "range": [ 117, 120 ] }, { "text": "his", "range": [ 208, 211 ] } ], "source_mask": "selected_person_mask" }, { "id": "entity_2", "label": "lab coat", "spans": [ { "text": "a white lab coat", "range": [ 9, 25 ] } ], "source_mask": null }, { "id": "entity_3", "label": "tie", "spans": [ { "text": "red tie", "range": [ 30, 37 ] } ], "source_mask": null }, { "id": "entity_4", "label": "table", "spans": [ { "text": "a table", "range": [ 46, 53 ] }, { "text": "the table", "range": [ 195, 204 ] } ], "source_mask": null }, { "id": "entity_5", "label": "hands", "spans": [ { "text": "his hands", "range": [ 94, 103 ] } ], "source_mask": null }, { "id": "entity_6", "label": "bottles", "spans": [ { "text": "several white bottles", "range": [ 125, 146 ] } ], "source_mask": null }, { "id": "entity_7", "label": "containers", "spans": [ { "text": "containers", "range": [ 151, 161 ] } ], "source_mask": null }, { "id": "entity_8", "label": "woman", "spans": [ { "text": "woman's", "range": [ 169, 176 ] } ], "source_mask": null }, { "id": "entity_9", "label": "arm", "spans": [ { "text": "a woman's arm", "range": [ 167, 180 ] } ], "source_mask": null } ], "actions": [ { "id": "action_1", "mentions": [ { "text": "sits", "occurrence": 1, "ranges": [ [ 38, 42 ] ], "particle": { "text": null, "occurrence": 1 } } ], "type": "posture", "subject_refs": [ "entity_1" ], "object_refs": [ "entity_4" ], "confidence": "high", "temporal_status": "localized", "intervals": [ { "start_s": 0.0, "end_s": 35.5 } ], "evidence": "The mask-selected man remains in a seated posture at the table throughout all sampled frames.", "mask_relation": "The action subject matches the supplied selected-person mask." }, { "id": "action_2", "mentions": [ { "text": "speaks", "occurrence": 1, "ranges": [ [ 58, 64 ] ], "particle": { "text": null, "occurrence": 1 } } ], "type": "simple", "subject_refs": [ "entity_1" ], "object_refs": [], "confidence": "high", "temporal_status": "localized", "intervals": [ { "start_s": 1.0, "end_s": 35.5 } ], "evidence": "Across the sampled sequence, the mask-selected man's mouth repeatedly changes between open and closed speech configurations, continuing through the final frame.", "mask_relation": "The action subject matches the supplied selected-person mask." }, { "id": "action_3", "mentions": [ { "text": "gesturing", "occurrence": 1, "ranges": [ [ 79, 88 ] ], "particle": { "text": null, "occurrence": 1 } } ], "type": "simple", "subject_refs": [ "entity_1" ], "object_refs": [ "entity_5" ], "confidence": "high", "temporal_status": "localized", "intervals": [ { "start_s": 29.0, "end_s": 30.0 } ], "evidence": "At 29.0 seconds the mask-selected man raises an open hand in a clear conversational gesture; it is lowered again by 30.0 seconds.", "mask_relation": "The action subject matches the supplied selected-person mask." } ], "notes": [ "The supplied masklet consistently identifies the seated man and no other person or object.", "The coat, tie, table, hands, bottles, containers, woman, and woman's arm have no separate supplied source masks, so their source-mask links are null.", "The captioned action \"is resting\" has the woman's arm as its subject. It is omitted from actions because neither the woman nor her arm has a corresponding supplied mask." ] }, "normalization_repairs": 0 }, { "source_video_id": "v_-ypKYhjrlXw", "caption_count_category": 10, "clip": { "caption_index": 2, "caption": "The man gestures towards the various products on the table, pointing to different bottles. He then picks up a round compact, opens it to show the mirror and powder inside, and places it back on the table.", "source_interval_sec": [ 35.5, 48.5 ], "loom_clip_id": "v_-ypKYhjrlXw_1064_1454", "clip_frame_interval_4fps": [ 1064, 1454 ], "sample_count_1fps": 13, "sampled_frames": [ { "frame": "00000", "timestamp_sec": 35.5, "offset_sec": 0, "rgb_path": "assets/cat_10/v_-ypKYhjrlXw_1064_1454/0000s_00000_rgb.jpg", "mask_path": "assets/cat_10/v_-ypKYhjrlXw_1064_1454/0000s_00000_mask.png", "mask_status": "available" }, { "frame": "00028", "timestamp_sec": 36.5, "offset_sec": 1, "rgb_path": "assets/cat_10/v_-ypKYhjrlXw_1064_1454/0001s_00028_rgb.jpg", "mask_path": "assets/cat_10/v_-ypKYhjrlXw_1064_1454/0001s_00028_mask.png", "mask_status": "available" }, { "frame": "00060", "timestamp_sec": 37.5, "offset_sec": 2, "rgb_path": "assets/cat_10/v_-ypKYhjrlXw_1064_1454/0002s_00060_rgb.jpg", "mask_path": "assets/cat_10/v_-ypKYhjrlXw_1064_1454/0002s_00060_mask.png", "mask_status": "available" }, { "frame": "00088", "timestamp_sec": 38.5, "offset_sec": 3, "rgb_path": "assets/cat_10/v_-ypKYhjrlXw_1064_1454/0003s_00088_rgb.jpg", "mask_path": "assets/cat_10/v_-ypKYhjrlXw_1064_1454/0003s_00088_mask.png", "mask_status": "available" }, { "frame": "00120", "timestamp_sec": 39.5, "offset_sec": 4, "rgb_path": "assets/cat_10/v_-ypKYhjrlXw_1064_1454/0004s_00120_rgb.jpg", "mask_path": "assets/cat_10/v_-ypKYhjrlXw_1064_1454/0004s_00120_mask.png", "mask_status": "available" }, { "frame": "00148", "timestamp_sec": 40.5, "offset_sec": 5, "rgb_path": "assets/cat_10/v_-ypKYhjrlXw_1064_1454/0005s_00148_rgb.jpg", "mask_path": "assets/cat_10/v_-ypKYhjrlXw_1064_1454/0005s_00148_mask.png", "mask_status": "available" }, { "frame": "00180", "timestamp_sec": 41.5, "offset_sec": 6, "rgb_path": "assets/cat_10/v_-ypKYhjrlXw_1064_1454/0006s_00180_rgb.jpg", "mask_path": "assets/cat_10/v_-ypKYhjrlXw_1064_1454/0006s_00180_mask.png", "mask_status": "available" }, { "frame": "00208", "timestamp_sec": 42.5, "offset_sec": 7, "rgb_path": "assets/cat_10/v_-ypKYhjrlXw_1064_1454/0007s_00208_rgb.jpg", "mask_path": "assets/cat_10/v_-ypKYhjrlXw_1064_1454/0007s_00208_mask.png", "mask_status": "available" }, { "frame": "00240", "timestamp_sec": 43.5, "offset_sec": 8, "rgb_path": "assets/cat_10/v_-ypKYhjrlXw_1064_1454/0008s_00240_rgb.jpg", "mask_path": "assets/cat_10/v_-ypKYhjrlXw_1064_1454/0008s_00240_mask.png", "mask_status": "available" }, { "frame": "00268", "timestamp_sec": 44.5, "offset_sec": 9, "rgb_path": "assets/cat_10/v_-ypKYhjrlXw_1064_1454/0009s_00268_rgb.jpg", "mask_path": "assets/cat_10/v_-ypKYhjrlXw_1064_1454/0009s_00268_mask.png", "mask_status": "available" }, { "frame": "00300", "timestamp_sec": 45.5, "offset_sec": 10, "rgb_path": "assets/cat_10/v_-ypKYhjrlXw_1064_1454/0010s_00300_rgb.jpg", "mask_path": "assets/cat_10/v_-ypKYhjrlXw_1064_1454/0010s_00300_mask.png", "mask_status": "available" }, { "frame": "00328", "timestamp_sec": 46.5, "offset_sec": 11, "rgb_path": "assets/cat_10/v_-ypKYhjrlXw_1064_1454/0011s_00328_rgb.jpg", "mask_path": "assets/cat_10/v_-ypKYhjrlXw_1064_1454/0011s_00328_mask.png", "mask_status": "available" }, { "frame": "00360", "timestamp_sec": 47.5, "offset_sec": 12, "rgb_path": "assets/cat_10/v_-ypKYhjrlXw_1064_1454/0012s_00360_rgb.jpg", "mask_path": "assets/cat_10/v_-ypKYhjrlXw_1064_1454/0012s_00360_mask.png", "mask_status": "available" } ], "obj_id": [ 0 ], "anno_id": [ 133 ], "fps_inferred_from_clip_interval": 30.0 }, "usage": { "input_tokens": 79695, "input_tokens_details": { "cache_write_tokens": 79692, "cached_tokens": 0 }, "output_tokens": 4281, "output_tokens_details": { "reasoning_tokens": 2588 }, "total_tokens": 83976 }, "model": "gpt-5.6-sol", "effort": "medium", "custom_id": "cat10_cap02_v_-ypKYhjrlXw_1064_1454", "annotation": { "caption": "The man gestures towards the various products on the table, pointing to different bottles. He then picks up a round compact, opens it to show the mirror and powder inside, and places it back on the table.", "entity_groups": [ { "id": "man_1", "label": "man", "spans": [ { "text": "The man", "range": [ 0, 7 ] }, { "text": "He", "range": [ 91, 93 ] } ], "source_mask_link": "selected_person" }, { "id": "products_1", "label": "products", "spans": [ { "text": "the various products", "range": [ 25, 45 ] } ], "source_mask_link": null }, { "id": "table_1", "label": "table", "spans": [ { "text": "the table", "range": [ 49, 58 ] }, { "text": "the table", "range": [ 194, 203 ] } ], "source_mask_link": null }, { "id": "bottles_1", "label": "bottles", "spans": [ { "text": "different bottles", "range": [ 72, 89 ] } ], "source_mask_link": null }, { "id": "compact_1", "label": "compact", "spans": [ { "text": "a round compact", "range": [ 108, 123 ] }, { "text": "it", "range": [ 131, 133 ] }, { "text": "it", "range": [ 183, 185 ] } ], "source_mask_link": null }, { "id": "mirror_1", "label": "mirror", "spans": [ { "text": "the mirror", "range": [ 142, 152 ] } ], "source_mask_link": null }, { "id": "powder_1", "label": "powder", "spans": [ { "text": "powder", "range": [ 157, 163 ] } ], "source_mask_link": null } ], "actions": [ { "id": "action_1", "mentions": [ { "text": "gestures", "ranges": [ [ 8, 16 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } }, { "text": "pointing", "ranges": [ [ 60, 68 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "type": "prepositional", "subject_refs": [ "man_1" ], "object_refs": [ "products_1", "bottles_1" ], "confidence": "high", "temporal_status": "localized", "intervals": [ { "start_s": 0.0, "end_s": 11.0, "source_start_s": 35.5, "source_end_s": 46.5 } ], "evidence": "Across samples at 0–10 seconds, the masked man repeatedly raises, sweeps, and directs his hand toward the bottles and other products.", "mask_relation": "subject matches the supplied selected-person mask", "source_subject_refs": [ "man_1" ] }, { "id": "action_2", "mentions": [ { "text": "picks up", "ranges": [ [ 99, 107 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "type": "phrasal", "subject_refs": [ "man_1" ], "object_refs": [ "compact_1" ], "confidence": "high", "temporal_status": "localized", "intervals": [ { "start_s": 11.0, "end_s": 12.0, "source_start_s": 46.5, "source_end_s": 47.5 } ], "evidence": "At 11 seconds the masked man's hand contacts the compact; by 12 seconds he has lifted it from the table.", "mask_relation": "subject matches the supplied selected-person mask", "source_subject_refs": [ "man_1" ] }, { "id": "action_3", "mentions": [ { "text": "opens", "ranges": [ [ 125, 130 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "type": "simple", "subject_refs": [ "man_1" ], "object_refs": [ "compact_1" ], "confidence": "high", "temporal_status": "not_localized", "intervals": [], "evidence": "No opening transition is visible in the sampled frames; the compact already appears open before it is lifted.", "mask_relation": "subject matches the supplied selected-person mask", "source_subject_refs": [ "man_1" ] }, { "id": "action_4", "mentions": [ { "text": "show", "ranges": [ [ 137, 141 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "type": "simple", "subject_refs": [ "man_1" ], "object_refs": [ "mirror_1", "powder_1" ], "confidence": "high", "temporal_status": "localized", "intervals": [ { "start_s": 12.0, "end_s": 13.0, "source_start_s": 47.5, "source_end_s": 48.5 } ], "evidence": "In the final sampled frame, the masked man holds the open compact upright with its mirror and powder facing outward.", "mask_relation": "subject matches the supplied selected-person mask", "source_subject_refs": [ "man_1" ] }, { "id": "action_5", "mentions": [ { "text": "places", "ranges": [ [ 176, 182 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "type": "simple", "subject_refs": [ "man_1" ], "object_refs": [ "compact_1" ], "confidence": "high", "temporal_status": "not_localized", "intervals": [], "evidence": "The compact remains held by the man in the final frame, so placement back onto the table is not supported by the supplied samples.", "mask_relation": "subject matches the supplied selected-person mask", "source_subject_refs": [ "man_1" ] } ], "notes": [ "The supplied masklet consistently identifies the man and includes his moving arms and hands.", "No source masks were supplied for the products, table, bottles, compact, mirror, or powder.", "The woman visible in the frames is not mentioned in the caption and therefore has no text entity group.", "The gesture and pointing predicates are grouped as two mentions of the same continuing demonstrative event." ] }, "model_output_raw": { "caption": "The man gestures towards the various products on the table, pointing to different bottles. He then picks up a round compact, opens it to show the mirror and powder inside, and places it back on the table.", "entity_groups": [ { "id": "man_1", "label": "man", "spans": [ { "text": "The man", "range": [ 0, 7 ] }, { "text": "He", "range": [ 91, 93 ] } ], "source_mask": "selected_person_masklet" }, { "id": "products_1", "label": "products", "spans": [ { "text": "the various products", "range": [ 25, 45 ] } ], "source_mask": null }, { "id": "table_1", "label": "table", "spans": [ { "text": "the table", "range": [ 49, 58 ] }, { "text": "the table", "range": [ 194, 203 ] } ], "source_mask": null }, { "id": "bottles_1", "label": "bottles", "spans": [ { "text": "different bottles", "range": [ 72, 89 ] } ], "source_mask": null }, { "id": "compact_1", "label": "compact", "spans": [ { "text": "a round compact", "range": [ 108, 123 ] }, { "text": "it", "range": [ 131, 133 ] }, { "text": "it", "range": [ 183, 185 ] } ], "source_mask": null }, { "id": "mirror_1", "label": "mirror", "spans": [ { "text": "the mirror", "range": [ 142, 152 ] } ], "source_mask": null }, { "id": "powder_1", "label": "powder", "spans": [ { "text": "powder", "range": [ 157, 163 ] } ], "source_mask": null } ], "actions": [ { "id": "action_1", "mentions": [ { "text": "gestures", "ranges": [ [ 8, 16 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } }, { "text": "pointing", "ranges": [ [ 60, 68 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "type": "prepositional", "subject_refs": [ "man_1" ], "object_refs": [ "products_1", "bottles_1" ], "confidence": "high", "mask_relation": "subject matches the supplied selected-person mask", "temporal_status": "localized", "intervals": [ { "start_s": 0.0, "end_s": 11.0, "source_start_s": 35.5, "source_end_s": 46.5 } ], "evidence": "Across samples at 0–10 seconds, the masked man repeatedly raises, sweeps, and directs his hand toward the bottles and other products." }, { "id": "action_2", "mentions": [ { "text": "picks up", "ranges": [ [ 99, 107 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "type": "phrasal", "subject_refs": [ "man_1" ], "object_refs": [ "compact_1" ], "confidence": "high", "mask_relation": "subject matches the supplied selected-person mask", "temporal_status": "localized", "intervals": [ { "start_s": 11.0, "end_s": 12.0, "source_start_s": 46.5, "source_end_s": 47.5 } ], "evidence": "At 11 seconds the masked man's hand contacts the compact; by 12 seconds he has lifted it from the table." }, { "id": "action_3", "mentions": [ { "text": "opens", "ranges": [ [ 125, 130 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "type": "simple", "subject_refs": [ "man_1" ], "object_refs": [ "compact_1" ], "confidence": "high", "mask_relation": "subject matches the supplied selected-person mask", "temporal_status": "not_localized", "intervals": [], "evidence": "No opening transition is visible in the sampled frames; the compact already appears open before it is lifted." }, { "id": "action_4", "mentions": [ { "text": "show", "ranges": [ [ 137, 141 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "type": "simple", "subject_refs": [ "man_1" ], "object_refs": [ "mirror_1", "powder_1" ], "confidence": "high", "mask_relation": "subject matches the supplied selected-person mask", "temporal_status": "localized", "intervals": [ { "start_s": 12.0, "end_s": 13.0, "source_start_s": 47.5, "source_end_s": 48.5 } ], "evidence": "In the final sampled frame, the masked man holds the open compact upright with its mirror and powder facing outward." }, { "id": "action_5", "mentions": [ { "text": "places", "ranges": [ [ 176, 182 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "type": "simple", "subject_refs": [ "man_1" ], "object_refs": [ "compact_1" ], "confidence": "high", "mask_relation": "subject matches the supplied selected-person mask", "temporal_status": "not_localized", "intervals": [], "evidence": "The compact remains held by the man in the final frame, so placement back onto the table is not supported by the supplied samples." } ], "notes": [ "The supplied masklet consistently identifies the man and includes his moving arms and hands.", "No source masks were supplied for the products, table, bottles, compact, mirror, or powder.", "The woman visible in the frames is not mentioned in the caption and therefore has no text entity group.", "The gesture and pointing predicates are grouped as two mentions of the same continuing demonstrative event." ] }, "normalization_repairs": 0 }, { "source_video_id": "v_-ypKYhjrlXw", "caption_count_category": 10, "clip": { "caption_index": 3, "caption": "The man picks up a tall white spray bottle with his left hand and holds it up while gesturing with his right hand. He continues speaking, explaining the product while holding it.", "source_interval_sec": [ 48.5, 59.0 ], "loom_clip_id": "v_-ypKYhjrlXw_1454_1769", "clip_frame_interval_4fps": [ 1454, 1769 ], "sample_count_1fps": 11, "sampled_frames": [ { "frame": "00000", "timestamp_sec": 48.5, "offset_sec": 0, "rgb_path": "assets/cat_10/v_-ypKYhjrlXw_1454_1769/0000s_00000_rgb.jpg", "mask_path": "assets/cat_10/v_-ypKYhjrlXw_1454_1769/0000s_00000_mask.png", "mask_status": "available" }, { "frame": "00028", "timestamp_sec": 49.5, "offset_sec": 1, "rgb_path": "assets/cat_10/v_-ypKYhjrlXw_1454_1769/0001s_00028_rgb.jpg", "mask_path": "assets/cat_10/v_-ypKYhjrlXw_1454_1769/0001s_00028_mask.png", "mask_status": "available" }, { "frame": "00060", "timestamp_sec": 50.5, "offset_sec": 2, "rgb_path": "assets/cat_10/v_-ypKYhjrlXw_1454_1769/0002s_00060_rgb.jpg", "mask_path": "assets/cat_10/v_-ypKYhjrlXw_1454_1769/0002s_00060_mask.png", "mask_status": "available" }, { "frame": "00088", "timestamp_sec": 51.5, "offset_sec": 3, "rgb_path": "assets/cat_10/v_-ypKYhjrlXw_1454_1769/0003s_00088_rgb.jpg", "mask_path": "assets/cat_10/v_-ypKYhjrlXw_1454_1769/0003s_00088_mask.png", "mask_status": "available" }, { "frame": "00120", "timestamp_sec": 52.5, "offset_sec": 4, "rgb_path": "assets/cat_10/v_-ypKYhjrlXw_1454_1769/0004s_00120_rgb.jpg", "mask_path": "assets/cat_10/v_-ypKYhjrlXw_1454_1769/0004s_00120_mask.png", "mask_status": "available" }, { "frame": "00148", "timestamp_sec": 53.5, "offset_sec": 5, "rgb_path": "assets/cat_10/v_-ypKYhjrlXw_1454_1769/0005s_00148_rgb.jpg", "mask_path": "assets/cat_10/v_-ypKYhjrlXw_1454_1769/0005s_00148_mask.png", "mask_status": "available" }, { "frame": "00180", "timestamp_sec": 54.5, "offset_sec": 6, "rgb_path": "assets/cat_10/v_-ypKYhjrlXw_1454_1769/0006s_00180_rgb.jpg", "mask_path": "assets/cat_10/v_-ypKYhjrlXw_1454_1769/0006s_00180_mask.png", "mask_status": "available" }, { "frame": "00208", "timestamp_sec": 55.5, "offset_sec": 7, "rgb_path": "assets/cat_10/v_-ypKYhjrlXw_1454_1769/0007s_00208_rgb.jpg", "mask_path": "assets/cat_10/v_-ypKYhjrlXw_1454_1769/0007s_00208_mask.png", "mask_status": "available" }, { "frame": "00240", "timestamp_sec": 56.5, "offset_sec": 8, "rgb_path": "assets/cat_10/v_-ypKYhjrlXw_1454_1769/0008s_00240_rgb.jpg", "mask_path": "assets/cat_10/v_-ypKYhjrlXw_1454_1769/0008s_00240_mask.png", "mask_status": "available" }, { "frame": "00268", "timestamp_sec": 57.5, "offset_sec": 9, "rgb_path": "assets/cat_10/v_-ypKYhjrlXw_1454_1769/0009s_00268_rgb.jpg", "mask_path": "assets/cat_10/v_-ypKYhjrlXw_1454_1769/0009s_00268_mask.png", "mask_status": "available" }, { "frame": "00300", "timestamp_sec": 58.5, "offset_sec": 10, "rgb_path": "assets/cat_10/v_-ypKYhjrlXw_1454_1769/0010s_00300_rgb.jpg", "mask_path": "assets/cat_10/v_-ypKYhjrlXw_1454_1769/0010s_00300_mask.png", "mask_status": "available" } ], "obj_id": [ 0 ], "anno_id": [ 134 ], "fps_inferred_from_clip_interval": 30.0 }, "usage": { "input_tokens": 69787, "input_tokens_details": { "cache_write_tokens": 69784, "cached_tokens": 0 }, "output_tokens": 4200, "output_tokens_details": { "reasoning_tokens": 2501 }, "total_tokens": 73987 }, "model": "gpt-5.6-sol", "effort": "medium", "custom_id": "cat10_cap03_v_-ypKYhjrlXw_1454_1769", "annotation": { "caption": "The man picks up a tall white spray bottle with his left hand and holds it up while gesturing with his right hand. He continues speaking, explaining the product while holding it.", "entity_groups": [ { "id": "person_1", "label": "man", "spans": [ { "text": "The man", "range": [ 0, 7 ] }, { "text": "his", "range": [ 48, 51 ] }, { "text": "his", "range": [ 99, 102 ] }, { "text": "He", "range": [ 115, 117 ] } ], "source_mask_link": "selected_person" }, { "id": "spray_bottle_1", "label": "spray bottle", "spans": [ { "text": "a tall white spray bottle", "range": [ 17, 42 ] }, { "text": "it", "range": [ 72, 74 ] }, { "text": "the product", "range": [ 149, 160 ] }, { "text": "it", "range": [ 175, 177 ] } ], "source_mask_link": null }, { "id": "left_hand_1", "label": "left hand", "spans": [ { "text": "his left hand", "range": [ 48, 61 ] } ], "source_mask_link": null }, { "id": "right_hand_1", "label": "right hand", "spans": [ { "text": "his right hand", "range": [ 99, 113 ] } ], "source_mask_link": null } ], "actions": [ { "id": "action_1", "mentions": [ { "text": "picks up", "ranges": [ [ 8, 16 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "type": "phrasal", "subject_refs": [ "person_1" ], "object_refs": [ "spray_bottle_1" ], "confidence": "high", "temporal_status": "supported", "intervals": [ { "start_s": 1.0, "end_s": 3.0, "source_start_s": 49.5, "source_end_s": 51.5 } ], "evidence": "At 1.0 s the masked man reaches for and grasps the tall bottle; by 2.0 s he has lifted it from the table.", "mask_relation": "The action subject matches the supplied selected-person mask.", "source_subject_refs": [ "person_1" ] }, { "id": "action_2", "mentions": [ { "text": "holds", "ranges": [ [ 66, 71 ] ], "occurrence": 1, "particle": { "text": "up", "occurrence": 2 } }, { "text": "holding", "ranges": [ [ 167, 174 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "type": "phrasal", "subject_refs": [ "person_1" ], "object_refs": [ "spray_bottle_1" ], "confidence": "high", "temporal_status": "supported", "intervals": [ { "start_s": 2.0, "end_s": 10.5, "source_start_s": 50.5, "source_end_s": 59.0 } ], "evidence": "From 2.0 s through the final sample, the masked man visibly keeps the white bottle raised in his left hand.", "mask_relation": "The action subject matches the supplied selected-person mask.", "source_subject_refs": [ "person_1" ] }, { "id": "action_3", "mentions": [ { "text": "gesturing", "ranges": [ [ 84, 93 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "type": "simple", "subject_refs": [ "person_1" ], "object_refs": [], "confidence": "high", "temporal_status": "supported", "intervals": [ { "start_s": 2.0, "end_s": 3.0, "source_start_s": 50.5, "source_end_s": 51.5 }, { "start_s": 4.0, "end_s": 5.0, "source_start_s": 52.5, "source_end_s": 53.5 }, { "start_s": 6.0, "end_s": 10.5, "source_start_s": 54.5, "source_end_s": 59.0 } ], "evidence": "The masked man repeatedly raises, opens, and moves his free right hand while displaying the bottle; frames where that hand is instead handling the bottle are excluded.", "mask_relation": "The action subject matches the supplied selected-person mask.", "source_subject_refs": [ "person_1" ] }, { "id": "action_4", "mentions": [ { "text": "speaking", "ranges": [ [ 128, 136 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } }, { "text": "explaining", "ranges": [ [ 138, 148 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "type": "simple", "subject_refs": [ "person_1" ], "object_refs": [ "spray_bottle_1" ], "confidence": "medium", "temporal_status": "supported", "intervals": [ { "start_s": 0.0, "end_s": 10.5, "source_start_s": 48.5, "source_end_s": 59.0 } ], "evidence": "Across the sampled sequence, the masked man addresses the viewer with changing mouth positions while displaying the product and accompanying his speech with gestures; the explanatory content itself is not visually recoverable.", "mask_relation": "The action subject matches the supplied selected-person mask.", "source_subject_refs": [ "person_1" ] } ], "notes": [ "The supplied mask consistently identifies the man described by “The man” and “He.”", "No independent masks were supplied for the spray bottle or either hand, so those entity groups have null source-mask links.", "The visible woman is not mentioned in the immutable caption and therefore has no text entity group.", "The two holding expressions describe the same continuing bottle-holding event. “continues” is excluded as an aspectual verb.", "Temporal intervals are sparse support intervals derived from the sampled clean RGB frames and are bounded by the original source interval [48.5, 59.0)." ] }, "model_output_raw": { "caption": "The man picks up a tall white spray bottle with his left hand and holds it up while gesturing with his right hand. He continues speaking, explaining the product while holding it.", "entity_groups": [ { "id": "person_1", "label": "man", "spans": [ { "text": "The man", "range": [ 0, 7 ] }, { "text": "his", "range": [ 48, 51 ] }, { "text": "his", "range": [ 99, 102 ] }, { "text": "He", "range": [ 115, 117 ] } ], "source_mask": "selected_person_mask" }, { "id": "spray_bottle_1", "label": "spray bottle", "spans": [ { "text": "a tall white spray bottle", "range": [ 17, 42 ] }, { "text": "it", "range": [ 72, 74 ] }, { "text": "the product", "range": [ 149, 160 ] }, { "text": "it", "range": [ 175, 177 ] } ], "source_mask": null }, { "id": "left_hand_1", "label": "left hand", "spans": [ { "text": "his left hand", "range": [ 48, 61 ] } ], "source_mask": null }, { "id": "right_hand_1", "label": "right hand", "spans": [ { "text": "his right hand", "range": [ 99, 113 ] } ], "source_mask": null } ], "actions": [ { "id": "action_1", "mentions": [ { "text": "picks up", "ranges": [ [ 8, 16 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "type": "phrasal", "subject_refs": [ "person_1" ], "object_refs": [ "spray_bottle_1" ], "related_entity_refs": [ "left_hand_1" ], "confidence": "high", "temporal_status": "supported", "intervals": [ { "clip_relative_s": [ 1.0, 3.0 ], "source_video_s": [ 49.5, 51.5 ] } ], "evidence": "At 1.0 s the masked man reaches for and grasps the tall bottle; by 2.0 s he has lifted it from the table.", "mask_relation": "The action subject matches the supplied selected-person mask." }, { "id": "action_2", "mentions": [ { "text": "holds", "ranges": [ [ 66, 71 ] ], "occurrence": 1, "particle": { "text": "up", "occurrence": 2 } }, { "text": "holding", "ranges": [ [ 167, 174 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "type": "phrasal", "subject_refs": [ "person_1" ], "object_refs": [ "spray_bottle_1" ], "related_entity_refs": [ "left_hand_1" ], "confidence": "high", "temporal_status": "supported", "intervals": [ { "clip_relative_s": [ 2.0, 10.5 ], "source_video_s": [ 50.5, 59.0 ] } ], "evidence": "From 2.0 s through the final sample, the masked man visibly keeps the white bottle raised in his left hand.", "mask_relation": "The action subject matches the supplied selected-person mask." }, { "id": "action_3", "mentions": [ { "text": "gesturing", "ranges": [ [ 84, 93 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "type": "simple", "subject_refs": [ "person_1" ], "object_refs": [], "related_entity_refs": [ "right_hand_1" ], "confidence": "high", "temporal_status": "supported", "intervals": [ { "clip_relative_s": [ 2.0, 3.0 ], "source_video_s": [ 50.5, 51.5 ] }, { "clip_relative_s": [ 4.0, 5.0 ], "source_video_s": [ 52.5, 53.5 ] }, { "clip_relative_s": [ 6.0, 10.5 ], "source_video_s": [ 54.5, 59.0 ] } ], "evidence": "The masked man repeatedly raises, opens, and moves his free right hand while displaying the bottle; frames where that hand is instead handling the bottle are excluded.", "mask_relation": "The action subject matches the supplied selected-person mask." }, { "id": "action_4", "mentions": [ { "text": "speaking", "ranges": [ [ 128, 136 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } }, { "text": "explaining", "ranges": [ [ 138, 148 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "type": "simple", "subject_refs": [ "person_1" ], "object_refs": [ "spray_bottle_1" ], "confidence": "medium", "temporal_status": "supported", "intervals": [ { "clip_relative_s": [ 0.0, 10.5 ], "source_video_s": [ 48.5, 59.0 ] } ], "evidence": "Across the sampled sequence, the masked man addresses the viewer with changing mouth positions while displaying the product and accompanying his speech with gestures; the explanatory content itself is not visually recoverable.", "mask_relation": "The action subject matches the supplied selected-person mask." } ], "notes": [ "The supplied mask consistently identifies the man described by “The man” and “He.”", "No independent masks were supplied for the spray bottle or either hand, so those entity groups have null source-mask links.", "The visible woman is not mentioned in the immutable caption and therefore has no text entity group.", "The two holding expressions describe the same continuing bottle-holding event. “continues” is excluded as an aspectual verb.", "Temporal intervals are sparse support intervals derived from the sampled clean RGB frames and are bounded by the original source interval [48.5, 59.0)." ] }, "normalization_repairs": 0 }, { "source_video_id": "v_-ypKYhjrlXw", "caption_count_category": 10, "clip": { "caption_index": 4, "caption": "The man leans forward and sprays a white substance from the bottle onto the woman's forearm. He then uses his right index finger to spread the product in a line along her arm.", "source_interval_sec": [ 59.0, 83.0 ], "loom_clip_id": "v_-ypKYhjrlXw_1769_2488", "clip_frame_interval_4fps": [ 1769, 2488 ], "sample_count_1fps": 24, "sampled_frames": [ { "frame": "00000", "timestamp_sec": 59.0, "offset_sec": 0, "rgb_path": "assets/cat_10/v_-ypKYhjrlXw_1769_2488/0000s_00000_rgb.jpg", "mask_path": "assets/cat_10/v_-ypKYhjrlXw_1769_2488/0000s_00000_mask.png", "mask_status": "available" }, { "frame": "00028", "timestamp_sec": 60.0, "offset_sec": 1, "rgb_path": "assets/cat_10/v_-ypKYhjrlXw_1769_2488/0001s_00028_rgb.jpg", "mask_path": "assets/cat_10/v_-ypKYhjrlXw_1769_2488/0001s_00028_mask.png", "mask_status": "available" }, { "frame": "00060", "timestamp_sec": 61.0, "offset_sec": 2, "rgb_path": "assets/cat_10/v_-ypKYhjrlXw_1769_2488/0002s_00060_rgb.jpg", "mask_path": "assets/cat_10/v_-ypKYhjrlXw_1769_2488/0002s_00060_mask.png", "mask_status": "available" }, { "frame": "00088", "timestamp_sec": 62.0, "offset_sec": 3, "rgb_path": "assets/cat_10/v_-ypKYhjrlXw_1769_2488/0003s_00088_rgb.jpg", "mask_path": "assets/cat_10/v_-ypKYhjrlXw_1769_2488/0003s_00088_mask.png", "mask_status": "available" }, { "frame": "00120", "timestamp_sec": 63.0, "offset_sec": 4, "rgb_path": "assets/cat_10/v_-ypKYhjrlXw_1769_2488/0004s_00120_rgb.jpg", "mask_path": "assets/cat_10/v_-ypKYhjrlXw_1769_2488/0004s_00120_mask.png", "mask_status": "available" }, { "frame": "00148", "timestamp_sec": 64.0, "offset_sec": 5, "rgb_path": "assets/cat_10/v_-ypKYhjrlXw_1769_2488/0005s_00148_rgb.jpg", "mask_path": "assets/cat_10/v_-ypKYhjrlXw_1769_2488/0005s_00148_mask.png", "mask_status": "available" }, { "frame": "00180", "timestamp_sec": 65.0, "offset_sec": 6, "rgb_path": "assets/cat_10/v_-ypKYhjrlXw_1769_2488/0006s_00180_rgb.jpg", "mask_path": "assets/cat_10/v_-ypKYhjrlXw_1769_2488/0006s_00180_mask.png", "mask_status": "available" }, { "frame": "00208", "timestamp_sec": 66.0, "offset_sec": 7, "rgb_path": "assets/cat_10/v_-ypKYhjrlXw_1769_2488/0007s_00208_rgb.jpg", "mask_path": "assets/cat_10/v_-ypKYhjrlXw_1769_2488/0007s_00208_mask.png", "mask_status": "available" }, { "frame": "00240", "timestamp_sec": 67.0, "offset_sec": 8, "rgb_path": "assets/cat_10/v_-ypKYhjrlXw_1769_2488/0008s_00240_rgb.jpg", "mask_path": "assets/cat_10/v_-ypKYhjrlXw_1769_2488/0008s_00240_mask.png", "mask_status": "available" }, { "frame": "00268", "timestamp_sec": 68.0, "offset_sec": 9, "rgb_path": "assets/cat_10/v_-ypKYhjrlXw_1769_2488/0009s_00268_rgb.jpg", "mask_path": "assets/cat_10/v_-ypKYhjrlXw_1769_2488/0009s_00268_mask.png", "mask_status": "available" }, { "frame": "00300", "timestamp_sec": 69.0, "offset_sec": 10, "rgb_path": "assets/cat_10/v_-ypKYhjrlXw_1769_2488/0010s_00300_rgb.jpg", "mask_path": "assets/cat_10/v_-ypKYhjrlXw_1769_2488/0010s_00300_mask.png", "mask_status": "available" }, { "frame": "00328", "timestamp_sec": 70.0, "offset_sec": 11, "rgb_path": "assets/cat_10/v_-ypKYhjrlXw_1769_2488/0011s_00328_rgb.jpg", "mask_path": "assets/cat_10/v_-ypKYhjrlXw_1769_2488/0011s_00328_mask.png", "mask_status": "available" }, { "frame": "00360", "timestamp_sec": 71.0, "offset_sec": 12, "rgb_path": "assets/cat_10/v_-ypKYhjrlXw_1769_2488/0012s_00360_rgb.jpg", "mask_path": "assets/cat_10/v_-ypKYhjrlXw_1769_2488/0012s_00360_mask.png", "mask_status": "available" }, { "frame": "00388", "timestamp_sec": 72.0, "offset_sec": 13, "rgb_path": "assets/cat_10/v_-ypKYhjrlXw_1769_2488/0013s_00388_rgb.jpg", "mask_path": "assets/cat_10/v_-ypKYhjrlXw_1769_2488/0013s_00388_mask.png", "mask_status": "available" }, { "frame": "00420", "timestamp_sec": 73.0, "offset_sec": 14, "rgb_path": "assets/cat_10/v_-ypKYhjrlXw_1769_2488/0014s_00420_rgb.jpg", "mask_path": "assets/cat_10/v_-ypKYhjrlXw_1769_2488/0014s_00420_mask.png", "mask_status": "available" }, { "frame": "00448", "timestamp_sec": 74.0, "offset_sec": 15, "rgb_path": "assets/cat_10/v_-ypKYhjrlXw_1769_2488/0015s_00448_rgb.jpg", "mask_path": "assets/cat_10/v_-ypKYhjrlXw_1769_2488/0015s_00448_mask.png", "mask_status": "available" }, { "frame": "00480", "timestamp_sec": 75.0, "offset_sec": 16, "rgb_path": "assets/cat_10/v_-ypKYhjrlXw_1769_2488/0016s_00480_rgb.jpg", "mask_path": "assets/cat_10/v_-ypKYhjrlXw_1769_2488/0016s_00480_mask.png", "mask_status": "available" }, { "frame": "00508", "timestamp_sec": 76.0, "offset_sec": 17, "rgb_path": "assets/cat_10/v_-ypKYhjrlXw_1769_2488/0017s_00508_rgb.jpg", "mask_path": "assets/cat_10/v_-ypKYhjrlXw_1769_2488/0017s_00508_mask.png", "mask_status": "available" }, { "frame": "00540", "timestamp_sec": 77.0, "offset_sec": 18, "rgb_path": "assets/cat_10/v_-ypKYhjrlXw_1769_2488/0018s_00540_rgb.jpg", "mask_path": "assets/cat_10/v_-ypKYhjrlXw_1769_2488/0018s_00540_mask.png", "mask_status": "available" }, { "frame": "00568", "timestamp_sec": 78.0, "offset_sec": 19, "rgb_path": "assets/cat_10/v_-ypKYhjrlXw_1769_2488/0019s_00568_rgb.jpg", "mask_path": "assets/cat_10/v_-ypKYhjrlXw_1769_2488/0019s_00568_mask.png", "mask_status": "available" }, { "frame": "00600", "timestamp_sec": 79.0, "offset_sec": 20, "rgb_path": "assets/cat_10/v_-ypKYhjrlXw_1769_2488/0020s_00600_rgb.jpg", "mask_path": "assets/cat_10/v_-ypKYhjrlXw_1769_2488/0020s_00600_mask.png", "mask_status": "available" }, { "frame": "00628", "timestamp_sec": 80.0, "offset_sec": 21, "rgb_path": "assets/cat_10/v_-ypKYhjrlXw_1769_2488/0021s_00628_rgb.jpg", "mask_path": "assets/cat_10/v_-ypKYhjrlXw_1769_2488/0021s_00628_mask.png", "mask_status": "available" }, { "frame": "00660", "timestamp_sec": 81.0, "offset_sec": 22, "rgb_path": "assets/cat_10/v_-ypKYhjrlXw_1769_2488/0022s_00660_rgb.jpg", "mask_path": "assets/cat_10/v_-ypKYhjrlXw_1769_2488/0022s_00660_mask.png", "mask_status": "available" }, { "frame": "00688", "timestamp_sec": 82.0, "offset_sec": 23, "rgb_path": "assets/cat_10/v_-ypKYhjrlXw_1769_2488/0023s_00688_rgb.jpg", "mask_path": "assets/cat_10/v_-ypKYhjrlXw_1769_2488/0023s_00688_mask.png", "mask_status": "available" } ], "obj_id": [ 0 ], "anno_id": [ 135 ], "fps_inferred_from_clip_interval": 29.9583 }, "usage": { "input_tokens": 134151, "input_tokens_details": { "cache_write_tokens": 134148, "cached_tokens": 0 }, "output_tokens": 3850, "output_tokens_details": { "reasoning_tokens": 2213 }, "total_tokens": 138001 }, "model": "gpt-5.6-sol", "effort": "medium", "custom_id": "cat10_cap04_v_-ypKYhjrlXw_1769_2488", "annotation": { "caption": "The man leans forward and sprays a white substance from the bottle onto the woman's forearm. He then uses his right index finger to spread the product in a line along her arm.", "entity_groups": [ { "id": "entity_man", "label": "man", "spans": [ { "text": "The man", "range": [ 0, 7 ] }, { "text": "He", "range": [ 93, 95 ] }, { "text": "his", "range": [ 106, 109 ] } ], "source_mask_link": "selected_person" }, { "id": "entity_product", "label": "white substance/product", "spans": [ { "text": "a white substance", "range": [ 33, 50 ] }, { "text": "the product", "range": [ 139, 150 ] } ], "source_mask_link": null }, { "id": "entity_bottle", "label": "bottle", "spans": [ { "text": "the bottle", "range": [ 56, 66 ] } ], "source_mask_link": null }, { "id": "entity_woman", "label": "woman", "spans": [ { "text": "woman's", "range": [ 76, 83 ] }, { "text": "her", "range": [ 167, 170 ] } ], "source_mask_link": null }, { "id": "entity_woman_forearm", "label": "woman's forearm/arm", "spans": [ { "text": "the woman's forearm", "range": [ 72, 91 ] }, { "text": "her arm", "range": [ 167, 174 ] } ], "source_mask_link": null }, { "id": "entity_man_index_finger", "label": "man's right index finger", "spans": [ { "text": "his right index finger", "range": [ 106, 128 ] } ], "source_mask_link": null } ], "actions": [ { "id": "action_lean", "mentions": [ { "text": "leans", "ranges": [ [ 8, 13 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "type": "posture", "subject_refs": [ "entity_man" ], "object_refs": [], "confidence": "high", "temporal_status": "localized_sparse", "intervals": [ { "start_s": 1.0, "end_s": 4.0, "source_start_s": 60.0, "source_end_s": 63.0 } ], "evidence": "At local samples 1–3, the masked man bends his upper body toward the woman's extended arm while positioning the bottle.", "mask_relation": "subject matches the supplied selected-person mask", "source_subject_refs": "entity_man" }, { "id": "action_spray", "mentions": [ { "text": "sprays", "ranges": [ [ 26, 32 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "type": "simple", "subject_refs": [ "entity_man" ], "object_refs": [ "entity_product", "entity_bottle", "entity_woman_forearm" ], "confidence": "high", "temporal_status": "localized_sparse", "intervals": [ { "start_s": 1.0, "end_s": 4.0, "source_start_s": 60.0, "source_end_s": 63.0 } ], "evidence": "At local samples 1–3, the masked man aims the bottle nozzle at the forearm and depresses the dispenser.", "mask_relation": "subject matches the supplied selected-person mask", "source_subject_refs": "entity_man" }, { "id": "action_use_finger", "mentions": [ { "text": "uses", "ranges": [ [ 101, 105 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "type": "simple", "subject_refs": [ "entity_man" ], "object_refs": [ "entity_man_index_finger" ], "confidence": "high", "temporal_status": "localized_sparse", "intervals": [ { "start_s": 4.0, "end_s": 12.0, "source_start_s": 63.0, "source_end_s": 71.0 } ], "evidence": "At local samples 4–11, the masked man's right index finger is extended to and repeatedly contacts the woman's forearm.", "mask_relation": "subject matches the supplied selected-person mask", "source_subject_refs": "entity_man" }, { "id": "action_spread", "mentions": [ { "text": "spread", "ranges": [ [ 132, 138 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "type": "simple", "subject_refs": [ "entity_man" ], "object_refs": [ "entity_product", "entity_woman_forearm" ], "confidence": "high", "temporal_status": "localized_sparse", "intervals": [ { "start_s": 4.0, "end_s": 12.0, "source_start_s": 63.0, "source_end_s": 71.0 } ], "evidence": "Across local samples 4–11, the masked man places his index finger on the forearm and moves the contact point along it; his hand is withdrawn by sample 12.", "mask_relation": "subject matches the supplied selected-person mask", "source_subject_refs": "entity_man" } ], "notes": [ "The supplied mask consistently identifies the man, including his moving arms and hands; it does not provide separate instance masks for his finger or held bottle.", "No source masks are supplied for the woman, her forearm, the bottle, or the applied substance.", "The product references \"a white substance\" and \"the product\" are grouped as the same material, and \"the woman's forearm\" and \"her arm\" are grouped as the same body-part instance.", "\"a line\" is treated as the product's spatial arrangement rather than as a separate tangible entity." ] }, "model_output_raw": { "caption": "The man leans forward and sprays a white substance from the bottle onto the woman's forearm. He then uses his right index finger to spread the product in a line along her arm.", "entity_groups": [ { "id": "entity_man", "label": "man", "spans": [ { "text": "The man", "start": 0, "end": 7 }, { "text": "He", "start": 93, "end": 95 }, { "text": "his", "start": 106, "end": 109 } ], "source_mask": "supplied_selected_person_mask" }, { "id": "entity_product", "label": "white substance/product", "spans": [ { "text": "a white substance", "start": 33, "end": 50 }, { "text": "the product", "start": 139, "end": 150 } ], "source_mask": null }, { "id": "entity_bottle", "label": "bottle", "spans": [ { "text": "the bottle", "start": 56, "end": 66 } ], "source_mask": null }, { "id": "entity_woman", "label": "woman", "spans": [ { "text": "woman's", "start": 76, "end": 83 }, { "text": "her", "start": 167, "end": 170 } ], "source_mask": null }, { "id": "entity_woman_forearm", "label": "woman's forearm/arm", "spans": [ { "text": "the woman's forearm", "start": 72, "end": 91 }, { "text": "her arm", "start": 167, "end": 174 } ], "source_mask": null }, { "id": "entity_man_index_finger", "label": "man's right index finger", "spans": [ { "text": "his right index finger", "start": 106, "end": 128 } ], "source_mask": null } ], "actions": [ { "id": "action_lean", "mentions": [ { "text": "leans", "ranges": [ [ 8, 13 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "type": "posture", "subject_ref": "entity_man", "object_refs": [], "confidence": "high", "temporal_status": "localized_sparse", "intervals": [ { "clip_relative_s": [ 1.0, 4.0 ], "source_video_s": [ 60.0, 63.0 ] } ], "evidence": "At local samples 1–3, the masked man bends his upper body toward the woman's extended arm while positioning the bottle.", "mask_relation": "subject matches the supplied selected-person mask" }, { "id": "action_spray", "mentions": [ { "text": "sprays", "ranges": [ [ 26, 32 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "type": "simple", "subject_ref": "entity_man", "object_refs": [ "entity_product", "entity_bottle", "entity_woman_forearm" ], "confidence": "high", "temporal_status": "localized_sparse", "intervals": [ { "clip_relative_s": [ 1.0, 4.0 ], "source_video_s": [ 60.0, 63.0 ] } ], "evidence": "At local samples 1–3, the masked man aims the bottle nozzle at the forearm and depresses the dispenser.", "mask_relation": "subject matches the supplied selected-person mask" }, { "id": "action_use_finger", "mentions": [ { "text": "uses", "ranges": [ [ 101, 105 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "type": "simple", "subject_ref": "entity_man", "object_refs": [ "entity_man_index_finger" ], "confidence": "high", "temporal_status": "localized_sparse", "intervals": [ { "clip_relative_s": [ 4.0, 12.0 ], "source_video_s": [ 63.0, 71.0 ] } ], "evidence": "At local samples 4–11, the masked man's right index finger is extended to and repeatedly contacts the woman's forearm.", "mask_relation": "subject matches the supplied selected-person mask" }, { "id": "action_spread", "mentions": [ { "text": "spread", "ranges": [ [ 132, 138 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "type": "simple", "subject_ref": "entity_man", "object_refs": [ "entity_product", "entity_woman_forearm" ], "confidence": "high", "temporal_status": "localized_sparse", "intervals": [ { "clip_relative_s": [ 4.0, 12.0 ], "source_video_s": [ 63.0, 71.0 ] } ], "evidence": "Across local samples 4–11, the masked man places his index finger on the forearm and moves the contact point along it; his hand is withdrawn by sample 12.", "mask_relation": "subject matches the supplied selected-person mask" } ], "notes": [ "The supplied mask consistently identifies the man, including his moving arms and hands; it does not provide separate instance masks for his finger or held bottle.", "No source masks are supplied for the woman, her forearm, the bottle, or the applied substance.", "The product references \"a white substance\" and \"the product\" are grouped as the same material, and \"the woman's forearm\" and \"her arm\" are grouped as the same body-part instance.", "\"a line\" is treated as the product's spatial arrangement rather than as a separate tangible entity." ] }, "normalization_repairs": 0 }, { "source_video_id": "v_-ypKYhjrlXw", "caption_count_category": 10, "clip": { "caption_index": 5, "caption": "He picks up a small white pump bottle and dispenses a dot of white cream onto the woman's arm. He uses his index finger to spread the cream over a small area.", "source_interval_sec": [ 83.0, 96.0 ], "loom_clip_id": "v_-ypKYhjrlXw_2488_2878", "clip_frame_interval_4fps": [ 2488, 2878 ], "sample_count_1fps": 13, "sampled_frames": [ { "frame": "00000", "timestamp_sec": 83.0, "offset_sec": 0, "rgb_path": "assets/cat_10/v_-ypKYhjrlXw_2488_2878/0000s_00000_rgb.jpg", "mask_path": "assets/cat_10/v_-ypKYhjrlXw_2488_2878/0000s_00000_mask.png", "mask_status": "available" }, { "frame": "00028", "timestamp_sec": 84.0, "offset_sec": 1, "rgb_path": "assets/cat_10/v_-ypKYhjrlXw_2488_2878/0001s_00028_rgb.jpg", "mask_path": "assets/cat_10/v_-ypKYhjrlXw_2488_2878/0001s_00028_mask.png", "mask_status": "available" }, { "frame": "00060", "timestamp_sec": 85.0, "offset_sec": 2, "rgb_path": "assets/cat_10/v_-ypKYhjrlXw_2488_2878/0002s_00060_rgb.jpg", "mask_path": "assets/cat_10/v_-ypKYhjrlXw_2488_2878/0002s_00060_mask.png", "mask_status": "available" }, { "frame": "00088", "timestamp_sec": 86.0, "offset_sec": 3, "rgb_path": "assets/cat_10/v_-ypKYhjrlXw_2488_2878/0003s_00088_rgb.jpg", "mask_path": "assets/cat_10/v_-ypKYhjrlXw_2488_2878/0003s_00088_mask.png", "mask_status": "available" }, { "frame": "00120", "timestamp_sec": 87.0, "offset_sec": 4, "rgb_path": "assets/cat_10/v_-ypKYhjrlXw_2488_2878/0004s_00120_rgb.jpg", "mask_path": "assets/cat_10/v_-ypKYhjrlXw_2488_2878/0004s_00120_mask.png", "mask_status": "available" }, { "frame": "00148", "timestamp_sec": 88.0, "offset_sec": 5, "rgb_path": "assets/cat_10/v_-ypKYhjrlXw_2488_2878/0005s_00148_rgb.jpg", "mask_path": "assets/cat_10/v_-ypKYhjrlXw_2488_2878/0005s_00148_mask.png", "mask_status": "available" }, { "frame": "00180", "timestamp_sec": 89.0, "offset_sec": 6, "rgb_path": "assets/cat_10/v_-ypKYhjrlXw_2488_2878/0006s_00180_rgb.jpg", "mask_path": "assets/cat_10/v_-ypKYhjrlXw_2488_2878/0006s_00180_mask.png", "mask_status": "available" }, { "frame": "00208", "timestamp_sec": 90.0, "offset_sec": 7, "rgb_path": "assets/cat_10/v_-ypKYhjrlXw_2488_2878/0007s_00208_rgb.jpg", "mask_path": "assets/cat_10/v_-ypKYhjrlXw_2488_2878/0007s_00208_mask.png", "mask_status": "available" }, { "frame": "00240", "timestamp_sec": 91.0, "offset_sec": 8, "rgb_path": "assets/cat_10/v_-ypKYhjrlXw_2488_2878/0008s_00240_rgb.jpg", "mask_path": "assets/cat_10/v_-ypKYhjrlXw_2488_2878/0008s_00240_mask.png", "mask_status": "available" }, { "frame": "00268", "timestamp_sec": 92.0, "offset_sec": 9, "rgb_path": "assets/cat_10/v_-ypKYhjrlXw_2488_2878/0009s_00268_rgb.jpg", "mask_path": "assets/cat_10/v_-ypKYhjrlXw_2488_2878/0009s_00268_mask.png", "mask_status": "available" }, { "frame": "00300", "timestamp_sec": 93.0, "offset_sec": 10, "rgb_path": "assets/cat_10/v_-ypKYhjrlXw_2488_2878/0010s_00300_rgb.jpg", "mask_path": "assets/cat_10/v_-ypKYhjrlXw_2488_2878/0010s_00300_mask.png", "mask_status": "available" }, { "frame": "00328", "timestamp_sec": 94.0, "offset_sec": 11, "rgb_path": "assets/cat_10/v_-ypKYhjrlXw_2488_2878/0011s_00328_rgb.jpg", "mask_path": "assets/cat_10/v_-ypKYhjrlXw_2488_2878/0011s_00328_mask.png", "mask_status": "available" }, { "frame": "00360", "timestamp_sec": 95.0, "offset_sec": 12, "rgb_path": "assets/cat_10/v_-ypKYhjrlXw_2488_2878/0012s_00360_rgb.jpg", "mask_path": "assets/cat_10/v_-ypKYhjrlXw_2488_2878/0012s_00360_mask.png", "mask_status": "available" } ], "obj_id": [ 0 ], "anno_id": [ 137 ], "fps_inferred_from_clip_interval": 30.0 }, "usage": { "input_tokens": 79688, "input_tokens_details": { "cache_write_tokens": 79685, "cached_tokens": 0 }, "output_tokens": 3493, "output_tokens_details": { "reasoning_tokens": 2070 }, "total_tokens": 83181 }, "model": "gpt-5.6-sol", "effort": "medium", "custom_id": "cat10_cap05_v_-ypKYhjrlXw_2488_2878", "annotation": { "caption": "He picks up a small white pump bottle and dispenses a dot of white cream onto the woman's arm. He uses his index finger to spread the cream over a small area.", "entity_groups": [ { "id": "person_man", "label": "man", "spans": [ { "text": "He", "range": [ 0, 2 ] }, { "text": "He", "range": [ 95, 97 ] }, { "text": "his", "range": [ 103, 106 ] } ], "source_mask_link": "selected_person" }, { "id": "pump_bottle", "label": "pump bottle", "spans": [ { "text": "a small white pump bottle", "range": [ 12, 37 ] } ], "source_mask_link": null }, { "id": "cream", "label": "cream", "spans": [ { "text": "a dot of white cream", "range": [ 52, 72 ] }, { "text": "the cream", "range": [ 130, 139 ] } ], "source_mask_link": null }, { "id": "woman", "label": "woman", "spans": [ { "text": "the woman's", "range": [ 78, 89 ] } ], "source_mask_link": null }, { "id": "woman_arm", "label": "arm", "spans": [ { "text": "the woman's arm", "range": [ 78, 93 ] } ], "source_mask_link": null }, { "id": "man_index_finger", "label": "index finger", "spans": [ { "text": "his index finger", "range": [ 103, 119 ] } ], "source_mask_link": null } ], "actions": [ { "id": "action_pick_up", "mentions": [ { "text": "picks up", "ranges": [ [ 3, 11 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "type": "phrasal", "subject_refs": [ "person_man" ], "object_refs": [ "pump_bottle" ], "confidence": "high", "temporal_status": "not_observed_in_sampled_frames", "intervals": [], "evidence": "At 0.000 seconds the masked man already holds the bottle; no sampled frame shows the pickup transition.", "mask_relation": "subject matches the supplied selected-person mask", "source_subject_refs": "person_man" }, { "id": "action_dispense", "mentions": [ { "text": "dispenses", "ranges": [ [ 42, 51 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "type": "simple", "subject_refs": [ "person_man" ], "object_refs": [ "cream", "woman_arm" ], "confidence": "high", "temporal_status": "localized", "intervals": [ { "start_s": 0.0, "end_s": 1.0, "source_start_s": 83.0, "source_end_s": 84.0 } ], "evidence": "At 0.000 seconds the masked man presses the pump over the woman's arm; by 1.000 seconds a white dot is visible on the arm.", "mask_relation": "subject matches the supplied selected-person mask", "source_subject_refs": "person_man" }, { "id": "action_use_finger", "mentions": [ { "text": "uses", "ranges": [ [ 98, 102 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "type": "simple", "subject_refs": [ "person_man" ], "object_refs": [ "man_index_finger" ], "confidence": "high", "temporal_status": "localized", "intervals": [ { "start_s": 2.0, "end_s": 10.0, "source_start_s": 85.0, "source_end_s": 93.0 } ], "evidence": "From 2.000 through 9.000 seconds, the masked man's index finger contacts and moves over the cream-covered part of the arm; it is lifted by 10.000 seconds.", "mask_relation": "subject matches the supplied selected-person mask", "source_subject_refs": "person_man" }, { "id": "action_spread", "mentions": [ { "text": "spread", "ranges": [ [ 123, 129 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "type": "simple", "subject_refs": [ "person_man" ], "object_refs": [ "cream", "woman_arm" ], "confidence": "high", "temporal_status": "localized", "intervals": [ { "start_s": 2.0, "end_s": 10.0, "source_start_s": 85.0, "source_end_s": 93.0 } ], "evidence": "Frames from 2.000 through 9.000 seconds show the masked man moving his fingertip across the cream on the woman's arm; contact has ended by 10.000 seconds.", "mask_relation": "subject matches the supplied selected-person mask", "source_subject_refs": "person_man" } ], "notes": [ "The supplied mask consistently identifies the man described by both occurrences of \"He\" and by \"his\".", "No separate source masks were supplied for the woman, bottle, cream, arm, or index finger, so their source-mask links are null.", "\"a small area\" is excluded as an abstract spatial region.", "Temporal intervals are sparse support intervals based on the sampled clean RGB frames, not exact continuous-video boundaries." ] }, "model_output_raw": { "caption": "He picks up a small white pump bottle and dispenses a dot of white cream onto the woman's arm. He uses his index finger to spread the cream over a small area.", "entity_groups": [ { "id": "person_man", "label": "man", "spans": [ { "text": "He", "range": [ 0, 2 ] }, { "text": "He", "range": [ 95, 97 ] }, { "text": "his", "range": [ 103, 106 ] } ], "source_mask": "selected_person_mask" }, { "id": "pump_bottle", "label": "pump bottle", "spans": [ { "text": "a small white pump bottle", "range": [ 12, 37 ] } ], "source_mask": null }, { "id": "cream", "label": "cream", "spans": [ { "text": "a dot of white cream", "range": [ 52, 72 ] }, { "text": "the cream", "range": [ 130, 139 ] } ], "source_mask": null }, { "id": "woman", "label": "woman", "spans": [ { "text": "the woman's", "range": [ 78, 89 ] } ], "source_mask": null }, { "id": "woman_arm", "label": "arm", "spans": [ { "text": "the woman's arm", "range": [ 78, 93 ] } ], "source_mask": null }, { "id": "man_index_finger", "label": "index finger", "spans": [ { "text": "his index finger", "range": [ 103, 119 ] } ], "source_mask": null } ], "actions": [ { "id": "action_pick_up", "mentions": [ { "text": "picks up", "ranges": [ [ 3, 11 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "type": "phrasal", "subject_ref": "person_man", "object_refs": [ "pump_bottle" ], "confidence": "high", "mask_relation": "subject matches the supplied selected-person mask", "temporal_status": "not_observed_in_sampled_frames", "intervals": [], "evidence": "At 0.000 seconds the masked man already holds the bottle; no sampled frame shows the pickup transition." }, { "id": "action_dispense", "mentions": [ { "text": "dispenses", "ranges": [ [ 42, 51 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "type": "simple", "subject_ref": "person_man", "object_refs": [ "cream", "woman_arm" ], "confidence": "high", "mask_relation": "subject matches the supplied selected-person mask", "temporal_status": "localized", "intervals": [ { "clip_relative_s": [ 0.0, 1.0 ], "source_video_s": [ 83.0, 84.0 ] } ], "evidence": "At 0.000 seconds the masked man presses the pump over the woman's arm; by 1.000 seconds a white dot is visible on the arm." }, { "id": "action_use_finger", "mentions": [ { "text": "uses", "ranges": [ [ 98, 102 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "type": "simple", "subject_ref": "person_man", "object_refs": [ "man_index_finger" ], "confidence": "high", "mask_relation": "subject matches the supplied selected-person mask", "temporal_status": "localized", "intervals": [ { "clip_relative_s": [ 2.0, 10.0 ], "source_video_s": [ 85.0, 93.0 ] } ], "evidence": "From 2.000 through 9.000 seconds, the masked man's index finger contacts and moves over the cream-covered part of the arm; it is lifted by 10.000 seconds." }, { "id": "action_spread", "mentions": [ { "text": "spread", "ranges": [ [ 123, 129 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "type": "simple", "subject_ref": "person_man", "object_refs": [ "cream", "woman_arm" ], "confidence": "high", "mask_relation": "subject matches the supplied selected-person mask", "temporal_status": "localized", "intervals": [ { "clip_relative_s": [ 2.0, 10.0 ], "source_video_s": [ 85.0, 93.0 ] } ], "evidence": "Frames from 2.000 through 9.000 seconds show the masked man moving his fingertip across the cream on the woman's arm; contact has ended by 10.000 seconds." } ], "notes": [ "The supplied mask consistently identifies the man described by both occurrences of \"He\" and by \"his\".", "No separate source masks were supplied for the woman, bottle, cream, arm, or index finger, so their source-mask links are null.", "\"a small area\" is excluded as an abstract spatial region.", "Temporal intervals are sparse support intervals based on the sampled clean RGB frames, not exact continuous-video boundaries." ] }, "normalization_repairs": 0 }, { "source_video_id": "v_-ypKYhjrlXw", "caption_count_category": 10, "clip": { "caption_index": 6, "caption": "The man picks up a white tissue and wipes the products off the woman's arm. He then picks up a small white tube with a blue tip and holds it while talking.", "source_interval_sec": [ 96.0, 113.0 ], "loom_clip_id": "v_-ypKYhjrlXw_2878_3387", "clip_frame_interval_4fps": [ 2878, 3387 ], "sample_count_1fps": 17, "sampled_frames": [ { "frame": "00000", "timestamp_sec": 96.0, "offset_sec": 0, "rgb_path": "assets/cat_10/v_-ypKYhjrlXw_2878_3387/0000s_00000_rgb.jpg", "mask_path": "assets/cat_10/v_-ypKYhjrlXw_2878_3387/0000s_00000_mask.png", "mask_status": "available" }, { "frame": "00028", "timestamp_sec": 97.0, "offset_sec": 1, "rgb_path": "assets/cat_10/v_-ypKYhjrlXw_2878_3387/0001s_00028_rgb.jpg", "mask_path": "assets/cat_10/v_-ypKYhjrlXw_2878_3387/0001s_00028_mask.png", "mask_status": "available" }, { "frame": "00060", "timestamp_sec": 98.0, "offset_sec": 2, "rgb_path": "assets/cat_10/v_-ypKYhjrlXw_2878_3387/0002s_00060_rgb.jpg", "mask_path": "assets/cat_10/v_-ypKYhjrlXw_2878_3387/0002s_00060_mask.png", "mask_status": "available" }, { "frame": "00088", "timestamp_sec": 99.0, "offset_sec": 3, "rgb_path": "assets/cat_10/v_-ypKYhjrlXw_2878_3387/0003s_00088_rgb.jpg", "mask_path": "assets/cat_10/v_-ypKYhjrlXw_2878_3387/0003s_00088_mask.png", "mask_status": "available" }, { "frame": "00120", "timestamp_sec": 100.0, "offset_sec": 4, "rgb_path": "assets/cat_10/v_-ypKYhjrlXw_2878_3387/0004s_00120_rgb.jpg", "mask_path": "assets/cat_10/v_-ypKYhjrlXw_2878_3387/0004s_00120_mask.png", "mask_status": "available" }, { "frame": "00148", "timestamp_sec": 101.0, "offset_sec": 5, "rgb_path": "assets/cat_10/v_-ypKYhjrlXw_2878_3387/0005s_00148_rgb.jpg", "mask_path": "assets/cat_10/v_-ypKYhjrlXw_2878_3387/0005s_00148_mask.png", "mask_status": "available" }, { "frame": "00180", "timestamp_sec": 102.0, "offset_sec": 6, "rgb_path": "assets/cat_10/v_-ypKYhjrlXw_2878_3387/0006s_00180_rgb.jpg", "mask_path": "assets/cat_10/v_-ypKYhjrlXw_2878_3387/0006s_00180_mask.png", "mask_status": "available" }, { "frame": "00208", "timestamp_sec": 103.0, "offset_sec": 7, "rgb_path": "assets/cat_10/v_-ypKYhjrlXw_2878_3387/0007s_00208_rgb.jpg", "mask_path": "assets/cat_10/v_-ypKYhjrlXw_2878_3387/0007s_00208_mask.png", "mask_status": "available" }, { "frame": "00240", "timestamp_sec": 104.0, "offset_sec": 8, "rgb_path": "assets/cat_10/v_-ypKYhjrlXw_2878_3387/0008s_00240_rgb.jpg", "mask_path": "assets/cat_10/v_-ypKYhjrlXw_2878_3387/0008s_00240_mask.png", "mask_status": "available" }, { "frame": "00268", "timestamp_sec": 105.0, "offset_sec": 9, "rgb_path": "assets/cat_10/v_-ypKYhjrlXw_2878_3387/0009s_00268_rgb.jpg", "mask_path": "assets/cat_10/v_-ypKYhjrlXw_2878_3387/0009s_00268_mask.png", "mask_status": "available" }, { "frame": "00300", "timestamp_sec": 106.0, "offset_sec": 10, "rgb_path": "assets/cat_10/v_-ypKYhjrlXw_2878_3387/0010s_00300_rgb.jpg", "mask_path": "assets/cat_10/v_-ypKYhjrlXw_2878_3387/0010s_00300_mask.png", "mask_status": "available" }, { "frame": "00328", "timestamp_sec": 107.0, "offset_sec": 11, "rgb_path": "assets/cat_10/v_-ypKYhjrlXw_2878_3387/0011s_00328_rgb.jpg", "mask_path": "assets/cat_10/v_-ypKYhjrlXw_2878_3387/0011s_00328_mask.png", "mask_status": "available" }, { "frame": "00360", "timestamp_sec": 108.0, "offset_sec": 12, "rgb_path": "assets/cat_10/v_-ypKYhjrlXw_2878_3387/0012s_00360_rgb.jpg", "mask_path": "assets/cat_10/v_-ypKYhjrlXw_2878_3387/0012s_00360_mask.png", "mask_status": "available" }, { "frame": "00388", "timestamp_sec": 109.0, "offset_sec": 13, "rgb_path": "assets/cat_10/v_-ypKYhjrlXw_2878_3387/0013s_00388_rgb.jpg", "mask_path": "assets/cat_10/v_-ypKYhjrlXw_2878_3387/0013s_00388_mask.png", "mask_status": "available" }, { "frame": "00420", "timestamp_sec": 110.0, "offset_sec": 14, "rgb_path": "assets/cat_10/v_-ypKYhjrlXw_2878_3387/0014s_00420_rgb.jpg", "mask_path": "assets/cat_10/v_-ypKYhjrlXw_2878_3387/0014s_00420_mask.png", "mask_status": "available" }, { "frame": "00448", "timestamp_sec": 111.0, "offset_sec": 15, "rgb_path": "assets/cat_10/v_-ypKYhjrlXw_2878_3387/0015s_00448_rgb.jpg", "mask_path": "assets/cat_10/v_-ypKYhjrlXw_2878_3387/0015s_00448_mask.png", "mask_status": "available" }, { "frame": "00480", "timestamp_sec": 112.0, "offset_sec": 16, "rgb_path": "assets/cat_10/v_-ypKYhjrlXw_2878_3387/0016s_00480_rgb.jpg", "mask_path": "assets/cat_10/v_-ypKYhjrlXw_2878_3387/0016s_00480_mask.png", "mask_status": "available" } ], "obj_id": [ 0 ], "anno_id": [ 138 ], "fps_inferred_from_clip_interval": 29.9412 }, "usage": { "input_tokens": 99491, "input_tokens_details": { "cache_write_tokens": 99488, "cached_tokens": 0 }, "output_tokens": 3972, "output_tokens_details": { "reasoning_tokens": 2070 }, "total_tokens": 103463 }, "model": "gpt-5.6-sol", "effort": "medium", "custom_id": "cat10_cap06_v_-ypKYhjrlXw_2878_3387", "annotation": { "caption": "The man picks up a white tissue and wipes the products off the woman's arm. He then picks up a small white tube with a blue tip and holds it while talking.", "entity_groups": [ { "id": "entity_1", "label": "man", "spans": [ { "text": "The man", "range": [ 0, 7 ] }, { "text": "He", "range": [ 76, 78 ] } ], "source_mask_link": "selected_person" }, { "id": "entity_2", "label": "tissue", "spans": [ { "text": "a white tissue", "range": [ 17, 31 ] } ], "source_mask_link": null }, { "id": "entity_3", "label": "products", "spans": [ { "text": "the products", "range": [ 42, 54 ] } ], "source_mask_link": null }, { "id": "entity_4", "label": "woman", "spans": [ { "text": "the woman's", "range": [ 59, 70 ] } ], "source_mask_link": null }, { "id": "entity_5", "label": "arm", "spans": [ { "text": "the woman's arm", "range": [ 59, 74 ] } ], "source_mask_link": null }, { "id": "entity_6", "label": "tube", "spans": [ { "text": "a small white tube", "range": [ 93, 111 ] }, { "text": "it", "range": [ 138, 140 ] } ], "source_mask_link": null }, { "id": "entity_7", "label": "tip", "spans": [ { "text": "a blue tip", "range": [ 117, 127 ] } ], "source_mask_link": null } ], "actions": [ { "id": "action_1", "mentions": [ { "text": "picks up", "ranges": [ [ 8, 16 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "type": "phrasal", "subject_refs": [ "entity_1" ], "object_refs": [ "entity_2" ], "confidence": "high", "temporal_status": "localized_sparse", "intervals": [ { "start_s": 0.0, "end_s": 2.0, "source_start_s": 96.0, "source_end_s": 98.0 } ], "evidence": "At 0 seconds the masked man reaches down, and at 1 second he has raised the white tissue in his hand.", "mask_relation": "subject matches the supplied selected-person mask", "source_subject_refs": "entity_1" }, { "id": "action_2", "mentions": [ { "text": "wipes", "ranges": [ [ 36, 41 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "type": "prepositional", "subject_refs": [ "entity_1" ], "object_refs": [ "entity_3" ], "confidence": "high", "temporal_status": "localized_sparse", "intervals": [ { "start_s": 2.0, "end_s": 4.0, "source_start_s": 98.0, "source_end_s": 100.0 } ], "evidence": "At 2 and 3 seconds the masked man presses and moves the white tissue against the woman's arm.", "mask_relation": "subject matches the supplied selected-person mask", "source_subject_refs": "entity_1" }, { "id": "action_3", "mentions": [ { "text": "picks up", "ranges": [ [ 84, 92 ] ], "occurrence": 2, "particle": { "text": null, "occurrence": 1 } } ], "type": "phrasal", "subject_refs": [ "entity_1" ], "object_refs": [ "entity_6" ], "confidence": "medium", "temporal_status": "localized_sparse", "intervals": [ { "start_s": 3.0, "end_s": 5.0, "source_start_s": 99.0, "source_end_s": 101.0 } ], "evidence": "The sampled transition shows the masked man changing from the tissue task to visibly holding the white tube by 4 seconds; the exact pickup occurs between sparse samples.", "mask_relation": "subject matches the supplied selected-person mask", "source_subject_refs": "entity_1" }, { "id": "action_4", "mentions": [ { "text": "holds", "ranges": [ [ 132, 137 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "type": "simple", "subject_refs": [ "entity_1" ], "object_refs": [ "entity_6" ], "confidence": "high", "temporal_status": "localized_sparse", "intervals": [ { "start_s": 4.0, "end_s": 13.0, "source_start_s": 100.0, "source_end_s": 109.0 } ], "evidence": "From 4 through 12 seconds, the masked man visibly keeps the small white tube in his left hand.", "mask_relation": "subject matches the supplied selected-person mask", "source_subject_refs": "entity_1" }, { "id": "action_5", "mentions": [ { "text": "talking", "ranges": [ [ 147, 154 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "type": "simple", "subject_refs": [ "entity_1" ], "object_refs": [], "confidence": "medium", "temporal_status": "localized_sparse", "intervals": [ { "start_s": 0.0, "end_s": 1.0, "source_start_s": 96.0, "source_end_s": 97.0 }, { "start_s": 15.0, "end_s": 17.0, "source_start_s": 111.0, "source_end_s": 113.0 } ], "evidence": "The masked man's mouth is visibly open in the 0-, 15-, and 16-second samples. His mouth is cropped out in many intervening close-up frames, so only directly supported samples are included.", "mask_relation": "subject matches the supplied selected-person mask", "source_subject_refs": "entity_1" } ], "notes": [ "The supplied masklet consistently identifies the man, who is the subject of every annotated action.", "The woman, her arm, the tissue, products, tube, and tube tip have no supplied instance masks; their source-mask links are therefore null.", "The woman is nevertheless visible at the right side of early and late RGB frames, and her arm is visible throughout much of the clip.", "In the wiping expression, “off” introduces the removal/source relation to the arm and is excluded from the action span as an ordinary preposition.", "Pickup timing is based on sparse frame-to-frame transitions rather than an exact continuous-video onset." ] }, "model_output_raw": { "caption": "The man picks up a white tissue and wipes the products off the woman's arm. He then picks up a small white tube with a blue tip and holds it while talking.", "entity_groups": [ { "id": "entity_1", "label": "man", "spans": [ { "text": "The man", "start": 0, "end": 7 }, { "text": "He", "start": 76, "end": 78 } ], "source_mask": "selected_person_masklet" }, { "id": "entity_2", "label": "tissue", "spans": [ { "text": "a white tissue", "start": 17, "end": 31 } ], "source_mask": null }, { "id": "entity_3", "label": "products", "spans": [ { "text": "the products", "start": 42, "end": 54 } ], "source_mask": null }, { "id": "entity_4", "label": "woman", "spans": [ { "text": "the woman's", "start": 59, "end": 70 } ], "source_mask": null }, { "id": "entity_5", "label": "arm", "spans": [ { "text": "the woman's arm", "start": 59, "end": 74 } ], "source_mask": null }, { "id": "entity_6", "label": "tube", "spans": [ { "text": "a small white tube", "start": 93, "end": 111 }, { "text": "it", "start": 138, "end": 140 } ], "source_mask": null }, { "id": "entity_7", "label": "tip", "spans": [ { "text": "a blue tip", "start": 117, "end": 127 } ], "source_mask": null } ], "actions": [ { "id": "action_1", "mentions": [ { "text": "picks up", "occurrence": 1, "ranges": [ { "start": 8, "end": 16 } ], "particle": { "text": null, "occurrence": 1 } } ], "type": "phrasal", "subject": "entity_1", "object": "entity_2", "confidence": "high", "mask_relation": "subject matches the supplied selected-person mask", "temporal_status": "localized_sparse", "intervals": [ { "clip_relative_s": [ 0.0, 2.0 ], "source_video_s": [ 96.0, 98.0 ] } ], "evidence": "At 0 seconds the masked man reaches down, and at 1 second he has raised the white tissue in his hand." }, { "id": "action_2", "mentions": [ { "text": "wipes", "occurrence": 1, "ranges": [ { "start": 36, "end": 41 } ], "particle": { "text": null, "occurrence": 1 } } ], "type": "prepositional", "subject": "entity_1", "object": "entity_3", "contact_target": "entity_5", "confidence": "high", "mask_relation": "subject matches the supplied selected-person mask", "temporal_status": "localized_sparse", "intervals": [ { "clip_relative_s": [ 2.0, 4.0 ], "source_video_s": [ 98.0, 100.0 ] } ], "evidence": "At 2 and 3 seconds the masked man presses and moves the white tissue against the woman's arm." }, { "id": "action_3", "mentions": [ { "text": "picks up", "occurrence": 2, "ranges": [ { "start": 84, "end": 92 } ], "particle": { "text": null, "occurrence": 1 } } ], "type": "phrasal", "subject": "entity_1", "object": "entity_6", "confidence": "medium", "mask_relation": "subject matches the supplied selected-person mask", "temporal_status": "localized_sparse", "intervals": [ { "clip_relative_s": [ 3.0, 5.0 ], "source_video_s": [ 99.0, 101.0 ] } ], "evidence": "The sampled transition shows the masked man changing from the tissue task to visibly holding the white tube by 4 seconds; the exact pickup occurs between sparse samples." }, { "id": "action_4", "mentions": [ { "text": "holds", "occurrence": 1, "ranges": [ { "start": 132, "end": 137 } ], "particle": { "text": null, "occurrence": 1 } } ], "type": "simple", "subject": "entity_1", "object": "entity_6", "confidence": "high", "mask_relation": "subject matches the supplied selected-person mask", "temporal_status": "localized_sparse", "intervals": [ { "clip_relative_s": [ 4.0, 13.0 ], "source_video_s": [ 100.0, 109.0 ] } ], "evidence": "From 4 through 12 seconds, the masked man visibly keeps the small white tube in his left hand." }, { "id": "action_5", "mentions": [ { "text": "talking", "occurrence": 1, "ranges": [ { "start": 147, "end": 154 } ], "particle": { "text": null, "occurrence": 1 } } ], "type": "simple", "subject": "entity_1", "object": null, "confidence": "medium", "mask_relation": "subject matches the supplied selected-person mask", "temporal_status": "localized_sparse", "intervals": [ { "clip_relative_s": [ 0.0, 1.0 ], "source_video_s": [ 96.0, 97.0 ] }, { "clip_relative_s": [ 15.0, 17.0 ], "source_video_s": [ 111.0, 113.0 ] } ], "evidence": "The masked man's mouth is visibly open in the 0-, 15-, and 16-second samples. His mouth is cropped out in many intervening close-up frames, so only directly supported samples are included." } ], "notes": [ "The supplied masklet consistently identifies the man, who is the subject of every annotated action.", "The woman, her arm, the tissue, products, tube, and tube tip have no supplied instance masks; their source-mask links are therefore null.", "The woman is nevertheless visible at the right side of early and late RGB frames, and her arm is visible throughout much of the clip.", "In the wiping expression, “off” introduces the removal/source relation to the arm and is excluded from the action span as an ordinary preposition.", "Pickup timing is based on sparse frame-to-frame transitions rather than an exact continuous-video onset." ] }, "normalization_repairs": 0 }, { "source_video_id": "v_-ypKYhjrlXw", "caption_count_category": 10, "clip": { "caption_index": 7, "caption": "While continuing to speak, the man applies a dot of cream from the blue-tipped tube onto the back of the woman's hand. He then uses his index finger to gently spread the cream across her hand.", "source_interval_sec": [ 113.0, 140.5 ], "loom_clip_id": "v_-ypKYhjrlXw_3387_4211", "clip_frame_interval_4fps": [ 3387, 4211 ], "sample_count_1fps": 28, "sampled_frames": [ { "frame": "00000", "timestamp_sec": 113.0, "offset_sec": 0, "rgb_path": "assets/cat_10/v_-ypKYhjrlXw_3387_4211/0000s_00000_rgb.jpg", "mask_path": "assets/cat_10/v_-ypKYhjrlXw_3387_4211/0000s_00000_mask.png", "mask_status": "available" }, { "frame": "00028", "timestamp_sec": 114.0, "offset_sec": 1, "rgb_path": "assets/cat_10/v_-ypKYhjrlXw_3387_4211/0001s_00028_rgb.jpg", "mask_path": "assets/cat_10/v_-ypKYhjrlXw_3387_4211/0001s_00028_mask.png", "mask_status": "available" }, { "frame": "00060", "timestamp_sec": 115.0, "offset_sec": 2, "rgb_path": "assets/cat_10/v_-ypKYhjrlXw_3387_4211/0002s_00060_rgb.jpg", "mask_path": "assets/cat_10/v_-ypKYhjrlXw_3387_4211/0002s_00060_mask.png", "mask_status": "available" }, { "frame": "00088", "timestamp_sec": 116.0, "offset_sec": 3, "rgb_path": "assets/cat_10/v_-ypKYhjrlXw_3387_4211/0003s_00088_rgb.jpg", "mask_path": "assets/cat_10/v_-ypKYhjrlXw_3387_4211/0003s_00088_mask.png", "mask_status": "available" }, { "frame": "00120", "timestamp_sec": 117.0, "offset_sec": 4, "rgb_path": "assets/cat_10/v_-ypKYhjrlXw_3387_4211/0004s_00120_rgb.jpg", "mask_path": "assets/cat_10/v_-ypKYhjrlXw_3387_4211/0004s_00120_mask.png", "mask_status": "available" }, { "frame": "00148", "timestamp_sec": 118.0, "offset_sec": 5, "rgb_path": "assets/cat_10/v_-ypKYhjrlXw_3387_4211/0005s_00148_rgb.jpg", "mask_path": "assets/cat_10/v_-ypKYhjrlXw_3387_4211/0005s_00148_mask.png", "mask_status": "available" }, { "frame": "00180", "timestamp_sec": 119.0, "offset_sec": 6, "rgb_path": "assets/cat_10/v_-ypKYhjrlXw_3387_4211/0006s_00180_rgb.jpg", "mask_path": "assets/cat_10/v_-ypKYhjrlXw_3387_4211/0006s_00180_mask.png", "mask_status": "available" }, { "frame": "00208", "timestamp_sec": 120.0, "offset_sec": 7, "rgb_path": "assets/cat_10/v_-ypKYhjrlXw_3387_4211/0007s_00208_rgb.jpg", "mask_path": "assets/cat_10/v_-ypKYhjrlXw_3387_4211/0007s_00208_mask.png", "mask_status": "available" }, { "frame": "00240", "timestamp_sec": 121.0, "offset_sec": 8, "rgb_path": "assets/cat_10/v_-ypKYhjrlXw_3387_4211/0008s_00240_rgb.jpg", "mask_path": "assets/cat_10/v_-ypKYhjrlXw_3387_4211/0008s_00240_mask.png", "mask_status": "available" }, { "frame": "00268", "timestamp_sec": 122.0, "offset_sec": 9, "rgb_path": "assets/cat_10/v_-ypKYhjrlXw_3387_4211/0009s_00268_rgb.jpg", "mask_path": "assets/cat_10/v_-ypKYhjrlXw_3387_4211/0009s_00268_mask.png", "mask_status": "available" }, { "frame": "00300", "timestamp_sec": 123.0, "offset_sec": 10, "rgb_path": "assets/cat_10/v_-ypKYhjrlXw_3387_4211/0010s_00300_rgb.jpg", "mask_path": "assets/cat_10/v_-ypKYhjrlXw_3387_4211/0010s_00300_mask.png", "mask_status": "available" }, { "frame": "00328", "timestamp_sec": 124.0, "offset_sec": 11, "rgb_path": "assets/cat_10/v_-ypKYhjrlXw_3387_4211/0011s_00328_rgb.jpg", "mask_path": "assets/cat_10/v_-ypKYhjrlXw_3387_4211/0011s_00328_mask.png", "mask_status": "available" }, { "frame": "00360", "timestamp_sec": 125.0, "offset_sec": 12, "rgb_path": "assets/cat_10/v_-ypKYhjrlXw_3387_4211/0012s_00360_rgb.jpg", "mask_path": "assets/cat_10/v_-ypKYhjrlXw_3387_4211/0012s_00360_mask.png", "mask_status": "available" }, { "frame": "00388", "timestamp_sec": 126.0, "offset_sec": 13, "rgb_path": "assets/cat_10/v_-ypKYhjrlXw_3387_4211/0013s_00388_rgb.jpg", "mask_path": "assets/cat_10/v_-ypKYhjrlXw_3387_4211/0013s_00388_mask.png", "mask_status": "available" }, { "frame": "00420", "timestamp_sec": 127.0, "offset_sec": 14, "rgb_path": "assets/cat_10/v_-ypKYhjrlXw_3387_4211/0014s_00420_rgb.jpg", "mask_path": "assets/cat_10/v_-ypKYhjrlXw_3387_4211/0014s_00420_mask.png", "mask_status": "available" }, { "frame": "00448", "timestamp_sec": 128.0, "offset_sec": 15, "rgb_path": "assets/cat_10/v_-ypKYhjrlXw_3387_4211/0015s_00448_rgb.jpg", "mask_path": "assets/cat_10/v_-ypKYhjrlXw_3387_4211/0015s_00448_mask.png", "mask_status": "available" }, { "frame": "00480", "timestamp_sec": 129.0, "offset_sec": 16, "rgb_path": "assets/cat_10/v_-ypKYhjrlXw_3387_4211/0016s_00480_rgb.jpg", "mask_path": "assets/cat_10/v_-ypKYhjrlXw_3387_4211/0016s_00480_mask.png", "mask_status": "available" }, { "frame": "00508", "timestamp_sec": 130.0, "offset_sec": 17, "rgb_path": "assets/cat_10/v_-ypKYhjrlXw_3387_4211/0017s_00508_rgb.jpg", "mask_path": "assets/cat_10/v_-ypKYhjrlXw_3387_4211/0017s_00508_mask.png", "mask_status": "available" }, { "frame": "00540", "timestamp_sec": 131.0, "offset_sec": 18, "rgb_path": "assets/cat_10/v_-ypKYhjrlXw_3387_4211/0018s_00540_rgb.jpg", "mask_path": "assets/cat_10/v_-ypKYhjrlXw_3387_4211/0018s_00540_mask.png", "mask_status": "available" }, { "frame": "00568", "timestamp_sec": 132.0, "offset_sec": 19, "rgb_path": "assets/cat_10/v_-ypKYhjrlXw_3387_4211/0019s_00568_rgb.jpg", "mask_path": "assets/cat_10/v_-ypKYhjrlXw_3387_4211/0019s_00568_mask.png", "mask_status": "available" }, { "frame": "00600", "timestamp_sec": 133.0, "offset_sec": 20, "rgb_path": "assets/cat_10/v_-ypKYhjrlXw_3387_4211/0020s_00600_rgb.jpg", "mask_path": "assets/cat_10/v_-ypKYhjrlXw_3387_4211/0020s_00600_mask.png", "mask_status": "available" }, { "frame": "00628", "timestamp_sec": 134.0, "offset_sec": 21, "rgb_path": "assets/cat_10/v_-ypKYhjrlXw_3387_4211/0021s_00628_rgb.jpg", "mask_path": "assets/cat_10/v_-ypKYhjrlXw_3387_4211/0021s_00628_mask.png", "mask_status": "available" }, { "frame": "00660", "timestamp_sec": 135.0, "offset_sec": 22, "rgb_path": "assets/cat_10/v_-ypKYhjrlXw_3387_4211/0022s_00660_rgb.jpg", "mask_path": "assets/cat_10/v_-ypKYhjrlXw_3387_4211/0022s_00660_mask.png", "mask_status": "available" }, { "frame": "00688", "timestamp_sec": 136.0, "offset_sec": 23, "rgb_path": "assets/cat_10/v_-ypKYhjrlXw_3387_4211/0023s_00688_rgb.jpg", "mask_path": "assets/cat_10/v_-ypKYhjrlXw_3387_4211/0023s_00688_mask.png", "mask_status": "available" }, { "frame": "00720", "timestamp_sec": 137.0, "offset_sec": 24, "rgb_path": "assets/cat_10/v_-ypKYhjrlXw_3387_4211/0024s_00720_rgb.jpg", "mask_path": "assets/cat_10/v_-ypKYhjrlXw_3387_4211/0024s_00720_mask.png", "mask_status": "available" }, { "frame": "00748", "timestamp_sec": 138.0, "offset_sec": 25, "rgb_path": "assets/cat_10/v_-ypKYhjrlXw_3387_4211/0025s_00748_rgb.jpg", "mask_path": "assets/cat_10/v_-ypKYhjrlXw_3387_4211/0025s_00748_mask.png", "mask_status": "available" }, { "frame": "00780", "timestamp_sec": 139.0, "offset_sec": 26, "rgb_path": "assets/cat_10/v_-ypKYhjrlXw_3387_4211/0026s_00780_rgb.jpg", "mask_path": "assets/cat_10/v_-ypKYhjrlXw_3387_4211/0026s_00780_mask.png", "mask_status": "available" }, { "frame": "00808", "timestamp_sec": 140.0, "offset_sec": 27, "rgb_path": "assets/cat_10/v_-ypKYhjrlXw_3387_4211/0027s_00808_rgb.jpg", "mask_path": "assets/cat_10/v_-ypKYhjrlXw_3387_4211/0027s_00808_mask.png", "mask_status": "available" } ], "obj_id": [ 0 ], "anno_id": [ 139 ], "fps_inferred_from_clip_interval": 29.9636 }, "usage": { "input_tokens": 153959, "input_tokens_details": { "cache_write_tokens": 153956, "cached_tokens": 0 }, "output_tokens": 5057, "output_tokens_details": { "reasoning_tokens": 3362 }, "total_tokens": 159016 }, "model": "gpt-5.6-sol", "effort": "medium", "custom_id": "cat10_cap07_v_-ypKYhjrlXw_3387_4211", "annotation": { "caption": "While continuing to speak, the man applies a dot of cream from the blue-tipped tube onto the back of the woman's hand. He then uses his index finger to gently spread the cream across her hand.", "entity_groups": [ { "id": "E1", "label": "man", "spans": [ { "text": "the man", "range": [ 27, 34 ] }, { "text": "He", "range": [ 119, 121 ] }, { "text": "his", "range": [ 132, 135 ] } ], "source_mask_link": "selected_person" }, { "id": "E2", "label": "cream", "spans": [ { "text": "a dot of cream", "range": [ 43, 57 ] }, { "text": "the cream", "range": [ 166, 175 ] } ], "source_mask_link": null }, { "id": "E3", "label": "tube", "spans": [ { "text": "the blue-tipped tube", "range": [ 63, 83 ] } ], "source_mask_link": null }, { "id": "E4", "label": "back of hand", "spans": [ { "text": "the back", "range": [ 89, 97 ] } ], "source_mask_link": null }, { "id": "E5", "label": "woman", "spans": [ { "text": "the woman's", "range": [ 101, 112 ] }, { "text": "her", "range": [ 183, 186 ] } ], "source_mask_link": null }, { "id": "E6", "label": "woman's hand", "spans": [ { "text": "the woman's hand", "range": [ 101, 117 ] }, { "text": "her hand", "range": [ 183, 191 ] } ], "source_mask_link": null }, { "id": "E7", "label": "man's index finger", "spans": [ { "text": "his index finger", "range": [ 132, 148 ] } ], "source_mask_link": null } ], "actions": [ { "id": "A1", "mentions": [ { "text": "speak", "ranges": [ [ 20, 25 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "type": "simple", "subject_refs": [ "E1" ], "object_refs": [], "confidence": "medium", "temporal_status": "localized_sparse", "intervals": [ { "start_s": 0.0, "end_s": 21.0, "source_start_s": 113.0, "source_end_s": 134.0 } ], "evidence": "Across samples 0–20, the masked man repeatedly changes mouth shape and is often visibly speaking while gesturing and preparing the tube.", "mask_relation": "subject matches the supplied selected-person mask", "source_subject_refs": "E1" }, { "id": "A2", "mentions": [ { "text": "applies", "ranges": [ [ 35, 42 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "type": "simple", "subject_refs": [ "E1" ], "object_refs": [ "E2" ], "confidence": "high", "temporal_status": "localized_sparse", "intervals": [ { "start_s": 20.0, "end_s": 25.0, "source_start_s": 133.0, "source_end_s": 138.0 } ], "evidence": "Samples 20–24 show the masked man placing the blue tube tip against the back of the woman's hand and dispensing a visible cream dot.", "mask_relation": "subject matches the supplied selected-person mask", "source_subject_refs": "E1" }, { "id": "A3", "mentions": [ { "text": "uses", "ranges": [ [ 127, 131 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "type": "simple", "subject_refs": [ "E1" ], "object_refs": [ "E7" ], "confidence": "medium", "temporal_status": "localized_sparse", "intervals": [ { "start_s": 25.0, "end_s": 27.5, "source_start_s": 138.0, "source_end_s": 140.5 } ], "evidence": "Samples 25–27 show the masked man bringing his index finger into contact with the cream on the woman's hand.", "mask_relation": "subject matches the supplied selected-person mask", "source_subject_refs": "E1" }, { "id": "A4", "mentions": [ { "text": "spread", "ranges": [ [ 159, 165 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "type": "simple", "subject_refs": [ "E1" ], "object_refs": [ "E2" ], "confidence": "high", "temporal_status": "localized_sparse", "intervals": [ { "start_s": 25.0, "end_s": 27.5, "source_start_s": 138.0, "source_end_s": 140.5 } ], "evidence": "Samples 25–27 show the masked man's index finger contacting and moving through the cream; the action remains supported at the final sample and therefore ends at the clip boundary.", "mask_relation": "subject matches the supplied selected-person mask", "source_subject_refs": "E1" } ], "notes": [ "The supplied mask consistently identifies the man, including his arms and hands; it does not identify the woman, her hand, the cream, the tube, or the finger as separate instances.", "Entity groups without source-mask links are retained because the masklet is not a complete set of scene-instance masks.", "The aspectual word \"continuing\" and infinitival \"to\" are excluded from the speaking action span, and the manner adverb \"gently\" is excluded from the spreading span.", "Temporal intervals are sparse sampled-frame support, not exact continuous-video onset and offset boundaries." ] }, "model_output_raw": { "caption": "While continuing to speak, the man applies a dot of cream from the blue-tipped tube onto the back of the woman's hand. He then uses his index finger to gently spread the cream across her hand.", "entity_groups": [ { "id": "E1", "label": "man", "spans": [ { "text": "the man", "range": [ 27, 34 ] }, { "text": "He", "range": [ 119, 121 ] }, { "text": "his", "range": [ 132, 135 ] } ], "source_mask": "selected_person_mask" }, { "id": "E2", "label": "cream", "spans": [ { "text": "a dot of cream", "range": [ 43, 57 ] }, { "text": "the cream", "range": [ 166, 175 ] } ], "source_mask": null }, { "id": "E3", "label": "tube", "spans": [ { "text": "the blue-tipped tube", "range": [ 63, 83 ] } ], "source_mask": null }, { "id": "E4", "label": "back of hand", "spans": [ { "text": "the back", "range": [ 89, 97 ] } ], "source_mask": null }, { "id": "E5", "label": "woman", "spans": [ { "text": "the woman's", "range": [ 101, 112 ] }, { "text": "her", "range": [ 183, 186 ] } ], "source_mask": null }, { "id": "E6", "label": "woman's hand", "spans": [ { "text": "the woman's hand", "range": [ 101, 117 ] }, { "text": "her hand", "range": [ 183, 191 ] } ], "source_mask": null }, { "id": "E7", "label": "man's index finger", "spans": [ { "text": "his index finger", "range": [ 132, 148 ] } ], "source_mask": null } ], "actions": [ { "id": "A1", "mentions": [ { "text": "speak", "ranges": [ [ 20, 25 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "type": "simple", "subject_group": "E1", "object_groups": [], "instrument_groups": [], "target_groups": [], "confidence": "medium", "mask_relation": "subject matches the supplied selected-person mask", "temporal_status": "localized_sparse", "intervals": [ { "start_s": 0.0, "end_s": 21.0, "source_start_s": 113.0, "source_end_s": 134.0 } ], "evidence": "Across samples 0–20, the masked man repeatedly changes mouth shape and is often visibly speaking while gesturing and preparing the tube." }, { "id": "A2", "mentions": [ { "text": "applies", "ranges": [ [ 35, 42 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "type": "simple", "subject_group": "E1", "object_groups": [ "E2" ], "instrument_groups": [ "E3" ], "target_groups": [ "E4", "E6" ], "confidence": "high", "mask_relation": "subject matches the supplied selected-person mask", "temporal_status": "localized_sparse", "intervals": [ { "start_s": 20.0, "end_s": 25.0, "source_start_s": 133.0, "source_end_s": 138.0 } ], "evidence": "Samples 20–24 show the masked man placing the blue tube tip against the back of the woman's hand and dispensing a visible cream dot." }, { "id": "A3", "mentions": [ { "text": "uses", "ranges": [ [ 127, 131 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "type": "simple", "subject_group": "E1", "object_groups": [ "E7" ], "instrument_groups": [ "E7" ], "target_groups": [ "E2", "E6" ], "confidence": "medium", "mask_relation": "subject matches the supplied selected-person mask", "temporal_status": "localized_sparse", "intervals": [ { "start_s": 25.0, "end_s": 27.5, "source_start_s": 138.0, "source_end_s": 140.5 } ], "evidence": "Samples 25–27 show the masked man bringing his index finger into contact with the cream on the woman's hand." }, { "id": "A4", "mentions": [ { "text": "spread", "ranges": [ [ 159, 165 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "type": "simple", "subject_group": "E1", "object_groups": [ "E2" ], "instrument_groups": [ "E7" ], "target_groups": [ "E6" ], "confidence": "high", "mask_relation": "subject matches the supplied selected-person mask", "temporal_status": "localized_sparse", "intervals": [ { "start_s": 25.0, "end_s": 27.5, "source_start_s": 138.0, "source_end_s": 140.5 } ], "evidence": "Samples 25–27 show the masked man's index finger contacting and moving through the cream; the action remains supported at the final sample and therefore ends at the clip boundary." } ], "notes": [ "The supplied mask consistently identifies the man, including his arms and hands; it does not identify the woman, her hand, the cream, the tube, or the finger as separate instances.", "Entity groups without source-mask links are retained because the masklet is not a complete set of scene-instance masks.", "The aspectual word \"continuing\" and infinitival \"to\" are excluded from the speaking action span, and the manner adverb \"gently\" is excluded from the spreading span.", "Temporal intervals are sparse sampled-frame support, not exact continuous-video onset and offset boundaries." ] }, "normalization_repairs": 0 }, { "source_video_id": "v_-ypKYhjrlXw", "caption_count_category": 10, "clip": { "caption_index": 8, "caption": "The man picks up a larger white tube, squeezes a line of white cream onto the woman's forearm, and then uses his right hand to spread it evenly along her arm.", "source_interval_sec": [ 140.5, 177.0 ], "loom_clip_id": "v_-ypKYhjrlXw_4211_5306", "clip_frame_interval_4fps": [ 4211, 5306 ], "sample_count_1fps": 37, "sampled_frames": [ { "frame": "00000", "timestamp_sec": 140.5, "offset_sec": 0, "rgb_path": "assets/cat_10/v_-ypKYhjrlXw_4211_5306/0000s_00000_rgb.jpg", "mask_path": "assets/cat_10/v_-ypKYhjrlXw_4211_5306/0000s_00000_mask.png", "mask_status": "available" }, { "frame": "00028", "timestamp_sec": 141.5, "offset_sec": 1, "rgb_path": "assets/cat_10/v_-ypKYhjrlXw_4211_5306/0001s_00028_rgb.jpg", "mask_path": "assets/cat_10/v_-ypKYhjrlXw_4211_5306/0001s_00028_mask.png", "mask_status": "available" }, { "frame": "00060", "timestamp_sec": 142.5, "offset_sec": 2, "rgb_path": "assets/cat_10/v_-ypKYhjrlXw_4211_5306/0002s_00060_rgb.jpg", "mask_path": "assets/cat_10/v_-ypKYhjrlXw_4211_5306/0002s_00060_mask.png", "mask_status": "available" }, { "frame": "00088", "timestamp_sec": 143.5, "offset_sec": 3, "rgb_path": "assets/cat_10/v_-ypKYhjrlXw_4211_5306/0003s_00088_rgb.jpg", "mask_path": "assets/cat_10/v_-ypKYhjrlXw_4211_5306/0003s_00088_mask.png", "mask_status": "available" }, { "frame": "00120", "timestamp_sec": 144.5, "offset_sec": 4, "rgb_path": "assets/cat_10/v_-ypKYhjrlXw_4211_5306/0004s_00120_rgb.jpg", "mask_path": "assets/cat_10/v_-ypKYhjrlXw_4211_5306/0004s_00120_mask.png", "mask_status": "available" }, { "frame": "00148", "timestamp_sec": 145.5, "offset_sec": 5, "rgb_path": "assets/cat_10/v_-ypKYhjrlXw_4211_5306/0005s_00148_rgb.jpg", "mask_path": "assets/cat_10/v_-ypKYhjrlXw_4211_5306/0005s_00148_mask.png", "mask_status": "available" }, { "frame": "00180", "timestamp_sec": 146.5, "offset_sec": 6, "rgb_path": "assets/cat_10/v_-ypKYhjrlXw_4211_5306/0006s_00180_rgb.jpg", "mask_path": "assets/cat_10/v_-ypKYhjrlXw_4211_5306/0006s_00180_mask.png", "mask_status": "available" }, { "frame": "00208", "timestamp_sec": 147.5, "offset_sec": 7, "rgb_path": "assets/cat_10/v_-ypKYhjrlXw_4211_5306/0007s_00208_rgb.jpg", "mask_path": "assets/cat_10/v_-ypKYhjrlXw_4211_5306/0007s_00208_mask.png", "mask_status": "available" }, { "frame": "00240", "timestamp_sec": 148.5, "offset_sec": 8, "rgb_path": "assets/cat_10/v_-ypKYhjrlXw_4211_5306/0008s_00240_rgb.jpg", "mask_path": "assets/cat_10/v_-ypKYhjrlXw_4211_5306/0008s_00240_mask.png", "mask_status": "available" }, { "frame": "00268", "timestamp_sec": 149.5, "offset_sec": 9, "rgb_path": "assets/cat_10/v_-ypKYhjrlXw_4211_5306/0009s_00268_rgb.jpg", "mask_path": "assets/cat_10/v_-ypKYhjrlXw_4211_5306/0009s_00268_mask.png", "mask_status": "available" }, { "frame": "00300", "timestamp_sec": 150.5, "offset_sec": 10, "rgb_path": "assets/cat_10/v_-ypKYhjrlXw_4211_5306/0010s_00300_rgb.jpg", "mask_path": "assets/cat_10/v_-ypKYhjrlXw_4211_5306/0010s_00300_mask.png", "mask_status": "available" }, { "frame": "00328", "timestamp_sec": 151.5, "offset_sec": 11, "rgb_path": "assets/cat_10/v_-ypKYhjrlXw_4211_5306/0011s_00328_rgb.jpg", "mask_path": "assets/cat_10/v_-ypKYhjrlXw_4211_5306/0011s_00328_mask.png", "mask_status": "available" }, { "frame": "00360", "timestamp_sec": 152.5, "offset_sec": 12, "rgb_path": "assets/cat_10/v_-ypKYhjrlXw_4211_5306/0012s_00360_rgb.jpg", "mask_path": "assets/cat_10/v_-ypKYhjrlXw_4211_5306/0012s_00360_mask.png", "mask_status": "available" }, { "frame": "00388", "timestamp_sec": 153.5, "offset_sec": 13, "rgb_path": "assets/cat_10/v_-ypKYhjrlXw_4211_5306/0013s_00388_rgb.jpg", "mask_path": "assets/cat_10/v_-ypKYhjrlXw_4211_5306/0013s_00388_mask.png", "mask_status": "available" }, { "frame": "00420", "timestamp_sec": 154.5, "offset_sec": 14, "rgb_path": "assets/cat_10/v_-ypKYhjrlXw_4211_5306/0014s_00420_rgb.jpg", "mask_path": "assets/cat_10/v_-ypKYhjrlXw_4211_5306/0014s_00420_mask.png", "mask_status": "available" }, { "frame": "00448", "timestamp_sec": 155.5, "offset_sec": 15, "rgb_path": "assets/cat_10/v_-ypKYhjrlXw_4211_5306/0015s_00448_rgb.jpg", "mask_path": "assets/cat_10/v_-ypKYhjrlXw_4211_5306/0015s_00448_mask.png", "mask_status": "available" }, { "frame": "00480", "timestamp_sec": 156.5, "offset_sec": 16, "rgb_path": "assets/cat_10/v_-ypKYhjrlXw_4211_5306/0016s_00480_rgb.jpg", "mask_path": "assets/cat_10/v_-ypKYhjrlXw_4211_5306/0016s_00480_mask.png", "mask_status": "available" }, { "frame": "00508", "timestamp_sec": 157.5, "offset_sec": 17, "rgb_path": "assets/cat_10/v_-ypKYhjrlXw_4211_5306/0017s_00508_rgb.jpg", "mask_path": "assets/cat_10/v_-ypKYhjrlXw_4211_5306/0017s_00508_mask.png", "mask_status": "available" }, { "frame": "00540", "timestamp_sec": 158.5, "offset_sec": 18, "rgb_path": "assets/cat_10/v_-ypKYhjrlXw_4211_5306/0018s_00540_rgb.jpg", "mask_path": "assets/cat_10/v_-ypKYhjrlXw_4211_5306/0018s_00540_mask.png", "mask_status": "available" }, { "frame": "00568", "timestamp_sec": 159.5, "offset_sec": 19, "rgb_path": "assets/cat_10/v_-ypKYhjrlXw_4211_5306/0019s_00568_rgb.jpg", "mask_path": "assets/cat_10/v_-ypKYhjrlXw_4211_5306/0019s_00568_mask.png", "mask_status": "available" }, { "frame": "00600", "timestamp_sec": 160.5, "offset_sec": 20, "rgb_path": "assets/cat_10/v_-ypKYhjrlXw_4211_5306/0020s_00600_rgb.jpg", "mask_path": "assets/cat_10/v_-ypKYhjrlXw_4211_5306/0020s_00600_mask.png", "mask_status": "available" }, { "frame": "00628", "timestamp_sec": 161.5, "offset_sec": 21, "rgb_path": "assets/cat_10/v_-ypKYhjrlXw_4211_5306/0021s_00628_rgb.jpg", "mask_path": "assets/cat_10/v_-ypKYhjrlXw_4211_5306/0021s_00628_mask.png", "mask_status": "available" }, { "frame": "00660", "timestamp_sec": 162.5, "offset_sec": 22, "rgb_path": "assets/cat_10/v_-ypKYhjrlXw_4211_5306/0022s_00660_rgb.jpg", "mask_path": "assets/cat_10/v_-ypKYhjrlXw_4211_5306/0022s_00660_mask.png", "mask_status": "available" }, { "frame": "00688", "timestamp_sec": 163.5, "offset_sec": 23, "rgb_path": "assets/cat_10/v_-ypKYhjrlXw_4211_5306/0023s_00688_rgb.jpg", "mask_path": "assets/cat_10/v_-ypKYhjrlXw_4211_5306/0023s_00688_mask.png", "mask_status": "available" }, { "frame": "00720", "timestamp_sec": 164.5, "offset_sec": 24, "rgb_path": "assets/cat_10/v_-ypKYhjrlXw_4211_5306/0024s_00720_rgb.jpg", "mask_path": "assets/cat_10/v_-ypKYhjrlXw_4211_5306/0024s_00720_mask.png", "mask_status": "available" }, { "frame": "00748", "timestamp_sec": 165.5, "offset_sec": 25, "rgb_path": "assets/cat_10/v_-ypKYhjrlXw_4211_5306/0025s_00748_rgb.jpg", "mask_path": "assets/cat_10/v_-ypKYhjrlXw_4211_5306/0025s_00748_mask.png", "mask_status": "available" }, { "frame": "00780", "timestamp_sec": 166.5, "offset_sec": 26, "rgb_path": "assets/cat_10/v_-ypKYhjrlXw_4211_5306/0026s_00780_rgb.jpg", "mask_path": "assets/cat_10/v_-ypKYhjrlXw_4211_5306/0026s_00780_mask.png", "mask_status": "available" }, { "frame": "00808", "timestamp_sec": 167.5, "offset_sec": 27, "rgb_path": "assets/cat_10/v_-ypKYhjrlXw_4211_5306/0027s_00808_rgb.jpg", "mask_path": "assets/cat_10/v_-ypKYhjrlXw_4211_5306/0027s_00808_mask.png", "mask_status": "available" }, { "frame": "00840", "timestamp_sec": 168.5, "offset_sec": 28, "rgb_path": "assets/cat_10/v_-ypKYhjrlXw_4211_5306/0028s_00840_rgb.jpg", "mask_path": "assets/cat_10/v_-ypKYhjrlXw_4211_5306/0028s_00840_mask.png", "mask_status": "available" }, { "frame": "00868", "timestamp_sec": 169.5, "offset_sec": 29, "rgb_path": "assets/cat_10/v_-ypKYhjrlXw_4211_5306/0029s_00868_rgb.jpg", "mask_path": "assets/cat_10/v_-ypKYhjrlXw_4211_5306/0029s_00868_mask.png", "mask_status": "available" }, { "frame": "00900", "timestamp_sec": 170.5, "offset_sec": 30, "rgb_path": "assets/cat_10/v_-ypKYhjrlXw_4211_5306/0030s_00900_rgb.jpg", "mask_path": "assets/cat_10/v_-ypKYhjrlXw_4211_5306/0030s_00900_mask.png", "mask_status": "available" }, { "frame": "00928", "timestamp_sec": 171.5, "offset_sec": 31, "rgb_path": "assets/cat_10/v_-ypKYhjrlXw_4211_5306/0031s_00928_rgb.jpg", "mask_path": "assets/cat_10/v_-ypKYhjrlXw_4211_5306/0031s_00928_mask.png", "mask_status": "available" }, { "frame": "00960", "timestamp_sec": 172.5, "offset_sec": 32, "rgb_path": "assets/cat_10/v_-ypKYhjrlXw_4211_5306/0032s_00960_rgb.jpg", "mask_path": "assets/cat_10/v_-ypKYhjrlXw_4211_5306/0032s_00960_mask.png", "mask_status": "available" }, { "frame": "00988", "timestamp_sec": 173.5, "offset_sec": 33, "rgb_path": "assets/cat_10/v_-ypKYhjrlXw_4211_5306/0033s_00988_rgb.jpg", "mask_path": "assets/cat_10/v_-ypKYhjrlXw_4211_5306/0033s_00988_mask.png", "mask_status": "available" }, { "frame": "01020", "timestamp_sec": 174.5, "offset_sec": 34, "rgb_path": "assets/cat_10/v_-ypKYhjrlXw_4211_5306/0034s_01020_rgb.jpg", "mask_path": "assets/cat_10/v_-ypKYhjrlXw_4211_5306/0034s_01020_mask.png", "mask_status": "available" }, { "frame": "01048", "timestamp_sec": 175.5, "offset_sec": 35, "rgb_path": "assets/cat_10/v_-ypKYhjrlXw_4211_5306/0035s_01048_rgb.jpg", "mask_path": "assets/cat_10/v_-ypKYhjrlXw_4211_5306/0035s_01048_mask.png", "mask_status": "available" }, { "frame": "01080", "timestamp_sec": 176.5, "offset_sec": 36, "rgb_path": "assets/cat_10/v_-ypKYhjrlXw_4211_5306/0036s_01080_rgb.jpg", "mask_path": "assets/cat_10/v_-ypKYhjrlXw_4211_5306/0036s_01080_mask.png", "mask_status": "available" } ], "obj_id": [ 0 ], "anno_id": [ 140 ], "fps_inferred_from_clip_interval": 30.0 }, "usage": { "input_tokens": 198513, "input_tokens_details": { "cache_write_tokens": 198510, "cached_tokens": 0 }, "output_tokens": 3789, "output_tokens_details": { "reasoning_tokens": 2469 }, "total_tokens": 202302 }, "model": "gpt-5.6-sol", "effort": "medium", "custom_id": "cat10_cap08_v_-ypKYhjrlXw_4211_5306", "annotation": { "caption": "The man picks up a larger white tube, squeezes a line of white cream onto the woman's forearm, and then uses his right hand to spread it evenly along her arm.", "entity_groups": [ { "id": "E1", "label": "man", "spans": [ { "text": "The man", "range": [ 0, 7 ] }, { "text": "his", "range": [ 109, 112 ] } ], "source_mask_link": "selected_person" }, { "id": "E2", "label": "tube", "spans": [ { "text": "a larger white tube", "range": [ 17, 36 ] } ], "source_mask_link": null }, { "id": "E3", "label": "white cream", "spans": [ { "text": "a line of white cream", "range": [ 47, 68 ] }, { "text": "it", "range": [ 134, 136 ] } ], "source_mask_link": null }, { "id": "E4", "label": "woman", "spans": [ { "text": "the woman's", "range": [ 74, 85 ] }, { "text": "her", "range": [ 150, 153 ] } ], "source_mask_link": null }, { "id": "E5", "label": "woman's forearm/arm", "spans": [ { "text": "the woman's forearm", "range": [ 74, 93 ] }, { "text": "her arm", "range": [ 150, 157 ] } ], "source_mask_link": null }, { "id": "E6", "label": "man's right hand", "spans": [ { "text": "his right hand", "range": [ 109, 123 ] } ], "source_mask_link": null } ], "actions": [ { "id": "A1", "mentions": [ { "text": "picks up", "ranges": [ [ 8, 16 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "type": "phrasal", "subject_refs": [ "E1" ], "object_refs": [ "E2" ], "confidence": "medium", "temporal_status": "supported_sparse", "intervals": [ { "start_s": 11.0, "end_s": 12.0, "source_start_s": 151.5, "source_end_s": 152.5 } ], "evidence": "At 11.0 s the masked man has newly brought the larger white tube into his hand; it was not visible in his hand at 10.0 s.", "mask_relation": "The action subject matches the supplied selected-person mask.", "source_subject_refs": "E1" }, { "id": "A2", "mentions": [ { "text": "squeezes", "ranges": [ [ 38, 46 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "type": "simple", "subject_refs": [ "E1" ], "object_refs": [ "E2", "E3" ], "confidence": "high", "temporal_status": "supported_sparse", "intervals": [ { "start_s": 20.0, "end_s": 25.0, "source_start_s": 160.5, "source_end_s": 165.5 } ], "evidence": "From 20.0–24.0 s the masked man holds the open tube against the woman's forearm and compresses it; a dollop of white cream is visible by 24.0 s.", "mask_relation": "The action subject matches the supplied selected-person mask.", "source_subject_refs": "E1" }, { "id": "A3", "mentions": [ { "text": "spread", "ranges": [ [ 127, 133 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "type": "simple", "subject_refs": [ "E1" ], "object_refs": [ "E3" ], "confidence": "high", "temporal_status": "supported_sparse", "intervals": [ { "start_s": 25.0, "end_s": 36.5, "source_start_s": 165.5, "source_end_s": 177.0 } ], "evidence": "At 25.0–26.0 s the masked man contacts and begins flattening the cream, then repeatedly rubs it along the forearm through the final 36.0 s sample.", "mask_relation": "The action subject matches the supplied selected-person mask.", "source_subject_refs": "E1" } ], "notes": [ "The supplied masklet consistently identifies the man in the white coat, so only actions performed by that man are annotated.", "No separate source masks were supplied for the woman, tube, cream, forearm/arm, or right hand; their source-mask links remain null.", "The clip interval is [140.5, 177.0) source-video seconds, corresponding to [0.0, 36.5) clip-relative seconds.", "\"uses\" is treated as instrumental framing for the explicitly stated spreading action rather than as a separate physical event." ] }, "model_output_raw": { "caption": "The man picks up a larger white tube, squeezes a line of white cream onto the woman's forearm, and then uses his right hand to spread it evenly along her arm.", "entity_groups": [ { "id": "E1", "label": "man", "spans": [ { "text": "The man", "range": [ 0, 7 ] }, { "text": "his", "range": [ 109, 112 ] } ], "source_mask": "selected_person_masklet" }, { "id": "E2", "label": "tube", "spans": [ { "text": "a larger white tube", "range": [ 17, 36 ] } ], "source_mask": null }, { "id": "E3", "label": "white cream", "spans": [ { "text": "a line of white cream", "range": [ 47, 68 ] }, { "text": "it", "range": [ 134, 136 ] } ], "source_mask": null }, { "id": "E4", "label": "woman", "spans": [ { "text": "the woman's", "range": [ 74, 85 ] }, { "text": "her", "range": [ 150, 153 ] } ], "source_mask": null }, { "id": "E5", "label": "woman's forearm/arm", "spans": [ { "text": "the woman's forearm", "range": [ 74, 93 ] }, { "text": "her arm", "range": [ 150, 157 ] } ], "source_mask": null }, { "id": "E6", "label": "man's right hand", "spans": [ { "text": "his right hand", "range": [ 109, 123 ] } ], "source_mask": null } ], "actions": [ { "id": "A1", "mentions": [ { "text": "picks up", "ranges": [ [ 8, 16 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "type": "phrasal", "subject": "E1", "objects": [ "E2" ], "confidence": "medium", "temporal_status": "supported_sparse", "intervals": [ { "clip_relative_s": [ 11.0, 12.0 ], "source_video_s": [ 151.5, 152.5 ] } ], "evidence": "At 11.0 s the masked man has newly brought the larger white tube into his hand; it was not visible in his hand at 10.0 s.", "mask_relation": "The action subject matches the supplied selected-person mask." }, { "id": "A2", "mentions": [ { "text": "squeezes", "ranges": [ [ 38, 46 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "type": "simple", "subject": "E1", "objects": [ "E2", "E3" ], "targets": [ "E5" ], "confidence": "high", "temporal_status": "supported_sparse", "intervals": [ { "clip_relative_s": [ 20.0, 25.0 ], "source_video_s": [ 160.5, 165.5 ] } ], "evidence": "From 20.0–24.0 s the masked man holds the open tube against the woman's forearm and compresses it; a dollop of white cream is visible by 24.0 s.", "mask_relation": "The action subject matches the supplied selected-person mask." }, { "id": "A3", "mentions": [ { "text": "spread", "ranges": [ [ 127, 133 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "type": "simple", "subject": "E1", "objects": [ "E3" ], "targets": [ "E5" ], "instruments": [ "E6" ], "confidence": "high", "temporal_status": "supported_sparse", "intervals": [ { "clip_relative_s": [ 25.0, 36.5 ], "source_video_s": [ 165.5, 177.0 ] } ], "evidence": "At 25.0–26.0 s the masked man contacts and begins flattening the cream, then repeatedly rubs it along the forearm through the final 36.0 s sample.", "mask_relation": "The action subject matches the supplied selected-person mask." } ], "notes": [ "The supplied masklet consistently identifies the man in the white coat, so only actions performed by that man are annotated.", "No separate source masks were supplied for the woman, tube, cream, forearm/arm, or right hand; their source-mask links remain null.", "The clip interval is [140.5, 177.0) source-video seconds, corresponding to [0.0, 36.5) clip-relative seconds.", "\"uses\" is treated as instrumental framing for the explicitly stated spreading action rather than as a separate physical event." ] }, "normalization_repairs": 0 }, { "source_video_id": "v_-ypKYhjrlXw", "caption_count_category": 10, "clip": { "caption_index": 9, "caption": "The man picks up the round compact again, holds it open with both hands, and gestures towards it while speaking. He then closes the compact and places it back on the table.", "source_interval_sec": [ 177.0, 188.5 ], "loom_clip_id": "v_-ypKYhjrlXw_5306_5650", "clip_frame_interval_4fps": [ 5306, 5650 ], "sample_count_1fps": 12, "sampled_frames": [ { "frame": "00000", "timestamp_sec": 177.0, "offset_sec": 0, "rgb_path": "assets/cat_10/v_-ypKYhjrlXw_5306_5650/0000s_00000_rgb.jpg", "mask_path": "assets/cat_10/v_-ypKYhjrlXw_5306_5650/0000s_00000_mask.png", "mask_status": "available" }, { "frame": "00028", "timestamp_sec": 178.0, "offset_sec": 1, "rgb_path": "assets/cat_10/v_-ypKYhjrlXw_5306_5650/0001s_00028_rgb.jpg", "mask_path": "assets/cat_10/v_-ypKYhjrlXw_5306_5650/0001s_00028_mask.png", "mask_status": "available" }, { "frame": "00060", "timestamp_sec": 179.0, "offset_sec": 2, "rgb_path": "assets/cat_10/v_-ypKYhjrlXw_5306_5650/0002s_00060_rgb.jpg", "mask_path": "assets/cat_10/v_-ypKYhjrlXw_5306_5650/0002s_00060_mask.png", "mask_status": "available" }, { "frame": "00088", "timestamp_sec": 180.0, "offset_sec": 3, "rgb_path": "assets/cat_10/v_-ypKYhjrlXw_5306_5650/0003s_00088_rgb.jpg", "mask_path": "assets/cat_10/v_-ypKYhjrlXw_5306_5650/0003s_00088_mask.png", "mask_status": "available" }, { "frame": "00120", "timestamp_sec": 181.0, "offset_sec": 4, "rgb_path": "assets/cat_10/v_-ypKYhjrlXw_5306_5650/0004s_00120_rgb.jpg", "mask_path": "assets/cat_10/v_-ypKYhjrlXw_5306_5650/0004s_00120_mask.png", "mask_status": "available" }, { "frame": "00148", "timestamp_sec": 182.0, "offset_sec": 5, "rgb_path": "assets/cat_10/v_-ypKYhjrlXw_5306_5650/0005s_00148_rgb.jpg", "mask_path": "assets/cat_10/v_-ypKYhjrlXw_5306_5650/0005s_00148_mask.png", "mask_status": "available" }, { "frame": "00180", "timestamp_sec": 183.0, "offset_sec": 6, "rgb_path": "assets/cat_10/v_-ypKYhjrlXw_5306_5650/0006s_00180_rgb.jpg", "mask_path": "assets/cat_10/v_-ypKYhjrlXw_5306_5650/0006s_00180_mask.png", "mask_status": "available" }, { "frame": "00208", "timestamp_sec": 184.0, "offset_sec": 7, "rgb_path": "assets/cat_10/v_-ypKYhjrlXw_5306_5650/0007s_00208_rgb.jpg", "mask_path": "assets/cat_10/v_-ypKYhjrlXw_5306_5650/0007s_00208_mask.png", "mask_status": "available" }, { "frame": "00240", "timestamp_sec": 185.0, "offset_sec": 8, "rgb_path": "assets/cat_10/v_-ypKYhjrlXw_5306_5650/0008s_00240_rgb.jpg", "mask_path": "assets/cat_10/v_-ypKYhjrlXw_5306_5650/0008s_00240_mask.png", "mask_status": "available" }, { "frame": "00268", "timestamp_sec": 186.0, "offset_sec": 9, "rgb_path": "assets/cat_10/v_-ypKYhjrlXw_5306_5650/0009s_00268_rgb.jpg", "mask_path": "assets/cat_10/v_-ypKYhjrlXw_5306_5650/0009s_00268_mask.png", "mask_status": "available" }, { "frame": "00300", "timestamp_sec": 187.0, "offset_sec": 10, "rgb_path": "assets/cat_10/v_-ypKYhjrlXw_5306_5650/0010s_00300_rgb.jpg", "mask_path": "assets/cat_10/v_-ypKYhjrlXw_5306_5650/0010s_00300_mask.png", "mask_status": "available" }, { "frame": "00328", "timestamp_sec": 188.0, "offset_sec": 11, "rgb_path": "assets/cat_10/v_-ypKYhjrlXw_5306_5650/0011s_00328_rgb.jpg", "mask_path": "assets/cat_10/v_-ypKYhjrlXw_5306_5650/0011s_00328_mask.png", "mask_status": "available" } ], "obj_id": [ 0 ], "anno_id": [ 141 ], "fps_inferred_from_clip_interval": 29.913 }, "usage": { "input_tokens": 74738, "input_tokens_details": { "cache_write_tokens": 74735, "cached_tokens": 0 }, "output_tokens": 4871, "output_tokens_details": { "reasoning_tokens": 3242 }, "total_tokens": 79609 }, "model": "gpt-5.6-sol", "effort": "medium", "custom_id": "cat10_cap09_v_-ypKYhjrlXw_5306_5650", "annotation": { "caption": "The man picks up the round compact again, holds it open with both hands, and gestures towards it while speaking. He then closes the compact and places it back on the table.", "entity_groups": [ { "id": "entity_man", "label": "man", "spans": [ { "text": "The man", "range": [ 0, 7 ] }, { "text": "He", "range": [ 113, 115 ] } ], "source_mask_link": "selected_person" }, { "id": "entity_compact", "label": "compact", "spans": [ { "text": "the round compact", "range": [ 17, 34 ] }, { "text": "it", "range": [ 48, 50 ] }, { "text": "it", "range": [ 94, 96 ] }, { "text": "the compact", "range": [ 128, 139 ] }, { "text": "it", "range": [ 151, 153 ] } ], "source_mask_link": null }, { "id": "entity_hands", "label": "hands", "spans": [ { "text": "both hands", "range": [ 61, 71 ] } ], "source_mask_link": null }, { "id": "entity_table", "label": "table", "spans": [ { "text": "the table", "range": [ 162, 171 ] } ], "source_mask_link": null } ], "actions": [ { "id": "action_1", "mentions": [ { "text": "picks up", "ranges": [ [ 8, 16 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "type": "phrasal", "subject_refs": [ "entity_man" ], "object_refs": [ "entity_compact" ], "confidence": "high", "temporal_status": "uncertain", "intervals": [], "evidence": "The compact is not clearly being lifted in any sampled frame; it is already open and held by 2.000 seconds.", "mask_relation": "The action subject matches the supplied selected-person mask.", "source_subject_refs": "entity_man" }, { "id": "action_2", "mentions": [ { "text": "holds", "ranges": [ [ 42, 47 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "type": "simple", "subject_refs": [ "entity_man" ], "object_refs": [ "entity_compact" ], "confidence": "high", "temporal_status": "localized", "intervals": [ { "start_s": 2.0, "end_s": 8.0, "source_start_s": 179.0, "source_end_s": 185.0 } ], "evidence": "At clip-relative 2.000 through 7.000 seconds, the masked man visibly supports and displays the open compact.", "mask_relation": "The action subject matches the supplied selected-person mask.", "source_subject_refs": "entity_man" }, { "id": "action_3", "mentions": [ { "text": "gestures", "ranges": [ [ 77, 85 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "type": "prepositional", "subject_refs": [ "entity_man" ], "object_refs": [ "entity_compact" ], "confidence": "high", "temporal_status": "localized", "intervals": [ { "start_s": 2.0, "end_s": 3.0, "source_start_s": 179.0, "source_end_s": 180.0 } ], "evidence": "At 2.000 seconds, the masked man presents an open hand beside and toward the displayed compact.", "mask_relation": "The action subject matches the supplied selected-person mask.", "source_subject_refs": "entity_man" }, { "id": "action_4", "mentions": [ { "text": "speaking", "ranges": [ [ 103, 111 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "type": "simple", "subject_refs": [ "entity_man" ], "object_refs": [], "confidence": "high", "temporal_status": "localized", "intervals": [ { "start_s": 3.0, "end_s": 11.5, "source_start_s": 180.0, "source_end_s": 188.5 } ], "evidence": "From 3.000 through the final 11.000-second sample, the masked man's changing open-mouth configurations support continued speech.", "mask_relation": "The action subject matches the supplied selected-person mask.", "source_subject_refs": "entity_man" }, { "id": "action_5", "mentions": [ { "text": "closes", "ranges": [ [ 121, 127 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "type": "simple", "subject_refs": [ "entity_man" ], "object_refs": [ "entity_compact" ], "confidence": "high", "temporal_status": "localized", "intervals": [ { "start_s": 8.0, "end_s": 9.0, "source_start_s": 185.0, "source_end_s": 186.0 } ], "evidence": "At 8.000 seconds the lid is being brought down; by 9.000 seconds the compact is closed.", "mask_relation": "The action subject matches the supplied selected-person mask.", "source_subject_refs": "entity_man" }, { "id": "action_6", "mentions": [ { "text": "places", "ranges": [ [ 144, 150 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "type": "prepositional", "subject_refs": [ "entity_man" ], "object_refs": [ "entity_compact" ], "confidence": "high", "temporal_status": "localized", "intervals": [ { "start_s": 9.0, "end_s": 11.0, "source_start_s": 186.0, "source_end_s": 188.0 } ], "evidence": "At 9.000 seconds the masked man holds the closed compact; at 10.000 seconds he reaches left and the compact is no longer in his hands, supporting placement on the adjacent table.", "mask_relation": "The action subject matches the supplied selected-person mask.", "source_subject_refs": "entity_man" } ], "notes": [ "The supplied masklet consistently identifies the captioned man.", "No separate source masks were supplied for the compact, the man's hands, or the table.", "The clip covers source-video interval [177.000000, 188.500000), corresponding to clip-relative [0.000000, 11.500000).", "\"open\" in \"holds it open\" describes the compact's state and is not annotated as a separate action." ] }, "model_output_raw": { "caption": "The man picks up the round compact again, holds it open with both hands, and gestures towards it while speaking. He then closes the compact and places it back on the table.", "entity_groups": [ { "id": "entity_man", "label": "man", "spans": [ { "text": "The man", "range": [ 0, 7 ] }, { "text": "He", "range": [ 113, 115 ] } ], "source_mask_link": "selected_person_masklet" }, { "id": "entity_compact", "label": "compact", "spans": [ { "text": "the round compact", "range": [ 17, 34 ] }, { "text": "it", "range": [ 48, 50 ] }, { "text": "it", "range": [ 94, 96 ] }, { "text": "the compact", "range": [ 128, 139 ] }, { "text": "it", "range": [ 151, 153 ] } ], "source_mask_link": null }, { "id": "entity_hands", "label": "hands", "spans": [ { "text": "both hands", "range": [ 61, 71 ] } ], "source_mask_link": null }, { "id": "entity_table", "label": "table", "spans": [ { "text": "the table", "range": [ 162, 171 ] } ], "source_mask_link": null } ], "actions": [ { "mentions": [ { "text": "picks up", "ranges": [ [ 8, 16 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "type": "phrasal", "subject_entity_group_id": "entity_man", "object_entity_group_ids": [ "entity_compact" ], "confidence": "high", "temporal_status": "uncertain", "intervals": [], "evidence": "The compact is not clearly being lifted in any sampled frame; it is already open and held by 2.000 seconds.", "mask_relation": "The action subject matches the supplied selected-person mask." }, { "mentions": [ { "text": "holds", "ranges": [ [ 42, 47 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "type": "simple", "subject_entity_group_id": "entity_man", "object_entity_group_ids": [ "entity_compact" ], "confidence": "high", "temporal_status": "localized", "intervals": [ { "start_s": 2.0, "end_s": 8.0 } ], "evidence": "At clip-relative 2.000 through 7.000 seconds, the masked man visibly supports and displays the open compact.", "mask_relation": "The action subject matches the supplied selected-person mask." }, { "mentions": [ { "text": "gestures", "ranges": [ [ 77, 85 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "type": "prepositional", "subject_entity_group_id": "entity_man", "object_entity_group_ids": [ "entity_compact" ], "confidence": "high", "temporal_status": "localized", "intervals": [ { "start_s": 2.0, "end_s": 3.0 } ], "evidence": "At 2.000 seconds, the masked man presents an open hand beside and toward the displayed compact.", "mask_relation": "The action subject matches the supplied selected-person mask." }, { "mentions": [ { "text": "speaking", "ranges": [ [ 103, 111 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "type": "simple", "subject_entity_group_id": "entity_man", "object_entity_group_ids": [], "confidence": "high", "temporal_status": "localized", "intervals": [ { "start_s": 3.0, "end_s": 11.5 } ], "evidence": "From 3.000 through the final 11.000-second sample, the masked man's changing open-mouth configurations support continued speech.", "mask_relation": "The action subject matches the supplied selected-person mask." }, { "mentions": [ { "text": "closes", "ranges": [ [ 121, 127 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "type": "simple", "subject_entity_group_id": "entity_man", "object_entity_group_ids": [ "entity_compact" ], "confidence": "high", "temporal_status": "localized", "intervals": [ { "start_s": 8.0, "end_s": 9.0 } ], "evidence": "At 8.000 seconds the lid is being brought down; by 9.000 seconds the compact is closed.", "mask_relation": "The action subject matches the supplied selected-person mask." }, { "mentions": [ { "text": "places", "ranges": [ [ 144, 150 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "type": "prepositional", "subject_entity_group_id": "entity_man", "object_entity_group_ids": [ "entity_compact" ], "location_entity_group_ids": [ "entity_table" ], "confidence": "high", "temporal_status": "localized", "intervals": [ { "start_s": 9.0, "end_s": 11.0 } ], "evidence": "At 9.000 seconds the masked man holds the closed compact; at 10.000 seconds he reaches left and the compact is no longer in his hands, supporting placement on the adjacent table.", "mask_relation": "The action subject matches the supplied selected-person mask." } ], "notes": [ "The supplied masklet consistently identifies the captioned man.", "No separate source masks were supplied for the compact, the man's hands, or the table.", "The clip covers source-video interval [177.000000, 188.500000), corresponding to clip-relative [0.000000, 11.500000).", "\"open\" in \"holds it open\" describes the compact's state and is not annotated as a separate action." ] }, "normalization_repairs": 0 }, { "source_video_id": "v_-ypKYhjrlXw", "caption_count_category": 10, "clip": { "caption_index": 10, "caption": "The man places his right hand on the woman's hand and continues speaking. He makes various gestures with his hands to emphasize his points before resting his hands on the table.", "source_interval_sec": [ 188.5, 212.68 ], "loom_clip_id": "v_-ypKYhjrlXw_5650_6375", "clip_frame_interval_4fps": [ 5650, 6375 ], "sample_count_1fps": 25, "sampled_frames": [ { "frame": "00000", "timestamp_sec": 188.5, "offset_sec": 0, "rgb_path": "assets/cat_10/v_-ypKYhjrlXw_5650_6375/0000s_00000_rgb.jpg", "mask_path": "assets/cat_10/v_-ypKYhjrlXw_5650_6375/0000s_00000_mask.png", "mask_status": "available" }, { "frame": "00028", "timestamp_sec": 189.5, "offset_sec": 1, "rgb_path": "assets/cat_10/v_-ypKYhjrlXw_5650_6375/0001s_00028_rgb.jpg", "mask_path": "assets/cat_10/v_-ypKYhjrlXw_5650_6375/0001s_00028_mask.png", "mask_status": "available" }, { "frame": "00060", "timestamp_sec": 190.5, "offset_sec": 2, "rgb_path": "assets/cat_10/v_-ypKYhjrlXw_5650_6375/0002s_00060_rgb.jpg", "mask_path": "assets/cat_10/v_-ypKYhjrlXw_5650_6375/0002s_00060_mask.png", "mask_status": "available" }, { "frame": "00088", "timestamp_sec": 191.5, "offset_sec": 3, "rgb_path": "assets/cat_10/v_-ypKYhjrlXw_5650_6375/0003s_00088_rgb.jpg", "mask_path": "assets/cat_10/v_-ypKYhjrlXw_5650_6375/0003s_00088_mask.png", "mask_status": "available" }, { "frame": "00120", "timestamp_sec": 192.5, "offset_sec": 4, "rgb_path": "assets/cat_10/v_-ypKYhjrlXw_5650_6375/0004s_00120_rgb.jpg", "mask_path": "assets/cat_10/v_-ypKYhjrlXw_5650_6375/0004s_00120_mask.png", "mask_status": "available" }, { "frame": "00148", "timestamp_sec": 193.5, "offset_sec": 5, "rgb_path": "assets/cat_10/v_-ypKYhjrlXw_5650_6375/0005s_00148_rgb.jpg", "mask_path": "assets/cat_10/v_-ypKYhjrlXw_5650_6375/0005s_00148_mask.png", "mask_status": "available" }, { "frame": "00180", "timestamp_sec": 194.5, "offset_sec": 6, "rgb_path": "assets/cat_10/v_-ypKYhjrlXw_5650_6375/0006s_00180_rgb.jpg", "mask_path": "assets/cat_10/v_-ypKYhjrlXw_5650_6375/0006s_00180_mask.png", "mask_status": "available" }, { "frame": "00208", "timestamp_sec": 195.5, "offset_sec": 7, "rgb_path": "assets/cat_10/v_-ypKYhjrlXw_5650_6375/0007s_00208_rgb.jpg", "mask_path": "assets/cat_10/v_-ypKYhjrlXw_5650_6375/0007s_00208_mask.png", "mask_status": "available" }, { "frame": "00240", "timestamp_sec": 196.5, "offset_sec": 8, "rgb_path": "assets/cat_10/v_-ypKYhjrlXw_5650_6375/0008s_00240_rgb.jpg", "mask_path": "assets/cat_10/v_-ypKYhjrlXw_5650_6375/0008s_00240_mask.png", "mask_status": "available" }, { "frame": "00268", "timestamp_sec": 197.5, "offset_sec": 9, "rgb_path": "assets/cat_10/v_-ypKYhjrlXw_5650_6375/0009s_00268_rgb.jpg", "mask_path": "assets/cat_10/v_-ypKYhjrlXw_5650_6375/0009s_00268_mask.png", "mask_status": "available" }, { "frame": "00300", "timestamp_sec": 198.5, "offset_sec": 10, "rgb_path": "assets/cat_10/v_-ypKYhjrlXw_5650_6375/0010s_00300_rgb.jpg", "mask_path": "assets/cat_10/v_-ypKYhjrlXw_5650_6375/0010s_00300_mask.png", "mask_status": "available" }, { "frame": "00328", "timestamp_sec": 199.5, "offset_sec": 11, "rgb_path": "assets/cat_10/v_-ypKYhjrlXw_5650_6375/0011s_00328_rgb.jpg", "mask_path": "assets/cat_10/v_-ypKYhjrlXw_5650_6375/0011s_00328_mask.png", "mask_status": "available" }, { "frame": "00360", "timestamp_sec": 200.5, "offset_sec": 12, "rgb_path": "assets/cat_10/v_-ypKYhjrlXw_5650_6375/0012s_00360_rgb.jpg", "mask_path": "assets/cat_10/v_-ypKYhjrlXw_5650_6375/0012s_00360_mask.png", "mask_status": "available" }, { "frame": "00388", "timestamp_sec": 201.5, "offset_sec": 13, "rgb_path": "assets/cat_10/v_-ypKYhjrlXw_5650_6375/0013s_00388_rgb.jpg", "mask_path": "assets/cat_10/v_-ypKYhjrlXw_5650_6375/0013s_00388_mask.png", "mask_status": "available" }, { "frame": "00420", "timestamp_sec": 202.5, "offset_sec": 14, "rgb_path": "assets/cat_10/v_-ypKYhjrlXw_5650_6375/0014s_00420_rgb.jpg", "mask_path": "assets/cat_10/v_-ypKYhjrlXw_5650_6375/0014s_00420_mask.png", "mask_status": "available" }, { "frame": "00448", "timestamp_sec": 203.5, "offset_sec": 15, "rgb_path": "assets/cat_10/v_-ypKYhjrlXw_5650_6375/0015s_00448_rgb.jpg", "mask_path": "assets/cat_10/v_-ypKYhjrlXw_5650_6375/0015s_00448_mask.png", "mask_status": "available" }, { "frame": "00480", "timestamp_sec": 204.5, "offset_sec": 16, "rgb_path": "assets/cat_10/v_-ypKYhjrlXw_5650_6375/0016s_00480_rgb.jpg", "mask_path": "assets/cat_10/v_-ypKYhjrlXw_5650_6375/0016s_00480_mask.png", "mask_status": "available" }, { "frame": "00508", "timestamp_sec": 205.5, "offset_sec": 17, "rgb_path": "assets/cat_10/v_-ypKYhjrlXw_5650_6375/0017s_00508_rgb.jpg", "mask_path": "assets/cat_10/v_-ypKYhjrlXw_5650_6375/0017s_00508_mask.png", "mask_status": "available" }, { "frame": "00540", "timestamp_sec": 206.5, "offset_sec": 18, "rgb_path": "assets/cat_10/v_-ypKYhjrlXw_5650_6375/0018s_00540_rgb.jpg", "mask_path": "assets/cat_10/v_-ypKYhjrlXw_5650_6375/0018s_00540_mask.png", "mask_status": "available" }, { "frame": "00568", "timestamp_sec": 207.5, "offset_sec": 19, "rgb_path": "assets/cat_10/v_-ypKYhjrlXw_5650_6375/0019s_00568_rgb.jpg", "mask_path": "assets/cat_10/v_-ypKYhjrlXw_5650_6375/0019s_00568_mask.png", "mask_status": "available" }, { "frame": "00600", "timestamp_sec": 208.5, "offset_sec": 20, "rgb_path": "assets/cat_10/v_-ypKYhjrlXw_5650_6375/0020s_00600_rgb.jpg", "mask_path": "assets/cat_10/v_-ypKYhjrlXw_5650_6375/0020s_00600_mask.png", "mask_status": "available" }, { "frame": "00628", "timestamp_sec": 209.5, "offset_sec": 21, "rgb_path": "assets/cat_10/v_-ypKYhjrlXw_5650_6375/0021s_00628_rgb.jpg", "mask_path": "assets/cat_10/v_-ypKYhjrlXw_5650_6375/0021s_00628_mask.png", "mask_status": "available" }, { "frame": "00660", "timestamp_sec": 210.5, "offset_sec": 22, "rgb_path": "assets/cat_10/v_-ypKYhjrlXw_5650_6375/0022s_00660_rgb.jpg", "mask_path": "assets/cat_10/v_-ypKYhjrlXw_5650_6375/0022s_00660_mask.png", "mask_status": "available" }, { "frame": "00688", "timestamp_sec": 211.5, "offset_sec": 23, "rgb_path": "assets/cat_10/v_-ypKYhjrlXw_5650_6375/0023s_00688_rgb.jpg", "mask_path": "assets/cat_10/v_-ypKYhjrlXw_5650_6375/0023s_00688_mask.png", "mask_status": "available" }, { "frame": "00720", "timestamp_sec": 212.5, "offset_sec": 24, "rgb_path": "assets/cat_10/v_-ypKYhjrlXw_5650_6375/0024s_00720_rgb.jpg", "mask_path": "assets/cat_10/v_-ypKYhjrlXw_5650_6375/0024s_00720_mask.png", "mask_status": "available" } ], "obj_id": [ 0 ], "anno_id": [ 142 ], "fps_inferred_from_clip_interval": 29.9835 }, "usage": { "input_tokens": 139098, "input_tokens_details": { "cache_write_tokens": 139095, "cached_tokens": 0 }, "output_tokens": 4005, "output_tokens_details": { "reasoning_tokens": 2355 }, "total_tokens": 143103 }, "model": "gpt-5.6-sol", "effort": "medium", "custom_id": "cat10_cap10_v_-ypKYhjrlXw_5650_6375", "annotation": { "caption": "The man places his right hand on the woman's hand and continues speaking. He makes various gestures with his hands to emphasize his points before resting his hands on the table.", "entity_groups": [ { "id": "E1", "label": "man", "spans": [ { "text": "The man", "range": [ 0, 7 ] }, { "text": "his", "range": [ 15, 18 ] }, { "text": "He", "range": [ 74, 76 ] }, { "text": "his", "range": [ 105, 108 ] }, { "text": "his", "range": [ 128, 131 ] }, { "text": "his", "range": [ 154, 157 ] } ], "source_mask_link": "selected_person" }, { "id": "E2", "label": "man's right hand", "spans": [ { "text": "his right hand", "range": [ 15, 29 ] } ], "source_mask_link": null }, { "id": "E3", "label": "woman", "spans": [ { "text": "the woman's", "range": [ 33, 44 ] } ], "source_mask_link": null }, { "id": "E4", "label": "woman's hand", "spans": [ { "text": "the woman's hand", "range": [ 33, 49 ] } ], "source_mask_link": null }, { "id": "E5", "label": "man's hands", "spans": [ { "text": "his hands", "range": [ 105, 114 ] }, { "text": "his hands", "range": [ 154, 163 ] } ], "source_mask_link": null }, { "id": "E6", "label": "table", "spans": [ { "text": "the table", "range": [ 167, 176 ] } ], "source_mask_link": null } ], "actions": [ { "id": "A1", "mentions": [ { "text": "places", "ranges": [ [ 8, 14 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "type": "simple", "subject_refs": [ "E1" ], "object_refs": [ "E4" ], "confidence": "high", "temporal_status": "localized", "intervals": [ { "start_s": 0.0, "end_s": 2.0, "source_start_s": 188.5, "source_end_s": 190.5 } ], "evidence": "At 0.000 seconds the masked man's right hand is raised above the table; by 1.000 seconds it has been placed onto the woman's outstretched hand.", "mask_relation": "The action subject matches the supplied selected-person mask.", "source_subject_refs": "E1" }, { "id": "A2", "mentions": [ { "text": "speaking", "ranges": [ [ 64, 72 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "type": "simple", "subject_refs": [ "E1" ], "object_refs": [], "confidence": "high", "temporal_status": "localized", "intervals": [ { "start_s": 2.0, "end_s": 23.0, "source_start_s": 190.5, "source_end_s": 211.5 } ], "evidence": "Across the sampled frames from 2.000 through 22.000 seconds, the masked man repeatedly has changing open-mouth configurations while addressing the woman.", "mask_relation": "The action subject matches the supplied selected-person mask.", "source_subject_refs": "E1" }, { "id": "A3", "mentions": [ { "text": "makes various gestures", "ranges": [ [ 77, 99 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "type": "light_verb", "subject_refs": [ "E1" ], "object_refs": [], "confidence": "high", "temporal_status": "localized", "intervals": [ { "start_s": 4.0, "end_s": 19.0, "source_start_s": 192.5, "source_end_s": 207.5 } ], "evidence": "From 4.000 through 18.000 seconds, the masked man repeatedly raises, extends, rotates, and brings together his hands in changing explanatory gestures.", "mask_relation": "The action subject matches the supplied selected-person mask.", "source_subject_refs": "E1" }, { "id": "A4", "mentions": [ { "text": "resting", "ranges": [ [ 146, 153 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "type": "simple", "subject_refs": [ "E1" ], "object_refs": [ "E5" ], "confidence": "medium", "temporal_status": "localized", "intervals": [ { "start_s": 19.0, "end_s": 24.180000000000007, "source_start_s": 207.5, "source_end_s": 212.68 } ], "evidence": "From 19.000 seconds to the clip boundary, the masked man stops the larger gestures and keeps his hands settled at the table surface, with one hand resting over the woman's hand.", "mask_relation": "The action subject matches the supplied selected-person mask.", "source_subject_refs": "E1" } ], "notes": [ "Only the man has a supplied source mask. The woman, body-part groups, and table are text-groundable but have no independent supplied masks.", "\"continues\" is excluded as an aspectual predicate; the embedded action span is \"speaking\".", "\"emphasize\" expresses the communicative purpose of the gestures rather than a separately visible physical action, so it is omitted.", "\"his points\" is abstract and is not annotated as a concrete entity.", "The final resting interval ends exactly at the clip boundary of 24.180 seconds." ] }, "model_output_raw": { "caption": "The man places his right hand on the woman's hand and continues speaking. He makes various gestures with his hands to emphasize his points before resting his hands on the table.", "entity_groups": [ { "id": "E1", "entity": "man", "spans": [ { "text": "The man", "range": [ 0, 7 ] }, { "text": "his", "range": [ 15, 18 ] }, { "text": "He", "range": [ 74, 76 ] }, { "text": "his", "range": [ 105, 108 ] }, { "text": "his", "range": [ 128, 131 ] }, { "text": "his", "range": [ 154, 157 ] } ], "source_mask": { "track_id": "selected_person", "description": "The supplied binary mask consistently identifies the man in the white coat." } }, { "id": "E2", "entity": "man's right hand", "spans": [ { "text": "his right hand", "range": [ 15, 29 ] } ], "source_mask": null }, { "id": "E3", "entity": "woman", "spans": [ { "text": "the woman's", "range": [ 33, 44 ] } ], "source_mask": null }, { "id": "E4", "entity": "woman's hand", "spans": [ { "text": "the woman's hand", "range": [ 33, 49 ] } ], "source_mask": null }, { "id": "E5", "entity": "man's hands", "spans": [ { "text": "his hands", "range": [ 105, 114 ] }, { "text": "his hands", "range": [ 154, 163 ] } ], "source_mask": null }, { "id": "E6", "entity": "table", "spans": [ { "text": "the table", "range": [ 167, 176 ] } ], "source_mask": null } ], "actions": [ { "id": "A1", "mentions": [ { "text": "places", "occurrence": 1, "ranges": [ [ 8, 14 ] ], "particle": { "text": null, "occurrence": 1 } } ], "type": "simple", "subject_entity_group_id": "E1", "object_entity_group_ids": [ "E4" ], "body_part_entity_group_ids": [ "E2" ], "confidence": "high", "mask_relation": "The action subject matches the supplied selected-person mask.", "temporal_status": "localized", "intervals": [ { "start_s": 0.0, "end_s": 2.0, "source_start_s": 188.5, "source_end_s": 190.5 } ], "evidence": "At 0.000 seconds the masked man's right hand is raised above the table; by 1.000 seconds it has been placed onto the woman's outstretched hand." }, { "id": "A2", "mentions": [ { "text": "speaking", "occurrence": 1, "ranges": [ [ 64, 72 ] ], "particle": { "text": null, "occurrence": 1 } } ], "type": "simple", "subject_entity_group_id": "E1", "object_entity_group_ids": [], "confidence": "high", "mask_relation": "The action subject matches the supplied selected-person mask.", "temporal_status": "localized", "intervals": [ { "start_s": 2.0, "end_s": 23.0, "source_start_s": 190.5, "source_end_s": 211.5 } ], "evidence": "Across the sampled frames from 2.000 through 22.000 seconds, the masked man repeatedly has changing open-mouth configurations while addressing the woman." }, { "id": "A3", "mentions": [ { "text": "makes various gestures", "occurrence": 1, "ranges": [ [ 77, 99 ] ], "particle": { "text": null, "occurrence": 1 } } ], "type": "light_verb", "subject_entity_group_id": "E1", "object_entity_group_ids": [], "body_part_entity_group_ids": [ "E5" ], "confidence": "high", "mask_relation": "The action subject matches the supplied selected-person mask.", "temporal_status": "localized", "intervals": [ { "start_s": 4.0, "end_s": 19.0, "source_start_s": 192.5, "source_end_s": 207.5 } ], "evidence": "From 4.000 through 18.000 seconds, the masked man repeatedly raises, extends, rotates, and brings together his hands in changing explanatory gestures." }, { "id": "A4", "mentions": [ { "text": "resting", "occurrence": 1, "ranges": [ [ 146, 153 ] ], "particle": { "text": null, "occurrence": 1 } } ], "type": "simple", "subject_entity_group_id": "E1", "object_entity_group_ids": [ "E5" ], "location_entity_group_ids": [ "E6" ], "confidence": "medium", "mask_relation": "The action subject matches the supplied selected-person mask.", "temporal_status": "localized", "intervals": [ { "start_s": 19.0, "end_s": 24.18, "source_start_s": 207.5, "source_end_s": 212.68 } ], "evidence": "From 19.000 seconds to the clip boundary, the masked man stops the larger gestures and keeps his hands settled at the table surface, with one hand resting over the woman's hand." } ], "notes": [ "Only the man has a supplied source mask. The woman, body-part groups, and table are text-groundable but have no independent supplied masks.", "\"continues\" is excluded as an aspectual predicate; the embedded action span is \"speaking\".", "\"emphasize\" expresses the communicative purpose of the gestures rather than a separately visible physical action, so it is omitted.", "\"his points\" is abstract and is not annotated as a concrete entity.", "The final resting interval ends exactly at the clip boundary of 24.180 seconds." ] }, "normalization_repairs": 0 }, { "source_video_id": "v_50MmJszF2XM", "caption_count_category": 11, "clip": { "caption_index": 1, "caption": "A young girl with her hair in a ponytail, wearing a black t-shirt and purple shorts, stands still in profile. Her arms are down by her sides. An adult stands behind her, placing hands on her hips and back to adjust her posture.", "source_interval_sec": [ 5.21, 21.21 ], "loom_clip_id": "v_50MmJszF2XM_156_636", "clip_frame_interval_4fps": [ 156, 636 ], "sample_count_1fps": 16, "sampled_frames": [ { "frame": "00000", "timestamp_sec": 5.21, "offset_sec": 0, "rgb_path": "assets/cat_11/v_50MmJszF2XM_156_636/0000s_00000_rgb.jpg", "mask_path": "assets/cat_11/v_50MmJszF2XM_156_636/0000s_00000_mask.png", "mask_status": "available" }, { "frame": "00028", "timestamp_sec": 6.21, "offset_sec": 1, "rgb_path": "assets/cat_11/v_50MmJszF2XM_156_636/0001s_00028_rgb.jpg", "mask_path": "assets/cat_11/v_50MmJszF2XM_156_636/0001s_00028_mask.png", "mask_status": "available" }, { "frame": "00060", "timestamp_sec": 7.21, "offset_sec": 2, "rgb_path": "assets/cat_11/v_50MmJszF2XM_156_636/0002s_00060_rgb.jpg", "mask_path": "assets/cat_11/v_50MmJszF2XM_156_636/0002s_00060_mask.png", "mask_status": "available" }, { "frame": "00088", "timestamp_sec": 8.21, "offset_sec": 3, "rgb_path": "assets/cat_11/v_50MmJszF2XM_156_636/0003s_00088_rgb.jpg", "mask_path": "assets/cat_11/v_50MmJszF2XM_156_636/0003s_00088_mask.png", "mask_status": "available" }, { "frame": "00120", "timestamp_sec": 9.21, "offset_sec": 4, "rgb_path": "assets/cat_11/v_50MmJszF2XM_156_636/0004s_00120_rgb.jpg", "mask_path": "assets/cat_11/v_50MmJszF2XM_156_636/0004s_00120_mask.png", "mask_status": "available" }, { "frame": "00148", "timestamp_sec": 10.21, "offset_sec": 5, "rgb_path": "assets/cat_11/v_50MmJszF2XM_156_636/0005s_00148_rgb.jpg", "mask_path": "assets/cat_11/v_50MmJszF2XM_156_636/0005s_00148_mask.png", "mask_status": "available" }, { "frame": "00180", "timestamp_sec": 11.21, "offset_sec": 6, "rgb_path": "assets/cat_11/v_50MmJszF2XM_156_636/0006s_00180_rgb.jpg", "mask_path": "assets/cat_11/v_50MmJszF2XM_156_636/0006s_00180_mask.png", "mask_status": "available" }, { "frame": "00208", "timestamp_sec": 12.21, "offset_sec": 7, "rgb_path": "assets/cat_11/v_50MmJszF2XM_156_636/0007s_00208_rgb.jpg", "mask_path": "assets/cat_11/v_50MmJszF2XM_156_636/0007s_00208_mask.png", "mask_status": "available" }, { "frame": "00240", "timestamp_sec": 13.21, "offset_sec": 8, "rgb_path": "assets/cat_11/v_50MmJszF2XM_156_636/0008s_00240_rgb.jpg", "mask_path": "assets/cat_11/v_50MmJszF2XM_156_636/0008s_00240_mask.png", "mask_status": "available" }, { "frame": "00268", "timestamp_sec": 14.21, "offset_sec": 9, "rgb_path": "assets/cat_11/v_50MmJszF2XM_156_636/0009s_00268_rgb.jpg", "mask_path": "assets/cat_11/v_50MmJszF2XM_156_636/0009s_00268_mask.png", "mask_status": "available" }, { "frame": "00300", "timestamp_sec": 15.21, "offset_sec": 10, "rgb_path": "assets/cat_11/v_50MmJszF2XM_156_636/0010s_00300_rgb.jpg", "mask_path": "assets/cat_11/v_50MmJszF2XM_156_636/0010s_00300_mask.png", "mask_status": "available" }, { "frame": "00328", "timestamp_sec": 16.21, "offset_sec": 11, "rgb_path": "assets/cat_11/v_50MmJszF2XM_156_636/0011s_00328_rgb.jpg", "mask_path": "assets/cat_11/v_50MmJszF2XM_156_636/0011s_00328_mask.png", "mask_status": "available" }, { "frame": "00360", "timestamp_sec": 17.21, "offset_sec": 12, "rgb_path": "assets/cat_11/v_50MmJszF2XM_156_636/0012s_00360_rgb.jpg", "mask_path": "assets/cat_11/v_50MmJszF2XM_156_636/0012s_00360_mask.png", "mask_status": "available" }, { "frame": "00388", "timestamp_sec": 18.21, "offset_sec": 13, "rgb_path": "assets/cat_11/v_50MmJszF2XM_156_636/0013s_00388_rgb.jpg", "mask_path": "assets/cat_11/v_50MmJszF2XM_156_636/0013s_00388_mask.png", "mask_status": "available" }, { "frame": "00420", "timestamp_sec": 19.21, "offset_sec": 14, "rgb_path": "assets/cat_11/v_50MmJszF2XM_156_636/0014s_00420_rgb.jpg", "mask_path": "assets/cat_11/v_50MmJszF2XM_156_636/0014s_00420_mask.png", "mask_status": "available" }, { "frame": "00448", "timestamp_sec": 20.21, "offset_sec": 15, "rgb_path": "assets/cat_11/v_50MmJszF2XM_156_636/0015s_00448_rgb.jpg", "mask_path": "assets/cat_11/v_50MmJszF2XM_156_636/0015s_00448_mask.png", "mask_status": "available" } ], "obj_id": [ 0 ], "anno_id": [ 723 ], "fps_inferred_from_clip_interval": 30.0 }, "usage": { "input_tokens": 51549, "input_tokens_details": { "cache_write_tokens": 51546, "cached_tokens": 0 }, "output_tokens": 4035, "output_tokens_details": { "reasoning_tokens": 2703 }, "total_tokens": 55584 }, "model": "gpt-5.6-sol", "effort": "medium", "custom_id": "cat11_cap01_v_50MmJszF2XM_156_636", "annotation": { "caption": "A young girl with her hair in a ponytail, wearing a black t-shirt and purple shorts, stands still in profile. Her arms are down by her sides. An adult stands behind her, placing hands on her hips and back to adjust her posture.", "entity_groups": [ { "id": "girl", "label": "young girl", "spans": [ { "text": "A young girl", "range": [ 0, 12 ] }, { "text": "her", "range": [ 18, 21 ] }, { "text": "Her", "range": [ 110, 113 ] }, { "text": "her", "range": [ 131, 134 ] }, { "text": "her", "range": [ 165, 168 ] }, { "text": "her", "range": [ 187, 190 ] }, { "text": "her", "range": [ 215, 218 ] } ], "source_mask_link": "selected_person" }, { "id": "hair", "label": "hair", "spans": [ { "text": "her hair", "range": [ 18, 26 ] } ], "source_mask_link": null }, { "id": "ponytail", "label": "ponytail", "spans": [ { "text": "a ponytail", "range": [ 30, 40 ] } ], "source_mask_link": null }, { "id": "t_shirt", "label": "t-shirt", "spans": [ { "text": "a black t-shirt", "range": [ 50, 65 ] } ], "source_mask_link": null }, { "id": "shorts", "label": "shorts", "spans": [ { "text": "purple shorts", "range": [ 70, 83 ] } ], "source_mask_link": null }, { "id": "arms", "label": "arms", "spans": [ { "text": "Her arms", "range": [ 110, 118 ] } ], "source_mask_link": null }, { "id": "sides", "label": "sides", "spans": [ { "text": "her sides", "range": [ 131, 140 ] } ], "source_mask_link": null }, { "id": "adult", "label": "adult", "spans": [ { "text": "An adult", "range": [ 142, 150 ] } ], "source_mask_link": null }, { "id": "adult_hands", "label": "hands", "spans": [ { "text": "hands", "range": [ 178, 183 ] } ], "source_mask_link": null }, { "id": "girl_hips", "label": "hips", "spans": [ { "text": "her hips", "range": [ 187, 195 ] } ], "source_mask_link": null }, { "id": "girl_back", "label": "back", "spans": [ { "text": "back", "range": [ 200, 204 ] } ], "source_mask_link": null } ], "actions": [ { "id": "action_wearing", "mentions": [ { "text": "wearing", "ranges": [ [ 42, 49 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "type": "simple", "subject_refs": [ "girl" ], "object_refs": [ "t_shirt", "shorts" ], "confidence": "high", "temporal_status": "supported", "intervals": [ { "start_s": 8.0, "end_s": 16.0, "source_start_s": 13.21, "source_end_s": 21.21 } ], "evidence": "At sampled times 8.0–15.0 seconds, the masked girl is visibly wearing the black shirt and purple shorts; support continues through the final sampled frame.", "mask_relation": "The action subject matches the supplied selected-person mask.", "source_subject_refs": "girl" }, { "id": "action_standing", "mentions": [ { "text": "stands", "ranges": [ [ 85, 91 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "type": "posture", "subject_refs": [ "girl" ], "object_refs": [], "confidence": "high", "temporal_status": "supported", "intervals": [ { "start_s": 1.0, "end_s": 16.0, "source_start_s": 6.21, "source_end_s": 21.21 } ], "evidence": "The selected masked girl is upright and standing in profile in every usable sampled frame from 1.0 through 15.0 seconds.", "mask_relation": "The action subject matches the supplied selected-person mask.", "source_subject_refs": "girl" } ], "notes": [ "The supplied masklet identifies the young girl, not the adult behind her.", "The initial 0.0-second RGB frame is black and does not provide visual action support.", "The adult predicates \"stands\", \"placing\", and \"adjust\" were omitted because the adult has no corresponding supplied mask.", "No separate masks were supplied for clothing, hair, body parts, or the adult." ] }, "model_output_raw": { "caption": "A young girl with her hair in a ponytail, wearing a black t-shirt and purple shorts, stands still in profile. Her arms are down by her sides. An adult stands behind her, placing hands on her hips and back to adjust her posture.", "entity_groups": [ { "id": "girl", "label": "young girl", "spans": [ { "text": "A young girl", "range": [ 0, 12 ] }, { "text": "her", "range": [ 18, 21 ] }, { "text": "Her", "range": [ 110, 113 ] }, { "text": "her", "range": [ 131, 134 ] }, { "text": "her", "range": [ 165, 168 ] }, { "text": "her", "range": [ 187, 190 ] }, { "text": "her", "range": [ 215, 218 ] } ], "source_mask": "selected_person_masklet" }, { "id": "hair", "label": "hair", "spans": [ { "text": "her hair", "range": [ 18, 26 ] } ], "source_mask": null }, { "id": "ponytail", "label": "ponytail", "spans": [ { "text": "a ponytail", "range": [ 30, 40 ] } ], "source_mask": null }, { "id": "t_shirt", "label": "t-shirt", "spans": [ { "text": "a black t-shirt", "range": [ 50, 65 ] } ], "source_mask": null }, { "id": "shorts", "label": "shorts", "spans": [ { "text": "purple shorts", "range": [ 70, 83 ] } ], "source_mask": null }, { "id": "arms", "label": "arms", "spans": [ { "text": "Her arms", "range": [ 110, 118 ] } ], "source_mask": null }, { "id": "sides", "label": "sides", "spans": [ { "text": "her sides", "range": [ 131, 140 ] } ], "source_mask": null }, { "id": "adult", "label": "adult", "spans": [ { "text": "An adult", "range": [ 142, 150 ] } ], "source_mask": null }, { "id": "adult_hands", "label": "hands", "spans": [ { "text": "hands", "range": [ 178, 183 ] } ], "source_mask": null }, { "id": "girl_hips", "label": "hips", "spans": [ { "text": "her hips", "range": [ 187, 195 ] } ], "source_mask": null }, { "id": "girl_back", "label": "back", "spans": [ { "text": "back", "range": [ 200, 204 ] } ], "source_mask": null } ], "actions": [ { "id": "action_wearing", "mentions": [ { "text": "wearing", "ranges": [ [ 42, 49 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "type": "simple", "subject_ref": "girl", "object_refs": [ "t_shirt", "shorts" ], "confidence": "high", "temporal_status": "supported", "intervals": [ { "start_s": 8.0, "end_s": 16.0 } ], "mask_relation": "The action subject matches the supplied selected-person mask.", "evidence": "At sampled times 8.0–15.0 seconds, the masked girl is visibly wearing the black shirt and purple shorts; support continues through the final sampled frame." }, { "id": "action_standing", "mentions": [ { "text": "stands", "ranges": [ [ 85, 91 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "type": "posture", "subject_ref": "girl", "object_refs": [], "confidence": "high", "temporal_status": "supported", "intervals": [ { "start_s": 1.0, "end_s": 16.0 } ], "mask_relation": "The action subject matches the supplied selected-person mask.", "evidence": "The selected masked girl is upright and standing in profile in every usable sampled frame from 1.0 through 15.0 seconds." } ], "notes": [ "The supplied masklet identifies the young girl, not the adult behind her.", "The initial 0.0-second RGB frame is black and does not provide visual action support.", "The adult predicates \"stands\", \"placing\", and \"adjust\" were omitted because the adult has no corresponding supplied mask.", "No separate masks were supplied for clothing, hair, body parts, or the adult." ] }, "normalization_repairs": 0 }, { "source_video_id": "v_50MmJszF2XM", "caption_count_category": 11, "clip": { "caption_index": 2, "caption": "The girl swings her arms forward and up, bends her knees into a squat, and then swings her arms down and back. She repeats the upward arm swing, arching her back while being supported at the waist by the adult.", "source_interval_sec": [ 21.21, 33.83 ], "loom_clip_id": "v_50MmJszF2XM_636_1014", "clip_frame_interval_4fps": [ 636, 1014 ], "sample_count_1fps": 13, "sampled_frames": [ { "frame": "00000", "timestamp_sec": 21.21, "offset_sec": 0, "rgb_path": "assets/cat_11/v_50MmJszF2XM_636_1014/0000s_00000_rgb.jpg", "mask_path": "assets/cat_11/v_50MmJszF2XM_636_1014/0000s_00000_mask.png", "mask_status": "available" }, { "frame": "00028", "timestamp_sec": 22.21, "offset_sec": 1, "rgb_path": "assets/cat_11/v_50MmJszF2XM_636_1014/0001s_00028_rgb.jpg", "mask_path": "assets/cat_11/v_50MmJszF2XM_636_1014/0001s_00028_mask.png", "mask_status": "available" }, { "frame": "00060", "timestamp_sec": 23.21, "offset_sec": 2, "rgb_path": "assets/cat_11/v_50MmJszF2XM_636_1014/0002s_00060_rgb.jpg", "mask_path": "assets/cat_11/v_50MmJszF2XM_636_1014/0002s_00060_mask.png", "mask_status": "available" }, { "frame": "00088", "timestamp_sec": 24.21, "offset_sec": 3, "rgb_path": "assets/cat_11/v_50MmJszF2XM_636_1014/0003s_00088_rgb.jpg", "mask_path": "assets/cat_11/v_50MmJszF2XM_636_1014/0003s_00088_mask.png", "mask_status": "available" }, { "frame": "00120", "timestamp_sec": 25.21, "offset_sec": 4, "rgb_path": "assets/cat_11/v_50MmJszF2XM_636_1014/0004s_00120_rgb.jpg", "mask_path": "assets/cat_11/v_50MmJszF2XM_636_1014/0004s_00120_mask.png", "mask_status": "available" }, { "frame": "00148", "timestamp_sec": 26.21, "offset_sec": 5, "rgb_path": "assets/cat_11/v_50MmJszF2XM_636_1014/0005s_00148_rgb.jpg", "mask_path": "assets/cat_11/v_50MmJszF2XM_636_1014/0005s_00148_mask.png", "mask_status": "available" }, { "frame": "00180", "timestamp_sec": 27.21, "offset_sec": 6, "rgb_path": "assets/cat_11/v_50MmJszF2XM_636_1014/0006s_00180_rgb.jpg", "mask_path": "assets/cat_11/v_50MmJszF2XM_636_1014/0006s_00180_mask.png", "mask_status": "available" }, { "frame": "00208", "timestamp_sec": 28.21, "offset_sec": 7, "rgb_path": "assets/cat_11/v_50MmJszF2XM_636_1014/0007s_00208_rgb.jpg", "mask_path": "assets/cat_11/v_50MmJszF2XM_636_1014/0007s_00208_mask.png", "mask_status": "available" }, { "frame": "00240", "timestamp_sec": 29.21, "offset_sec": 8, "rgb_path": "assets/cat_11/v_50MmJszF2XM_636_1014/0008s_00240_rgb.jpg", "mask_path": "assets/cat_11/v_50MmJszF2XM_636_1014/0008s_00240_mask.png", "mask_status": "available" }, { "frame": "00268", "timestamp_sec": 30.21, "offset_sec": 9, "rgb_path": "assets/cat_11/v_50MmJszF2XM_636_1014/0009s_00268_rgb.jpg", "mask_path": "assets/cat_11/v_50MmJszF2XM_636_1014/0009s_00268_mask.png", "mask_status": "available" }, { "frame": "00300", "timestamp_sec": 31.21, "offset_sec": 10, "rgb_path": "assets/cat_11/v_50MmJszF2XM_636_1014/0010s_00300_rgb.jpg", "mask_path": "assets/cat_11/v_50MmJszF2XM_636_1014/0010s_00300_mask.png", "mask_status": "available" }, { "frame": "00328", "timestamp_sec": 32.21, "offset_sec": 11, "rgb_path": "assets/cat_11/v_50MmJszF2XM_636_1014/0011s_00328_rgb.jpg", "mask_path": "assets/cat_11/v_50MmJszF2XM_636_1014/0011s_00328_mask.png", "mask_status": "available" }, { "frame": "00360", "timestamp_sec": 33.21, "offset_sec": 12, "rgb_path": "assets/cat_11/v_50MmJszF2XM_636_1014/0012s_00360_rgb.jpg", "mask_path": "assets/cat_11/v_50MmJszF2XM_636_1014/0012s_00360_mask.png", "mask_status": "available" } ], "obj_id": [ 0 ], "anno_id": [ 731 ], "fps_inferred_from_clip_interval": 29.9525 }, "usage": { "input_tokens": 44755, "input_tokens_details": { "cache_write_tokens": 44752, "cached_tokens": 0 }, "output_tokens": 5473, "output_tokens_details": { "reasoning_tokens": 3624 }, "total_tokens": 50228 }, "model": "gpt-5.6-sol", "effort": "medium", "custom_id": "cat11_cap02_v_50MmJszF2XM_636_1014", "annotation": { "caption": "The girl swings her arms forward and up, bends her knees into a squat, and then swings her arms down and back. She repeats the upward arm swing, arching her back while being supported at the waist by the adult.", "entity_groups": [ { "id": "girl", "label": "girl", "spans": [ { "text": "The girl", "range": [ 0, 8 ] }, { "text": "her", "range": [ 16, 19 ] }, { "text": "her", "range": [ 47, 50 ] }, { "text": "her", "range": [ 87, 90 ] }, { "text": "She", "range": [ 111, 114 ] }, { "text": "her", "range": [ 153, 156 ] } ], "source_mask_link": "selected_person" }, { "id": "girl_arms", "label": "arms", "spans": [ { "text": "her arms", "range": [ 16, 24 ] }, { "text": "her arms", "range": [ 87, 95 ] } ], "source_mask_link": null }, { "id": "girl_knees", "label": "knees", "spans": [ { "text": "her knees", "range": [ 47, 56 ] } ], "source_mask_link": null }, { "id": "girl_back", "label": "back", "spans": [ { "text": "her back", "range": [ 153, 161 ] } ], "source_mask_link": null }, { "id": "girl_waist", "label": "waist", "spans": [ { "text": "the waist", "range": [ 187, 196 ] } ], "source_mask_link": null }, { "id": "adult", "label": "adult", "spans": [ { "text": "the adult", "range": [ 200, 209 ] } ], "source_mask_link": null } ], "actions": [ { "id": "action_1", "mentions": [ { "text": "swings", "ranges": [ [ 9, 15 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "type": "simple", "subject_refs": [ "girl" ], "object_refs": [ "girl_arms" ], "confidence": "high", "temporal_status": "localized", "intervals": [ { "start_s": 0.0, "end_s": 6.0, "source_start_s": 21.21, "source_end_s": 27.21 } ], "evidence": "Across samples 0–5, the masked girl moves her arms from a lowered position to forward and raised positions.", "mask_relation": "The action subject matches the supplied selected-person mask.", "source_subject_refs": [ "girl" ] }, { "id": "action_2", "mentions": [ { "text": "bends", "ranges": [ [ 41, 46 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "type": "simple", "subject_refs": [ "girl" ], "object_refs": [ "girl_knees" ], "confidence": "high", "temporal_status": "localized", "intervals": [ { "start_s": 0.0, "end_s": 3.0, "source_start_s": 21.21, "source_end_s": 24.21 } ], "evidence": "The masked girl's knees change from relatively extended at sample 0 to visibly flexed in a squat by samples 1–2.", "mask_relation": "The action subject matches the supplied selected-person mask.", "source_subject_refs": [ "girl" ] }, { "id": "action_3", "mentions": [ { "text": "swings", "ranges": [ [ 80, 86 ] ], "occurrence": 2, "particle": { "text": null, "occurrence": 1 } } ], "type": "simple", "subject_refs": [ "girl" ], "object_refs": [ "girl_arms" ], "confidence": "high", "temporal_status": "localized", "intervals": [ { "start_s": 6.0, "end_s": 12.0, "source_start_s": 27.21, "source_end_s": 33.21 } ], "evidence": "Samples 6–11 show the masked girl's arms descending from an extended position toward low and rearward positions while she remains crouched.", "mask_relation": "The action subject matches the supplied selected-person mask.", "source_subject_refs": [ "girl" ] }, { "id": "action_4", "mentions": [ { "text": "repeats the upward arm swing", "ranges": [ [ 115, 143 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "type": "light_verb", "subject_refs": [ "girl" ], "object_refs": [ "girl_arms" ], "confidence": "high", "temporal_status": "localized", "intervals": [ { "start_s": 12.0, "end_s": 12.619999999999997, "source_start_s": 33.21, "source_end_s": 33.83 } ], "evidence": "At sample 12, the masked girl has swung her arms upward over her head, and the supported motion reaches the clip boundary.", "mask_relation": "The action subject matches the supplied selected-person mask.", "source_subject_refs": [ "girl" ] }, { "id": "action_5", "mentions": [ { "text": "arching", "ranges": [ [ 145, 152 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "type": "simple", "subject_refs": [ "girl" ], "object_refs": [ "girl_back" ], "confidence": "medium", "temporal_status": "localized", "intervals": [ { "start_s": 12.0, "end_s": 12.619999999999997, "source_start_s": 33.21, "source_end_s": 33.83 } ], "evidence": "Sample 12 shows the masked girl upright with her arms overhead and her torso/back curved while the adult supports her waist.", "mask_relation": "The action subject matches the supplied selected-person mask.", "source_subject_refs": [ "girl" ] }, { "id": "action_6", "mentions": [ { "text": "being supported", "ranges": [ [ 168, 183 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "type": "passive", "subject_refs": [ "girl" ], "object_refs": [], "confidence": "high", "temporal_status": "localized", "intervals": [ { "start_s": 0.0, "end_s": 12.619999999999997, "source_start_s": 21.21, "source_end_s": 33.83 } ], "evidence": "In every sampled frame, the adult stands behind the masked girl with hands at or near her waist, visibly supporting her through the clip boundary.", "mask_relation": "The passive action's grammatical subject matches the supplied selected-person mask.", "source_subject_refs": [ "girl" ] } ], "notes": [ "The supplied mask consistently identifies the girl in the foreground, not the adult behind her.", "No separate masks were supplied for the adult or for the girl's arms, knees, back, or waist; those entity groups therefore have null source-mask links.", "The nominal phrase \"a squat\" names the result of bending into a posture but is not separately extracted as an action mention because it is not expressed as an eligible predicate or qualifying light-verb construction." ] }, "model_output_raw": { "caption": "The girl swings her arms forward and up, bends her knees into a squat, and then swings her arms down and back. She repeats the upward arm swing, arching her back while being supported at the waist by the adult.", "entity_groups": [ { "id": "girl", "label": "girl", "spans": [ { "text": "The girl", "range": [ 0, 8 ] }, { "text": "her", "range": [ 16, 19 ] }, { "text": "her", "range": [ 47, 50 ] }, { "text": "her", "range": [ 87, 90 ] }, { "text": "She", "range": [ 111, 114 ] }, { "text": "her", "range": [ 153, 156 ] } ], "source_mask": "selected_person_masklet" }, { "id": "girl_arms", "label": "arms", "spans": [ { "text": "her arms", "range": [ 16, 24 ] }, { "text": "her arms", "range": [ 87, 95 ] } ], "source_mask": null }, { "id": "girl_knees", "label": "knees", "spans": [ { "text": "her knees", "range": [ 47, 56 ] } ], "source_mask": null }, { "id": "girl_back", "label": "back", "spans": [ { "text": "her back", "range": [ 153, 161 ] } ], "source_mask": null }, { "id": "girl_waist", "label": "waist", "spans": [ { "text": "the waist", "range": [ 187, 196 ] } ], "source_mask": null }, { "id": "adult", "label": "adult", "spans": [ { "text": "the adult", "range": [ 200, 209 ] } ], "source_mask": null } ], "actions": [ { "id": "action_1", "mentions": [ { "text": "swings", "ranges": [ [ 9, 15 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "type": "simple", "subject_entity_ids": [ "girl" ], "object_entity_ids": [ "girl_arms" ], "confidence": "high", "temporal_status": "localized", "clip_relative_intervals_s": [ [ 0.0, 6.0 ] ], "source_video_intervals_s": [ [ 21.21, 27.21 ] ], "evidence": "Across samples 0–5, the masked girl moves her arms from a lowered position to forward and raised positions.", "mask_relation": "The action subject matches the supplied selected-person mask." }, { "id": "action_2", "mentions": [ { "text": "bends", "ranges": [ [ 41, 46 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "type": "simple", "subject_entity_ids": [ "girl" ], "object_entity_ids": [ "girl_knees" ], "confidence": "high", "temporal_status": "localized", "clip_relative_intervals_s": [ [ 0.0, 3.0 ] ], "source_video_intervals_s": [ [ 21.21, 24.21 ] ], "evidence": "The masked girl's knees change from relatively extended at sample 0 to visibly flexed in a squat by samples 1–2.", "mask_relation": "The action subject matches the supplied selected-person mask." }, { "id": "action_3", "mentions": [ { "text": "swings", "ranges": [ [ 80, 86 ] ], "occurrence": 2, "particle": { "text": null, "occurrence": 1 } } ], "type": "simple", "subject_entity_ids": [ "girl" ], "object_entity_ids": [ "girl_arms" ], "confidence": "high", "temporal_status": "localized", "clip_relative_intervals_s": [ [ 6.0, 12.0 ] ], "source_video_intervals_s": [ [ 27.21, 33.21 ] ], "evidence": "Samples 6–11 show the masked girl's arms descending from an extended position toward low and rearward positions while she remains crouched.", "mask_relation": "The action subject matches the supplied selected-person mask." }, { "id": "action_4", "mentions": [ { "text": "repeats the upward arm swing", "ranges": [ [ 115, 143 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "type": "light_verb", "subject_entity_ids": [ "girl" ], "object_entity_ids": [ "girl_arms" ], "confidence": "high", "temporal_status": "localized", "clip_relative_intervals_s": [ [ 12.0, 12.62 ] ], "source_video_intervals_s": [ [ 33.21, 33.83 ] ], "evidence": "At sample 12, the masked girl has swung her arms upward over her head, and the supported motion reaches the clip boundary.", "mask_relation": "The action subject matches the supplied selected-person mask." }, { "id": "action_5", "mentions": [ { "text": "arching", "ranges": [ [ 145, 152 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "type": "simple", "subject_entity_ids": [ "girl" ], "object_entity_ids": [ "girl_back" ], "confidence": "medium", "temporal_status": "localized", "clip_relative_intervals_s": [ [ 12.0, 12.62 ] ], "source_video_intervals_s": [ [ 33.21, 33.83 ] ], "evidence": "Sample 12 shows the masked girl upright with her arms overhead and her torso/back curved while the adult supports her waist.", "mask_relation": "The action subject matches the supplied selected-person mask." }, { "id": "action_6", "mentions": [ { "text": "being supported", "ranges": [ [ 168, 183 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "type": "passive", "subject_entity_ids": [ "girl" ], "object_entity_ids": [], "agent_entity_ids": [ "adult" ], "confidence": "high", "temporal_status": "localized", "clip_relative_intervals_s": [ [ 0.0, 12.62 ] ], "source_video_intervals_s": [ [ 21.21, 33.83 ] ], "evidence": "In every sampled frame, the adult stands behind the masked girl with hands at or near her waist, visibly supporting her through the clip boundary.", "mask_relation": "The passive action's grammatical subject matches the supplied selected-person mask." } ], "notes": [ "The supplied mask consistently identifies the girl in the foreground, not the adult behind her.", "No separate masks were supplied for the adult or for the girl's arms, knees, back, or waist; those entity groups therefore have null source-mask links.", "The nominal phrase \"a squat\" names the result of bending into a posture but is not separately extracted as an action mention because it is not expressed as an eligible predicate or qualifying light-verb construction." ] }, "normalization_repairs": 0 }, { "source_video_id": "v_50MmJszF2XM", "caption_count_category": 11, "clip": { "caption_index": 3, "caption": "On a long blue tumbling mat, the girl performs a complete back handspring. She jumps backward onto her hands and springs back to her feet. She finishes by standing with her arms raised straight above her head.", "source_interval_sec": [ 33.83, 43.34 ], "loom_clip_id": "v_50MmJszF2XM_1014_1299", "clip_frame_interval_4fps": [ 1014, 1299 ], "sample_count_1fps": 10, "sampled_frames": [ { "frame": "00000", "timestamp_sec": 33.83, "offset_sec": 0, "rgb_path": "assets/cat_11/v_50MmJszF2XM_1014_1299/0000s_00000_rgb.jpg", "mask_path": "assets/cat_11/v_50MmJszF2XM_1014_1299/0000s_00000_mask.png", "mask_status": "available" }, { "frame": "00028", "timestamp_sec": 34.83, "offset_sec": 1, "rgb_path": "assets/cat_11/v_50MmJszF2XM_1014_1299/0001s_00028_rgb.jpg", "mask_path": "assets/cat_11/v_50MmJszF2XM_1014_1299/0001s_00028_mask.png", "mask_status": "available" }, { "frame": "00060", "timestamp_sec": 35.83, "offset_sec": 2, "rgb_path": "assets/cat_11/v_50MmJszF2XM_1014_1299/0002s_00060_rgb.jpg", "mask_path": "assets/cat_11/v_50MmJszF2XM_1014_1299/0002s_00060_mask.png", "mask_status": "available" }, { "frame": "00088", "timestamp_sec": 36.83, "offset_sec": 3, "rgb_path": "assets/cat_11/v_50MmJszF2XM_1014_1299/0003s_00088_rgb.jpg", "mask_path": "assets/cat_11/v_50MmJszF2XM_1014_1299/0003s_00088_mask.png", "mask_status": "available" }, { "frame": "00120", "timestamp_sec": 37.83, "offset_sec": 4, "rgb_path": "assets/cat_11/v_50MmJszF2XM_1014_1299/0004s_00120_rgb.jpg", "mask_path": "assets/cat_11/v_50MmJszF2XM_1014_1299/0004s_00120_mask.png", "mask_status": "available" }, { "frame": "00148", "timestamp_sec": 38.83, "offset_sec": 5, "rgb_path": "assets/cat_11/v_50MmJszF2XM_1014_1299/0005s_00148_rgb.jpg", "mask_path": "assets/cat_11/v_50MmJszF2XM_1014_1299/0005s_00148_mask.png", "mask_status": "available" }, { "frame": "00180", "timestamp_sec": 39.83, "offset_sec": 6, "rgb_path": "assets/cat_11/v_50MmJszF2XM_1014_1299/0006s_00180_rgb.jpg", "mask_path": "assets/cat_11/v_50MmJszF2XM_1014_1299/0006s_00180_mask.png", "mask_status": "available" }, { "frame": "00208", "timestamp_sec": 40.83, "offset_sec": 7, "rgb_path": "assets/cat_11/v_50MmJszF2XM_1014_1299/0007s_00208_rgb.jpg", "mask_path": "assets/cat_11/v_50MmJszF2XM_1014_1299/0007s_00208_mask.png", "mask_status": "available" }, { "frame": "00240", "timestamp_sec": 41.83, "offset_sec": 8, "rgb_path": "assets/cat_11/v_50MmJszF2XM_1014_1299/0008s_00240_rgb.jpg", "mask_path": "assets/cat_11/v_50MmJszF2XM_1014_1299/0008s_00240_mask.png", "mask_status": "available" }, { "frame": "00268", "timestamp_sec": 42.83, "offset_sec": 9, "rgb_path": "assets/cat_11/v_50MmJszF2XM_1014_1299/0009s_00268_rgb.jpg", "mask_path": "assets/cat_11/v_50MmJszF2XM_1014_1299/0009s_00268_mask.png", "mask_status": "available" } ], "obj_id": [ 0 ], "anno_id": [ 721 ], "fps_inferred_from_clip_interval": 29.9685 }, "usage": { "input_tokens": 37963, "input_tokens_details": { "cache_write_tokens": 37960, "cached_tokens": 0 }, "output_tokens": 3890, "output_tokens_details": { "reasoning_tokens": 2347 }, "total_tokens": 41853 }, "model": "gpt-5.6-sol", "effort": "medium", "custom_id": "cat11_cap03_v_50MmJszF2XM_1014_1299", "annotation": { "caption": "On a long blue tumbling mat, the girl performs a complete back handspring. She jumps backward onto her hands and springs back to her feet. She finishes by standing with her arms raised straight above her head.", "entity_groups": [ { "id": "entity_mat", "label": "mat", "spans": [ { "text": "a long blue tumbling mat", "range": [ 3, 27 ] } ], "source_mask_link": null }, { "id": "entity_girl", "label": "person", "spans": [ { "text": "the girl", "range": [ 29, 37 ] }, { "text": "She", "range": [ 75, 78 ] }, { "text": "her", "range": [ 99, 102 ] }, { "text": "her", "range": [ 129, 132 ] }, { "text": "She", "range": [ 139, 142 ] }, { "text": "her", "range": [ 169, 172 ] }, { "text": "her", "range": [ 200, 203 ] } ], "source_mask_link": "selected_person" }, { "id": "entity_hands", "label": "hands", "spans": [ { "text": "her hands", "range": [ 99, 108 ] } ], "source_mask_link": null }, { "id": "entity_feet", "label": "feet", "spans": [ { "text": "her feet", "range": [ 129, 137 ] } ], "source_mask_link": null }, { "id": "entity_arms", "label": "arms", "spans": [ { "text": "her arms", "range": [ 169, 177 ] } ], "source_mask_link": null }, { "id": "entity_head", "label": "head", "spans": [ { "text": "her head", "range": [ 200, 208 ] } ], "source_mask_link": null } ], "actions": [ { "id": "action_handspring", "mentions": [ { "text": "performs a complete back handspring", "ranges": [ [ 38, 73 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "type": "light_verb", "subject_refs": [ "entity_girl" ], "object_refs": [], "confidence": "high", "temporal_status": "localized", "intervals": [ { "start_s": 3.0, "end_s": 7.0, "source_start_s": 36.83, "source_end_s": 40.83 } ], "evidence": "At 3.000 seconds the masked girl crouches and initiates the maneuver; at 4.000 seconds she is inverted on her hands, at 5.000 seconds she is returning toward her feet, and by 6.000 seconds she has recovered upright.", "mask_relation": "subject matches the supplied selected-person mask", "source_subject_refs": [ "entity_girl" ] }, { "id": "action_jump", "mentions": [ { "text": "jumps", "ranges": [ [ 79, 84 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "type": "simple", "subject_refs": [ "entity_girl" ], "object_refs": [], "confidence": "high", "temporal_status": "localized", "intervals": [ { "start_s": 3.0, "end_s": 5.0, "source_start_s": 36.83, "source_end_s": 38.83 } ], "evidence": "The masked girl bends to take off at 3.000 seconds and is inverted after moving backward by 4.000 seconds.", "mask_relation": "subject matches the supplied selected-person mask", "source_subject_refs": [ "entity_girl" ] }, { "id": "action_spring", "mentions": [ { "text": "springs", "ranges": [ [ 113, 120 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "type": "simple", "subject_refs": [ "entity_girl" ], "object_refs": [], "confidence": "high", "temporal_status": "localized", "intervals": [ { "start_s": 4.0, "end_s": 7.0, "source_start_s": 37.83, "source_end_s": 40.83 } ], "evidence": "From the inverted hand-supported position at 4.000 seconds, the masked girl rotates down at 5.000 seconds and returns upright by 6.000 seconds.", "mask_relation": "subject matches the supplied selected-person mask", "source_subject_refs": [ "entity_girl" ] }, { "id": "action_standing", "mentions": [ { "text": "standing", "ranges": [ [ 155, 163 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "type": "posture", "subject_refs": [ "entity_girl" ], "object_refs": [], "confidence": "high", "temporal_status": "localized", "intervals": [ { "start_s": 6.0, "end_s": 9.510000000000005, "source_start_s": 39.83, "source_end_s": 43.34 } ], "evidence": "The masked girl is upright at 6.000 seconds and remains standing with her arms overhead through the final sampled frame at 9.000 seconds.", "mask_relation": "subject matches the supplied selected-person mask", "source_subject_refs": [ "entity_girl" ] } ], "notes": [ "The supplied masklet tracks the captioned girl; the mat and individual body-part entities have no separate supplied masks.", "The overall back handspring and its explicitly described jump and spring phases are retained as distinct captioned events.", "\"finishes\" is excluded because it frames the standing action rather than expressing an independent physical action.", "\"raised\" is not annotated as a separate action: in this construction it describes the resulting posture of the arms, and the arms do not have a separate supplied instance mask." ] }, "model_output_raw": { "caption": "On a long blue tumbling mat, the girl performs a complete back handspring. She jumps backward onto her hands and springs back to her feet. She finishes by standing with her arms raised straight above her head.", "entity_groups": [ { "id": "entity_mat", "entity_type": "mat", "spans": [ { "text": "a long blue tumbling mat", "range": [ 3, 27 ] } ], "source_mask_link": null }, { "id": "entity_girl", "entity_type": "person", "spans": [ { "text": "the girl", "range": [ 29, 37 ] }, { "text": "She", "range": [ 75, 78 ] }, { "text": "her", "range": [ 99, 102 ] }, { "text": "her", "range": [ 129, 132 ] }, { "text": "She", "range": [ 139, 142 ] }, { "text": "her", "range": [ 169, 172 ] }, { "text": "her", "range": [ 200, 203 ] } ], "source_mask_link": "selected_person_mask" }, { "id": "entity_hands", "entity_type": "hands", "spans": [ { "text": "her hands", "range": [ 99, 108 ] } ], "source_mask_link": null }, { "id": "entity_feet", "entity_type": "feet", "spans": [ { "text": "her feet", "range": [ 129, 137 ] } ], "source_mask_link": null }, { "id": "entity_arms", "entity_type": "arms", "spans": [ { "text": "her arms", "range": [ 169, 177 ] } ], "source_mask_link": null }, { "id": "entity_head", "entity_type": "head", "spans": [ { "text": "her head", "range": [ 200, 208 ] } ], "source_mask_link": null } ], "actions": [ { "id": "action_handspring", "mentions": [ { "text": "performs a complete back handspring", "ranges": [ [ 38, 73 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "type": "light_verb", "subject_refs": [ "entity_girl" ], "object_refs": [], "confidence": "high", "mask_relation": "subject matches the supplied selected-person mask", "temporal_status": "localized", "intervals": [ { "clip_relative": [ 3.0, 7.0 ], "source_video": [ 36.83, 40.83 ] } ], "evidence": "At 3.000 seconds the masked girl crouches and initiates the maneuver; at 4.000 seconds she is inverted on her hands, at 5.000 seconds she is returning toward her feet, and by 6.000 seconds she has recovered upright." }, { "id": "action_jump", "mentions": [ { "text": "jumps", "ranges": [ [ 79, 84 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "type": "simple", "subject_refs": [ "entity_girl" ], "object_refs": [], "confidence": "high", "mask_relation": "subject matches the supplied selected-person mask", "temporal_status": "localized", "intervals": [ { "clip_relative": [ 3.0, 5.0 ], "source_video": [ 36.83, 38.83 ] } ], "evidence": "The masked girl bends to take off at 3.000 seconds and is inverted after moving backward by 4.000 seconds." }, { "id": "action_spring", "mentions": [ { "text": "springs", "ranges": [ [ 113, 120 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "type": "simple", "subject_refs": [ "entity_girl" ], "object_refs": [], "confidence": "high", "mask_relation": "subject matches the supplied selected-person mask", "temporal_status": "localized", "intervals": [ { "clip_relative": [ 4.0, 7.0 ], "source_video": [ 37.83, 40.83 ] } ], "evidence": "From the inverted hand-supported position at 4.000 seconds, the masked girl rotates down at 5.000 seconds and returns upright by 6.000 seconds." }, { "id": "action_standing", "mentions": [ { "text": "standing", "ranges": [ [ 155, 163 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "type": "posture", "subject_refs": [ "entity_girl" ], "object_refs": [], "confidence": "high", "mask_relation": "subject matches the supplied selected-person mask", "temporal_status": "localized", "intervals": [ { "clip_relative": [ 6.0, 9.51 ], "source_video": [ 39.83, 43.34 ] } ], "evidence": "The masked girl is upright at 6.000 seconds and remains standing with her arms overhead through the final sampled frame at 9.000 seconds." } ], "notes": [ "The supplied masklet tracks the captioned girl; the mat and individual body-part entities have no separate supplied masks.", "The overall back handspring and its explicitly described jump and spring phases are retained as distinct captioned events.", "\"finishes\" is excluded because it frames the standing action rather than expressing an independent physical action.", "\"raised\" is not annotated as a separate action: in this construction it describes the resulting posture of the arms, and the arms do not have a separate supplied instance mask." ] }, "normalization_repairs": 0 }, { "source_video_id": "v_50MmJszF2XM", "caption_count_category": 11, "clip": { "caption_index": 4, "caption": "The girl performs a back walkover, starting from a standing position. She raises her arms, arches her back until her hands touch the floor, and kicks her legs over one by one. She returns to a standing position with her arms raised.", "source_interval_sec": [ 43.34, 53.21 ], "loom_clip_id": "v_50MmJszF2XM_1299_1595", "clip_frame_interval_4fps": [ 1299, 1595 ], "sample_count_1fps": 10, "sampled_frames": [ { "frame": "00000", "timestamp_sec": 43.34, "offset_sec": 0, "rgb_path": "assets/cat_11/v_50MmJszF2XM_1299_1595/0000s_00000_rgb.jpg", "mask_path": "assets/cat_11/v_50MmJszF2XM_1299_1595/0000s_00000_mask.png", "mask_status": "available" }, { "frame": "00028", "timestamp_sec": 44.34, "offset_sec": 1, "rgb_path": "assets/cat_11/v_50MmJszF2XM_1299_1595/0001s_00028_rgb.jpg", "mask_path": "assets/cat_11/v_50MmJszF2XM_1299_1595/0001s_00028_mask.png", "mask_status": "available" }, { "frame": "00060", "timestamp_sec": 45.34, "offset_sec": 2, "rgb_path": "assets/cat_11/v_50MmJszF2XM_1299_1595/0002s_00060_rgb.jpg", "mask_path": "assets/cat_11/v_50MmJszF2XM_1299_1595/0002s_00060_mask.png", "mask_status": "available" }, { "frame": "00088", "timestamp_sec": 46.34, "offset_sec": 3, "rgb_path": "assets/cat_11/v_50MmJszF2XM_1299_1595/0003s_00088_rgb.jpg", "mask_path": "assets/cat_11/v_50MmJszF2XM_1299_1595/0003s_00088_mask.png", "mask_status": "available" }, { "frame": "00120", "timestamp_sec": 47.34, "offset_sec": 4, "rgb_path": "assets/cat_11/v_50MmJszF2XM_1299_1595/0004s_00120_rgb.jpg", "mask_path": "assets/cat_11/v_50MmJszF2XM_1299_1595/0004s_00120_mask.png", "mask_status": "available" }, { "frame": "00148", "timestamp_sec": 48.34, "offset_sec": 5, "rgb_path": "assets/cat_11/v_50MmJszF2XM_1299_1595/0005s_00148_rgb.jpg", "mask_path": "assets/cat_11/v_50MmJszF2XM_1299_1595/0005s_00148_mask.png", "mask_status": "available" }, { "frame": "00180", "timestamp_sec": 49.34, "offset_sec": 6, "rgb_path": "assets/cat_11/v_50MmJszF2XM_1299_1595/0006s_00180_rgb.jpg", "mask_path": "assets/cat_11/v_50MmJszF2XM_1299_1595/0006s_00180_mask.png", "mask_status": "available" }, { "frame": "00208", "timestamp_sec": 50.34, "offset_sec": 7, "rgb_path": "assets/cat_11/v_50MmJszF2XM_1299_1595/0007s_00208_rgb.jpg", "mask_path": "assets/cat_11/v_50MmJszF2XM_1299_1595/0007s_00208_mask.png", "mask_status": "available" }, { "frame": "00240", "timestamp_sec": 51.34, "offset_sec": 8, "rgb_path": "assets/cat_11/v_50MmJszF2XM_1299_1595/0008s_00240_rgb.jpg", "mask_path": "assets/cat_11/v_50MmJszF2XM_1299_1595/0008s_00240_mask.png", "mask_status": "available" }, { "frame": "00268", "timestamp_sec": 52.34, "offset_sec": 9, "rgb_path": "assets/cat_11/v_50MmJszF2XM_1299_1595/0009s_00268_rgb.jpg", "mask_path": "assets/cat_11/v_50MmJszF2XM_1299_1595/0009s_00268_mask.png", "mask_status": "available" } ], "obj_id": [ 0 ], "anno_id": [ 722 ], "fps_inferred_from_clip_interval": 29.9899 }, "usage": { "input_tokens": 37969, "input_tokens_details": { "cache_write_tokens": 37966, "cached_tokens": 0 }, "output_tokens": 5195, "output_tokens_details": { "reasoning_tokens": 3106 }, "total_tokens": 43164 }, "model": "gpt-5.6-sol", "effort": "medium", "custom_id": "cat11_cap04_v_50MmJszF2XM_1299_1595", "annotation": { "caption": "The girl performs a back walkover, starting from a standing position. She raises her arms, arches her back until her hands touch the floor, and kicks her legs over one by one. She returns to a standing position with her arms raised.", "entity_groups": [ { "id": "girl", "label": "girl", "spans": [ { "text": "The girl", "range": [ 0, 8 ] }, { "text": "She", "range": [ 70, 73 ] }, { "text": "her", "range": [ 81, 84 ] }, { "text": "her", "range": [ 98, 101 ] }, { "text": "her", "range": [ 113, 116 ] }, { "text": "her", "range": [ 150, 153 ] }, { "text": "She", "range": [ 176, 179 ] }, { "text": "her", "range": [ 216, 219 ] } ], "source_mask_link": "selected_person" }, { "id": "arms", "label": "arms", "spans": [ { "text": "her arms", "range": [ 81, 89 ] }, { "text": "her arms", "range": [ 216, 224 ] } ], "source_mask_link": null }, { "id": "back", "label": "back", "spans": [ { "text": "her back", "range": [ 98, 106 ] } ], "source_mask_link": null }, { "id": "hands", "label": "hands", "spans": [ { "text": "her hands", "range": [ 113, 122 ] } ], "source_mask_link": null }, { "id": "floor", "label": "floor", "spans": [ { "text": "the floor", "range": [ 129, 138 ] } ], "source_mask_link": null }, { "id": "legs", "label": "legs", "spans": [ { "text": "her legs", "range": [ 150, 158 ] }, { "text": "one", "range": [ 164, 167 ] }, { "text": "one", "range": [ 171, 174 ] } ], "source_mask_link": null } ], "actions": [ { "id": "action_1", "mentions": [ { "text": "performs a back walkover", "ranges": [ [ 9, 33 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "type": "light_verb", "subject_refs": [ "girl" ], "object_refs": [], "confidence": "medium", "temporal_status": "localized", "intervals": [ { "start_s": 6.0, "end_s": 9.869999999999997, "source_start_s": 49.34, "source_end_s": 53.21 } ], "evidence": "The masked girl raises an arm at 6 seconds, bends and arches backward at 7–8 seconds, and is upright again with both arms overhead at 9 seconds. The inversion occurs between sparse samples.", "mask_relation": "subject matches the supplied selected-person mask", "source_subject_refs": "girl" }, { "id": "action_2", "mentions": [ { "text": "standing", "ranges": [ [ 51, 59 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "type": "posture", "subject_refs": [ "girl" ], "object_refs": [], "confidence": "high", "temporal_status": "localized", "intervals": [ { "start_s": 1.0, "end_s": 7.0, "source_start_s": 44.34, "source_end_s": 50.34 } ], "evidence": "The masked girl remains upright on her feet in the samples from 1 through 6 seconds.", "mask_relation": "subject matches the supplied selected-person mask", "source_subject_refs": "girl" }, { "id": "action_3", "mentions": [ { "text": "raises", "ranges": [ [ 74, 80 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "type": "simple", "subject_refs": [ "girl" ], "object_refs": [ "arms" ], "confidence": "high", "temporal_status": "localized", "intervals": [ { "start_s": 6.0, "end_s": 8.0, "source_start_s": 49.34, "source_end_s": 51.34 } ], "evidence": "The masked girl lifts an arm overhead at 6 seconds and moves her arms as she begins the maneuver at 7 seconds.", "mask_relation": "subject matches the supplied selected-person mask", "source_subject_refs": "girl" }, { "id": "action_4", "mentions": [ { "text": "arches", "ranges": [ [ 91, 97 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "type": "simple", "subject_refs": [ "girl" ], "object_refs": [ "back" ], "confidence": "high", "temporal_status": "localized", "intervals": [ { "start_s": 7.0, "end_s": 9.0, "source_start_s": 50.34, "source_end_s": 52.34 } ], "evidence": "At 7 and 8 seconds, the masked girl bends her knees and leans her torso backward into an arch.", "mask_relation": "subject matches the supplied selected-person mask", "source_subject_refs": "girl" }, { "id": "action_5", "mentions": [ { "text": "kicks", "ranges": [ [ 144, 149 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "type": "simple", "subject_refs": [ "girl" ], "object_refs": [ "legs" ], "confidence": "medium", "temporal_status": "uncertain", "intervals": [], "evidence": "No sampled frame clearly shows the masked girl's legs passing overhead; this phase likely falls between the sparse 8- and 9-second samples.", "mask_relation": "subject matches the supplied selected-person mask", "source_subject_refs": "girl" }, { "id": "action_6", "mentions": [ { "text": "returns", "ranges": [ [ 180, 187 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "type": "simple", "subject_refs": [ "girl" ], "object_refs": [], "confidence": "medium", "temporal_status": "localized", "intervals": [ { "start_s": 9.0, "end_s": 9.869999999999997, "source_start_s": 52.34, "source_end_s": 53.21 } ], "evidence": "The masked girl is upright again at 9 seconds after being arched backward at 8 seconds.", "mask_relation": "subject matches the supplied selected-person mask", "source_subject_refs": "girl" }, { "id": "action_7", "mentions": [ { "text": "standing", "ranges": [ [ 193, 201 ] ], "occurrence": 2, "particle": { "text": null, "occurrence": 1 } } ], "type": "posture", "subject_refs": [ "girl" ], "object_refs": [], "confidence": "high", "temporal_status": "localized", "intervals": [ { "start_s": 9.0, "end_s": 9.869999999999997, "source_start_s": 52.34, "source_end_s": 53.21 } ], "evidence": "At 9 seconds, the masked girl is standing upright with both feet on the floor.", "mask_relation": "subject matches the supplied selected-person mask", "source_subject_refs": "girl" } ], "notes": [ "The supplied masklet identifies the foreground girl from 1 through 9 seconds. At 0 seconds, its visible mask lies on a different person at the far-right edge, so that sample was not used as support for the girl's actions.", "No separate masks were supplied for the girl's arms, back, hands, legs, or the floor.", "The captioned action \"touch\" has the grammatical subject \"her hands\"; because the hands have no supplied source mask, it is omitted under the mask-linked-subject rule.", "The final participle \"raised\" has the body-part subject \"her arms\" in the construction \"with her arms raised\"; because the arms have no supplied source mask, it is not annotated as a separate action.", "The walkover's inversion and leg-overhead phase are not directly captured by the one-second samples." ] }, "model_output_raw": { "caption": "The girl performs a back walkover, starting from a standing position. She raises her arms, arches her back until her hands touch the floor, and kicks her legs over one by one. She returns to a standing position with her arms raised.", "entity_groups": [ { "id": "girl", "label": "girl", "spans": [ { "text": "The girl", "range": [ 0, 8 ] }, { "text": "She", "range": [ 70, 73 ] }, { "text": "her", "range": [ 81, 84 ] }, { "text": "her", "range": [ 98, 101 ] }, { "text": "her", "range": [ 113, 116 ] }, { "text": "her", "range": [ 150, 153 ] }, { "text": "She", "range": [ 176, 179 ] }, { "text": "her", "range": [ 216, 219 ] } ], "source_mask": "selected_person_masklet" }, { "id": "arms", "label": "arms", "spans": [ { "text": "her arms", "range": [ 81, 89 ] }, { "text": "her arms", "range": [ 216, 224 ] } ], "source_mask": null }, { "id": "back", "label": "back", "spans": [ { "text": "her back", "range": [ 98, 106 ] } ], "source_mask": null }, { "id": "hands", "label": "hands", "spans": [ { "text": "her hands", "range": [ 113, 122 ] } ], "source_mask": null }, { "id": "floor", "label": "floor", "spans": [ { "text": "the floor", "range": [ 129, 138 ] } ], "source_mask": null }, { "id": "legs", "label": "legs", "spans": [ { "text": "her legs", "range": [ 150, 158 ] }, { "text": "one", "range": [ 164, 167 ] }, { "text": "one", "range": [ 171, 174 ] } ], "source_mask": null } ], "actions": [ { "mentions": [ { "text": "performs a back walkover", "ranges": [ [ 9, 33 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "type": "light_verb", "subject": "girl", "object": null, "confidence": "medium", "mask_relation": "subject matches the supplied selected-person mask", "temporal_status": "localized", "intervals": [ { "clip_relative_s": [ 6.0, 9.87 ], "source_video_s": [ 49.34, 53.21 ] } ], "evidence": "The masked girl raises an arm at 6 seconds, bends and arches backward at 7–8 seconds, and is upright again with both arms overhead at 9 seconds. The inversion occurs between sparse samples." }, { "mentions": [ { "text": "standing", "ranges": [ [ 51, 59 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "type": "posture", "subject": "girl", "object": null, "confidence": "high", "mask_relation": "subject matches the supplied selected-person mask", "temporal_status": "localized", "intervals": [ { "clip_relative_s": [ 1.0, 7.0 ], "source_video_s": [ 44.34, 50.34 ] } ], "evidence": "The masked girl remains upright on her feet in the samples from 1 through 6 seconds." }, { "mentions": [ { "text": "raises", "ranges": [ [ 74, 80 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "type": "simple", "subject": "girl", "object": "arms", "confidence": "high", "mask_relation": "subject matches the supplied selected-person mask", "temporal_status": "localized", "intervals": [ { "clip_relative_s": [ 6.0, 8.0 ], "source_video_s": [ 49.34, 51.34 ] } ], "evidence": "The masked girl lifts an arm overhead at 6 seconds and moves her arms as she begins the maneuver at 7 seconds." }, { "mentions": [ { "text": "arches", "ranges": [ [ 91, 97 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "type": "simple", "subject": "girl", "object": "back", "confidence": "high", "mask_relation": "subject matches the supplied selected-person mask", "temporal_status": "localized", "intervals": [ { "clip_relative_s": [ 7.0, 9.0 ], "source_video_s": [ 50.34, 52.34 ] } ], "evidence": "At 7 and 8 seconds, the masked girl bends her knees and leans her torso backward into an arch." }, { "mentions": [ { "text": "kicks", "ranges": [ [ 144, 149 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "type": "simple", "subject": "girl", "object": "legs", "confidence": "medium", "mask_relation": "subject matches the supplied selected-person mask", "temporal_status": "uncertain", "intervals": [], "evidence": "No sampled frame clearly shows the masked girl's legs passing overhead; this phase likely falls between the sparse 8- and 9-second samples." }, { "mentions": [ { "text": "returns", "ranges": [ [ 180, 187 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "type": "simple", "subject": "girl", "object": null, "confidence": "medium", "mask_relation": "subject matches the supplied selected-person mask", "temporal_status": "localized", "intervals": [ { "clip_relative_s": [ 9.0, 9.87 ], "source_video_s": [ 52.34, 53.21 ] } ], "evidence": "The masked girl is upright again at 9 seconds after being arched backward at 8 seconds." }, { "mentions": [ { "text": "standing", "ranges": [ [ 193, 201 ] ], "occurrence": 2, "particle": { "text": null, "occurrence": 1 } } ], "type": "posture", "subject": "girl", "object": null, "confidence": "high", "mask_relation": "subject matches the supplied selected-person mask", "temporal_status": "localized", "intervals": [ { "clip_relative_s": [ 9.0, 9.87 ], "source_video_s": [ 52.34, 53.21 ] } ], "evidence": "At 9 seconds, the masked girl is standing upright with both feet on the floor." } ], "notes": [ "The supplied masklet identifies the foreground girl from 1 through 9 seconds. At 0 seconds, its visible mask lies on a different person at the far-right edge, so that sample was not used as support for the girl's actions.", "No separate masks were supplied for the girl's arms, back, hands, legs, or the floor.", "The captioned action \"touch\" has the grammatical subject \"her hands\"; because the hands have no supplied source mask, it is omitted under the mask-linked-subject rule.", "The final participle \"raised\" has the body-part subject \"her arms\" in the construction \"with her arms raised\"; because the arms have no supplied source mask, it is not annotated as a separate action.", "The walkover's inversion and leg-overhead phase are not directly captured by the one-second samples." ] }, "normalization_repairs": 0 }, { "source_video_id": "v_50MmJszF2XM", "caption_count_category": 11, "clip": { "caption_index": 5, "caption": "She lowers her arms, turns, and walks a few steps to her left. She then turns back to her original position and stands still, facing right with her arms at her sides.", "source_interval_sec": [ 53.21, 69.71 ], "loom_clip_id": "v_50MmJszF2XM_1595_2089", "clip_frame_interval_4fps": [ 1595, 2089 ], "sample_count_1fps": 17, "sampled_frames": [ { "frame": "00000", "timestamp_sec": 53.21, "offset_sec": 0, "rgb_path": "assets/cat_11/v_50MmJszF2XM_1595_2089/0000s_00000_rgb.jpg", "mask_path": "assets/cat_11/v_50MmJszF2XM_1595_2089/0000s_00000_mask.png", "mask_status": "available" }, { "frame": "00028", "timestamp_sec": 54.21, "offset_sec": 1, "rgb_path": "assets/cat_11/v_50MmJszF2XM_1595_2089/0001s_00028_rgb.jpg", "mask_path": "assets/cat_11/v_50MmJszF2XM_1595_2089/0001s_00028_mask.png", "mask_status": "available" }, { "frame": "00060", "timestamp_sec": 55.21, "offset_sec": 2, "rgb_path": "assets/cat_11/v_50MmJszF2XM_1595_2089/0002s_00060_rgb.jpg", "mask_path": "assets/cat_11/v_50MmJszF2XM_1595_2089/0002s_00060_mask.png", "mask_status": "available" }, { "frame": "00088", "timestamp_sec": 56.21, "offset_sec": 3, "rgb_path": "assets/cat_11/v_50MmJszF2XM_1595_2089/0003s_00088_rgb.jpg", "mask_path": "assets/cat_11/v_50MmJszF2XM_1595_2089/0003s_00088_mask.png", "mask_status": "available" }, { "frame": "00120", "timestamp_sec": 57.21, "offset_sec": 4, "rgb_path": "assets/cat_11/v_50MmJszF2XM_1595_2089/0004s_00120_rgb.jpg", "mask_path": "assets/cat_11/v_50MmJszF2XM_1595_2089/0004s_00120_mask.png", "mask_status": "available" }, { "frame": "00148", "timestamp_sec": 58.21, "offset_sec": 5, "rgb_path": "assets/cat_11/v_50MmJszF2XM_1595_2089/0005s_00148_rgb.jpg", "mask_path": "assets/cat_11/v_50MmJszF2XM_1595_2089/0005s_00148_mask.png", "mask_status": "available" }, { "frame": "00180", "timestamp_sec": 59.21, "offset_sec": 6, "rgb_path": "assets/cat_11/v_50MmJszF2XM_1595_2089/0006s_00180_rgb.jpg", "mask_path": "assets/cat_11/v_50MmJszF2XM_1595_2089/0006s_00180_mask.png", "mask_status": "available" }, { "frame": "00208", "timestamp_sec": 60.21, "offset_sec": 7, "rgb_path": "assets/cat_11/v_50MmJszF2XM_1595_2089/0007s_00208_rgb.jpg", "mask_path": "assets/cat_11/v_50MmJszF2XM_1595_2089/0007s_00208_mask.png", "mask_status": "available" }, { "frame": "00240", "timestamp_sec": 61.21, "offset_sec": 8, "rgb_path": "assets/cat_11/v_50MmJszF2XM_1595_2089/0008s_00240_rgb.jpg", "mask_path": "assets/cat_11/v_50MmJszF2XM_1595_2089/0008s_00240_mask.png", "mask_status": "available" }, { "frame": "00268", "timestamp_sec": 62.21, "offset_sec": 9, "rgb_path": "assets/cat_11/v_50MmJszF2XM_1595_2089/0009s_00268_rgb.jpg", "mask_path": "assets/cat_11/v_50MmJszF2XM_1595_2089/0009s_00268_mask.png", "mask_status": "available" }, { "frame": "00300", "timestamp_sec": 63.21, "offset_sec": 10, "rgb_path": "assets/cat_11/v_50MmJszF2XM_1595_2089/0010s_00300_rgb.jpg", "mask_path": "assets/cat_11/v_50MmJszF2XM_1595_2089/0010s_00300_mask.png", "mask_status": "available" }, { "frame": "00328", "timestamp_sec": 64.21, "offset_sec": 11, "rgb_path": "assets/cat_11/v_50MmJszF2XM_1595_2089/0011s_00328_rgb.jpg", "mask_path": "assets/cat_11/v_50MmJszF2XM_1595_2089/0011s_00328_mask.png", "mask_status": "available" }, { "frame": "00360", "timestamp_sec": 65.21, "offset_sec": 12, "rgb_path": "assets/cat_11/v_50MmJszF2XM_1595_2089/0012s_00360_rgb.jpg", "mask_path": "assets/cat_11/v_50MmJszF2XM_1595_2089/0012s_00360_mask.png", "mask_status": "available" }, { "frame": "00388", "timestamp_sec": 66.21, "offset_sec": 13, "rgb_path": "assets/cat_11/v_50MmJszF2XM_1595_2089/0013s_00388_rgb.jpg", "mask_path": "assets/cat_11/v_50MmJszF2XM_1595_2089/0013s_00388_mask.png", "mask_status": "available" }, { "frame": "00420", "timestamp_sec": 67.21, "offset_sec": 14, "rgb_path": "assets/cat_11/v_50MmJszF2XM_1595_2089/0014s_00420_rgb.jpg", "mask_path": "assets/cat_11/v_50MmJszF2XM_1595_2089/0014s_00420_mask.png", "mask_status": "available" }, { "frame": "00448", "timestamp_sec": 68.21, "offset_sec": 15, "rgb_path": "assets/cat_11/v_50MmJszF2XM_1595_2089/0015s_00448_rgb.jpg", "mask_path": "assets/cat_11/v_50MmJszF2XM_1595_2089/0015s_00448_mask.png", "mask_status": "available" }, { "frame": "00480", "timestamp_sec": 69.21, "offset_sec": 16, "rgb_path": "assets/cat_11/v_50MmJszF2XM_1595_2089/0016s_00480_rgb.jpg", "mask_path": "assets/cat_11/v_50MmJszF2XM_1595_2089/0016s_00480_mask.png", "mask_status": "available" } ], "obj_id": [ 0 ], "anno_id": [ 724 ], "fps_inferred_from_clip_interval": 29.9394 }, "usage": { "input_tokens": 53798, "input_tokens_details": { "cache_write_tokens": 53795, "cached_tokens": 0 }, "output_tokens": 3749, "output_tokens_details": { "reasoning_tokens": 2025 }, "total_tokens": 57547 }, "model": "gpt-5.6-sol", "effort": "medium", "custom_id": "cat11_cap05_v_50MmJszF2XM_1595_2089", "annotation": { "caption": "She lowers her arms, turns, and walks a few steps to her left. She then turns back to her original position and stands still, facing right with her arms at her sides.", "entity_groups": [ { "id": "entity_person_1", "label": "person", "spans": [ { "text": "She", "range": [ 0, 3 ] }, { "text": "her", "range": [ 11, 14 ] }, { "text": "her", "range": [ 53, 56 ] }, { "text": "She", "range": [ 63, 66 ] }, { "text": "her", "range": [ 86, 89 ] }, { "text": "her", "range": [ 144, 147 ] }, { "text": "her", "range": [ 156, 159 ] } ], "source_mask_link": "selected_person" }, { "id": "entity_arms_1", "label": "arms", "spans": [ { "text": "her arms", "range": [ 11, 19 ] }, { "text": "her arms", "range": [ 144, 152 ] } ], "source_mask_link": null }, { "id": "entity_sides_1", "label": "body sides", "spans": [ { "text": "her sides", "range": [ 156, 165 ] } ], "source_mask_link": null } ], "actions": [ { "id": "action_1", "mentions": [ { "text": "lowers", "ranges": [ [ 4, 10 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "type": "simple", "subject_refs": [ "entity_person_1" ], "object_refs": [ "entity_arms_1" ], "confidence": "high", "temporal_status": "localized", "intervals": [ { "start_s": 0.0, "end_s": 1.0, "source_start_s": 53.21, "source_end_s": 54.21 } ], "evidence": "The masked girl has both arms raised at 0.0 s and lowered beside her body by 1.0 s.", "mask_relation": "subject matches the supplied selected-person mask", "source_subject_refs": "entity_person_1" }, { "id": "action_2", "mentions": [ { "text": "turns", "ranges": [ [ 21, 26 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "type": "simple", "subject_refs": [ "entity_person_1" ], "object_refs": [], "confidence": "medium", "temporal_status": "localized", "intervals": [ { "start_s": 1.0, "end_s": 2.0, "source_start_s": 54.21, "source_end_s": 55.21 } ], "evidence": "Across 1.0–2.0 s, the masked girl changes her body orientation before moving across the floor.", "mask_relation": "subject matches the supplied selected-person mask", "source_subject_refs": "entity_person_1" }, { "id": "action_3", "mentions": [ { "text": "walks", "ranges": [ [ 32, 37 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "type": "simple", "subject_refs": [ "entity_person_1" ], "object_refs": [], "confidence": "high", "temporal_status": "localized", "intervals": [ { "start_s": 2.0, "end_s": 4.0, "source_start_s": 55.21, "source_end_s": 57.21 } ], "evidence": "At 2.0 and 3.0 s, the masked girl is visibly stepping toward the right side of the image and changing location.", "mask_relation": "subject matches the supplied selected-person mask", "source_subject_refs": "entity_person_1" }, { "id": "action_4", "mentions": [ { "text": "turns back", "ranges": [ [ 72, 82 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "type": "phrasal", "subject_refs": [ "entity_person_1" ], "object_refs": [], "confidence": "medium", "temporal_status": "localized", "intervals": [ { "start_s": 3.0, "end_s": 4.0, "source_start_s": 56.21, "source_end_s": 57.21 } ], "evidence": "The masked girl changes from a walking orientation at 3.0 s to a stationary profile orientation by 4.0 s.", "mask_relation": "subject matches the supplied selected-person mask", "source_subject_refs": "entity_person_1" }, { "id": "action_5", "mentions": [ { "text": "stands", "ranges": [ [ 112, 118 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "type": "posture", "subject_refs": [ "entity_person_1" ], "object_refs": [], "confidence": "high", "temporal_status": "localized", "intervals": [ { "start_s": 4.0, "end_s": 16.499999999999993, "source_start_s": 57.21, "source_end_s": 69.71 } ], "evidence": "From 4.0 s through the final sample at 16.0 s, the masked girl remains standing in approximately the same floor position.", "mask_relation": "subject matches the supplied selected-person mask", "source_subject_refs": "entity_person_1" }, { "id": "action_6", "mentions": [ { "text": "facing", "ranges": [ [ 126, 132 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "type": "posture", "subject_refs": [ "entity_person_1" ], "object_refs": [], "confidence": "low", "temporal_status": "uncertain", "intervals": [], "evidence": "The masked girl is shown in profile after 4.0 s, but the sampled frames do not clearly support the caption's stated rightward orientation.", "mask_relation": "subject matches the supplied selected-person mask", "source_subject_refs": "entity_person_1" } ], "notes": [ "The supplied mask consistently identifies the shorter girl in dark clothing, who performs the captioned sequence.", "The taller person visible in the frames is not mentioned by the caption and has no supplied mask.", "The arms and body sides are text-groundable body-part entities, but no independent masks are supplied for them.", "The nominal phrase \"a few steps\" names the extent of walking rather than a concrete segmentable entity.", "The direction stated for \"facing right\" is not clearly supported by the clean sampled frames, so that posture has no localized temporal interval." ] }, "model_output_raw": { "caption": "She lowers her arms, turns, and walks a few steps to her left. She then turns back to her original position and stands still, facing right with her arms at her sides.", "entity_groups": [ { "id": "entity_person_1", "label": "person", "spans": [ { "text": "She", "range": [ 0, 3 ] }, { "text": "her", "range": [ 11, 14 ] }, { "text": "her", "range": [ 53, 56 ] }, { "text": "She", "range": [ 63, 66 ] }, { "text": "her", "range": [ 86, 89 ] }, { "text": "her", "range": [ 144, 147 ] }, { "text": "her", "range": [ 156, 159 ] } ], "source_mask": "selected_person_masklet" }, { "id": "entity_arms_1", "label": "arms", "spans": [ { "text": "her arms", "range": [ 11, 19 ] }, { "text": "her arms", "range": [ 144, 152 ] } ], "source_mask": null }, { "id": "entity_sides_1", "label": "body sides", "spans": [ { "text": "her sides", "range": [ 156, 165 ] } ], "source_mask": null } ], "actions": [ { "id": "action_1", "mentions": [ { "text": "lowers", "ranges": [ [ 4, 10 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "type": "simple", "subject": "entity_person_1", "objects": [ "entity_arms_1" ], "confidence": "high", "mask_relation": "subject matches the supplied selected-person mask", "temporal_status": "localized", "intervals": [ { "clip_relative_s": [ 0.0, 1.0 ], "source_video_s": [ 53.21, 54.21 ] } ], "evidence": "The masked girl has both arms raised at 0.0 s and lowered beside her body by 1.0 s." }, { "id": "action_2", "mentions": [ { "text": "turns", "ranges": [ [ 21, 26 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "type": "simple", "subject": "entity_person_1", "objects": [], "confidence": "medium", "mask_relation": "subject matches the supplied selected-person mask", "temporal_status": "localized", "intervals": [ { "clip_relative_s": [ 1.0, 2.0 ], "source_video_s": [ 54.21, 55.21 ] } ], "evidence": "Across 1.0–2.0 s, the masked girl changes her body orientation before moving across the floor." }, { "id": "action_3", "mentions": [ { "text": "walks", "ranges": [ [ 32, 37 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "type": "simple", "subject": "entity_person_1", "objects": [], "confidence": "high", "mask_relation": "subject matches the supplied selected-person mask", "temporal_status": "localized", "intervals": [ { "clip_relative_s": [ 2.0, 4.0 ], "source_video_s": [ 55.21, 57.21 ] } ], "evidence": "At 2.0 and 3.0 s, the masked girl is visibly stepping toward the right side of the image and changing location." }, { "id": "action_4", "mentions": [ { "text": "turns back", "ranges": [ [ 72, 82 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "type": "phrasal", "subject": "entity_person_1", "objects": [], "confidence": "medium", "mask_relation": "subject matches the supplied selected-person mask", "temporal_status": "localized", "intervals": [ { "clip_relative_s": [ 3.0, 4.0 ], "source_video_s": [ 56.21, 57.21 ] } ], "evidence": "The masked girl changes from a walking orientation at 3.0 s to a stationary profile orientation by 4.0 s." }, { "id": "action_5", "mentions": [ { "text": "stands", "ranges": [ [ 112, 118 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "type": "posture", "subject": "entity_person_1", "objects": [], "confidence": "high", "mask_relation": "subject matches the supplied selected-person mask", "temporal_status": "localized", "intervals": [ { "clip_relative_s": [ 4.0, 16.5 ], "source_video_s": [ 57.21, 69.71 ] } ], "evidence": "From 4.0 s through the final sample at 16.0 s, the masked girl remains standing in approximately the same floor position." }, { "id": "action_6", "mentions": [ { "text": "facing", "ranges": [ [ 126, 132 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "type": "posture", "subject": "entity_person_1", "objects": [], "confidence": "low", "mask_relation": "subject matches the supplied selected-person mask", "temporal_status": "uncertain", "intervals": [], "evidence": "The masked girl is shown in profile after 4.0 s, but the sampled frames do not clearly support the caption's stated rightward orientation." } ], "notes": [ "The supplied mask consistently identifies the shorter girl in dark clothing, who performs the captioned sequence.", "The taller person visible in the frames is not mentioned by the caption and has no supplied mask.", "The arms and body sides are text-groundable body-part entities, but no independent masks are supplied for them.", "The nominal phrase \"a few steps\" names the extent of walking rather than a concrete segmentable entity.", "The direction stated for \"facing right\" is not clearly supported by the clean sampled frames, so that posture has no localized temporal interval." ] }, "normalization_repairs": 0 }, { "source_video_id": "v_50MmJszF2XM", "caption_count_category": 11, "clip": { "caption_index": 6, "caption": "After standing still for a moment, the girl performs a roundoff. She lunges forward, places her hands on the mat, and propels her body over, landing on both feet facing the opposite direction. She finishes with her arms raised.", "source_interval_sec": [ 69.71, 84.92 ], "loom_clip_id": "v_50MmJszF2XM_2089_2545", "clip_frame_interval_4fps": [ 2089, 2545 ], "sample_count_1fps": 16, "sampled_frames": [ { "frame": "00000", "timestamp_sec": 69.71, "offset_sec": 0, "rgb_path": "assets/cat_11/v_50MmJszF2XM_2089_2545/0000s_00000_rgb.jpg", "mask_path": "assets/cat_11/v_50MmJszF2XM_2089_2545/0000s_00000_mask.png", "mask_status": "available" }, { "frame": "00028", "timestamp_sec": 70.71, "offset_sec": 1, "rgb_path": "assets/cat_11/v_50MmJszF2XM_2089_2545/0001s_00028_rgb.jpg", "mask_path": "assets/cat_11/v_50MmJszF2XM_2089_2545/0001s_00028_mask.png", "mask_status": "available" }, { "frame": "00060", "timestamp_sec": 71.71, "offset_sec": 2, "rgb_path": "assets/cat_11/v_50MmJszF2XM_2089_2545/0002s_00060_rgb.jpg", "mask_path": "assets/cat_11/v_50MmJszF2XM_2089_2545/0002s_00060_mask.png", "mask_status": "available" }, { "frame": "00088", "timestamp_sec": 72.71, "offset_sec": 3, "rgb_path": "assets/cat_11/v_50MmJszF2XM_2089_2545/0003s_00088_rgb.jpg", "mask_path": "assets/cat_11/v_50MmJszF2XM_2089_2545/0003s_00088_mask.png", "mask_status": "available" }, { "frame": "00120", "timestamp_sec": 73.71, "offset_sec": 4, "rgb_path": "assets/cat_11/v_50MmJszF2XM_2089_2545/0004s_00120_rgb.jpg", "mask_path": "assets/cat_11/v_50MmJszF2XM_2089_2545/0004s_00120_mask.png", "mask_status": "available" }, { "frame": "00148", "timestamp_sec": 74.71, "offset_sec": 5, "rgb_path": "assets/cat_11/v_50MmJszF2XM_2089_2545/0005s_00148_rgb.jpg", "mask_path": "assets/cat_11/v_50MmJszF2XM_2089_2545/0005s_00148_mask.png", "mask_status": "available" }, { "frame": "00180", "timestamp_sec": 75.71, "offset_sec": 6, "rgb_path": "assets/cat_11/v_50MmJszF2XM_2089_2545/0006s_00180_rgb.jpg", "mask_path": "assets/cat_11/v_50MmJszF2XM_2089_2545/0006s_00180_mask.png", "mask_status": "available" }, { "frame": "00208", "timestamp_sec": 76.71, "offset_sec": 7, "rgb_path": "assets/cat_11/v_50MmJszF2XM_2089_2545/0007s_00208_rgb.jpg", "mask_path": "assets/cat_11/v_50MmJszF2XM_2089_2545/0007s_00208_mask.png", "mask_status": "available" }, { "frame": "00240", "timestamp_sec": 77.71, "offset_sec": 8, "rgb_path": "assets/cat_11/v_50MmJszF2XM_2089_2545/0008s_00240_rgb.jpg", "mask_path": "assets/cat_11/v_50MmJszF2XM_2089_2545/0008s_00240_mask.png", "mask_status": "available" }, { "frame": "00268", "timestamp_sec": 78.71, "offset_sec": 9, "rgb_path": "assets/cat_11/v_50MmJszF2XM_2089_2545/0009s_00268_rgb.jpg", "mask_path": "assets/cat_11/v_50MmJszF2XM_2089_2545/0009s_00268_mask.png", "mask_status": "available" }, { "frame": "00300", "timestamp_sec": 79.71, "offset_sec": 10, "rgb_path": "assets/cat_11/v_50MmJszF2XM_2089_2545/0010s_00300_rgb.jpg", "mask_path": "assets/cat_11/v_50MmJszF2XM_2089_2545/0010s_00300_mask.png", "mask_status": "available" }, { "frame": "00328", "timestamp_sec": 80.71, "offset_sec": 11, "rgb_path": "assets/cat_11/v_50MmJszF2XM_2089_2545/0011s_00328_rgb.jpg", "mask_path": "assets/cat_11/v_50MmJszF2XM_2089_2545/0011s_00328_mask.png", "mask_status": "available" }, { "frame": "00360", "timestamp_sec": 81.71, "offset_sec": 12, "rgb_path": "assets/cat_11/v_50MmJszF2XM_2089_2545/0012s_00360_rgb.jpg", "mask_path": "assets/cat_11/v_50MmJszF2XM_2089_2545/0012s_00360_mask.png", "mask_status": "available" }, { "frame": "00388", "timestamp_sec": 82.71, "offset_sec": 13, "rgb_path": "assets/cat_11/v_50MmJszF2XM_2089_2545/0013s_00388_rgb.jpg", "mask_path": "assets/cat_11/v_50MmJszF2XM_2089_2545/0013s_00388_mask.png", "mask_status": "available" }, { "frame": "00420", "timestamp_sec": 83.71, "offset_sec": 14, "rgb_path": "assets/cat_11/v_50MmJszF2XM_2089_2545/0014s_00420_rgb.jpg", "mask_path": "assets/cat_11/v_50MmJszF2XM_2089_2545/0014s_00420_mask.png", "mask_status": "available" }, { "frame": "00448", "timestamp_sec": 84.71, "offset_sec": 15, "rgb_path": "assets/cat_11/v_50MmJszF2XM_2089_2545/0015s_00448_rgb.jpg", "mask_path": "assets/cat_11/v_50MmJszF2XM_2089_2545/0015s_00448_mask.png", "mask_status": "available" } ], "obj_id": [ 0 ], "anno_id": [ 725 ], "fps_inferred_from_clip_interval": 29.9803 }, "usage": { "input_tokens": 51547, "input_tokens_details": { "cache_write_tokens": 51544, "cached_tokens": 0 }, "output_tokens": 5029, "output_tokens_details": { "reasoning_tokens": 2876 }, "total_tokens": 56576 }, "model": "gpt-5.6-sol", "effort": "medium", "custom_id": "cat11_cap06_v_50MmJszF2XM_2089_2545", "annotation": { "caption": "After standing still for a moment, the girl performs a roundoff. She lunges forward, places her hands on the mat, and propels her body over, landing on both feet facing the opposite direction. She finishes with her arms raised.", "entity_groups": [ { "id": "girl", "label": "girl", "spans": [ { "text": "the girl", "range": [ 35, 43 ] }, { "text": "She", "range": [ 65, 68 ] }, { "text": "her", "range": [ 92, 95 ] }, { "text": "her", "range": [ 126, 129 ] }, { "text": "She", "range": [ 193, 196 ] }, { "text": "her", "range": [ 211, 214 ] } ], "source_mask_link": "selected_person" }, { "id": "hands", "label": "hands", "spans": [ { "text": "her hands", "range": [ 92, 101 ] } ], "source_mask_link": null }, { "id": "mat", "label": "mat", "spans": [ { "text": "the mat", "range": [ 105, 112 ] } ], "source_mask_link": null }, { "id": "body", "label": "body", "spans": [ { "text": "her body", "range": [ 126, 134 ] } ], "source_mask_link": null }, { "id": "feet", "label": "feet", "spans": [ { "text": "both feet", "range": [ 152, 161 ] } ], "source_mask_link": null }, { "id": "arms", "label": "arms", "spans": [ { "text": "her arms", "range": [ 211, 219 ] } ], "source_mask_link": null } ], "actions": [ { "id": "action_1", "mentions": [ { "text": "standing", "ranges": [ [ 6, 14 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "type": "posture", "subject_refs": [ "girl" ], "object_refs": [], "confidence": "high", "temporal_status": "localized_sparse", "intervals": [ { "start_s": 0.0, "end_s": 10.0, "source_start_s": 69.71, "source_end_s": 79.71 } ], "evidence": "At samples 0–9 seconds, the masked girl remains upright and nearly motionless.", "mask_relation": "subject matches the supplied selected-person mask", "source_subject_refs": "girl" }, { "id": "action_2", "mentions": [ { "text": "performs a roundoff", "ranges": [ [ 44, 63 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "type": "light_verb", "subject_refs": [ "girl" ], "object_refs": [], "confidence": "high", "temporal_status": "localized_sparse", "intervals": [ { "start_s": 12.0, "end_s": 14.0, "source_start_s": 81.71, "source_end_s": 83.71 } ], "evidence": "The masked girl is pitched forward at 12 seconds and is upright after the maneuver at 13 seconds; the inversion occurs between sparse samples.", "mask_relation": "subject matches the supplied selected-person mask", "source_subject_refs": "girl" }, { "id": "action_3", "mentions": [ { "text": "lunges", "ranges": [ [ 69, 75 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "type": "simple", "subject_refs": [ "girl" ], "object_refs": [], "confidence": "high", "temporal_status": "localized_sparse", "intervals": [ { "start_s": 10.0, "end_s": 13.0, "source_start_s": 79.71, "source_end_s": 82.71 } ], "evidence": "At 10–12 seconds, the masked girl steps forward, extends one leg, raises her arms, and leans into the maneuver.", "mask_relation": "subject matches the supplied selected-person mask", "source_subject_refs": "girl" }, { "id": "action_4", "mentions": [ { "text": "places", "ranges": [ [ 85, 91 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "type": "simple", "subject_refs": [ "girl" ], "object_refs": [ "hands", "mat" ], "confidence": "high", "temporal_status": "not_observed_in_sampled_frames", "intervals": [], "evidence": "No sampled clean frame directly shows her hands contacting the mat; this phase falls between sparse samples.", "mask_relation": "subject matches the supplied selected-person mask", "source_subject_refs": "girl" }, { "id": "action_5", "mentions": [ { "text": "propels", "ranges": [ [ 118, 125 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "type": "simple", "subject_refs": [ "girl" ], "object_refs": [ "body" ], "confidence": "high", "temporal_status": "not_observed_in_sampled_frames", "intervals": [], "evidence": "The over-the-body propulsion phase is not directly visible in any sampled clean frame.", "mask_relation": "subject matches the supplied selected-person mask", "source_subject_refs": "girl" }, { "id": "action_6", "mentions": [ { "text": "landing", "ranges": [ [ 141, 148 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "type": "simple", "subject_refs": [ "girl" ], "object_refs": [ "feet" ], "confidence": "high", "temporal_status": "localized_sparse", "intervals": [ { "start_s": 13.0, "end_s": 14.0, "source_start_s": 82.71, "source_end_s": 83.71 } ], "evidence": "At 13 seconds, the masked girl is upright with both feet planted immediately after the maneuver.", "mask_relation": "subject matches the supplied selected-person mask", "source_subject_refs": "girl" }, { "id": "action_7", "mentions": [ { "text": "facing", "ranges": [ [ 162, 168 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "type": "posture", "subject_refs": [ "girl" ], "object_refs": [], "confidence": "high", "temporal_status": "localized_sparse", "intervals": [ { "start_s": 13.0, "end_s": 15.210000000000008, "source_start_s": 82.71, "source_end_s": 84.92 } ], "evidence": "From 13 seconds through the clip boundary, the masked girl maintains her post-landing body orientation.", "mask_relation": "subject matches the supplied selected-person mask", "source_subject_refs": "girl" }, { "id": "action_8", "mentions": [ { "text": "finishes", "ranges": [ [ 197, 205 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "type": "simple", "subject_refs": [ "girl" ], "object_refs": [ "arms" ], "confidence": "medium", "temporal_status": "localized_sparse", "intervals": [ { "start_s": 14.0, "end_s": 15.0, "source_start_s": 83.71, "source_end_s": 84.71 } ], "evidence": "At 14 seconds, the masked girl holds the concluding pose with both arms overhead; by 15 seconds she is lowering them.", "mask_relation": "subject matches the supplied selected-person mask", "source_subject_refs": "girl" } ], "notes": [ "The supplied masklet consistently identifies the girl described by the caption, not the unmentioned adult visible in early frames.", "No separate masks were supplied for the girl's hands, body, feet, arms, or the mat, so those entity groups have null source-mask links.", "The hand placement and inverted propulsion phases occur between the 12- and 13-second samples and therefore have no direct sampled-frame interval.", "The descriptive participle \"raised\" was not added as a separate action because it states the arms' resulting held configuration and the arms have no separate supplied mask." ] }, "model_output_raw": { "caption": "After standing still for a moment, the girl performs a roundoff. She lunges forward, places her hands on the mat, and propels her body over, landing on both feet facing the opposite direction. She finishes with her arms raised.", "entity_groups": [ { "id": "girl", "label": "girl", "spans": [ { "text": "the girl", "range": [ 35, 43 ] }, { "text": "She", "range": [ 65, 68 ] }, { "text": "her", "range": [ 92, 95 ] }, { "text": "her", "range": [ 126, 129 ] }, { "text": "She", "range": [ 193, 196 ] }, { "text": "her", "range": [ 211, 214 ] } ], "source_mask": "selected_person_masklet" }, { "id": "hands", "label": "hands", "spans": [ { "text": "her hands", "range": [ 92, 101 ] } ], "source_mask": null }, { "id": "mat", "label": "mat", "spans": [ { "text": "the mat", "range": [ 105, 112 ] } ], "source_mask": null }, { "id": "body", "label": "body", "spans": [ { "text": "her body", "range": [ 126, 134 ] } ], "source_mask": null }, { "id": "feet", "label": "feet", "spans": [ { "text": "both feet", "range": [ 152, 161 ] } ], "source_mask": null }, { "id": "arms", "label": "arms", "spans": [ { "text": "her arms", "range": [ 211, 219 ] } ], "source_mask": null } ], "actions": [ { "id": "action_1", "mentions": [ { "text": "standing", "ranges": [ [ 6, 14 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "type": "posture", "subject_ref": "girl", "object_refs": [], "confidence": "high", "mask_relation": "subject matches the supplied selected-person mask", "temporal_status": "localized_sparse", "intervals": [ { "clip_relative_s": [ 0.0, 10.0 ], "source_video_s": [ 69.71, 79.71 ] } ], "evidence": "At samples 0–9 seconds, the masked girl remains upright and nearly motionless." }, { "id": "action_2", "mentions": [ { "text": "performs a roundoff", "ranges": [ [ 44, 63 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "type": "light_verb", "subject_ref": "girl", "object_refs": [], "confidence": "high", "mask_relation": "subject matches the supplied selected-person mask", "temporal_status": "localized_sparse", "intervals": [ { "clip_relative_s": [ 12.0, 14.0 ], "source_video_s": [ 81.71, 83.71 ] } ], "evidence": "The masked girl is pitched forward at 12 seconds and is upright after the maneuver at 13 seconds; the inversion occurs between sparse samples." }, { "id": "action_3", "mentions": [ { "text": "lunges", "ranges": [ [ 69, 75 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "type": "simple", "subject_ref": "girl", "object_refs": [], "confidence": "high", "mask_relation": "subject matches the supplied selected-person mask", "temporal_status": "localized_sparse", "intervals": [ { "clip_relative_s": [ 10.0, 13.0 ], "source_video_s": [ 79.71, 82.71 ] } ], "evidence": "At 10–12 seconds, the masked girl steps forward, extends one leg, raises her arms, and leans into the maneuver." }, { "id": "action_4", "mentions": [ { "text": "places", "ranges": [ [ 85, 91 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "type": "simple", "subject_ref": "girl", "object_refs": [ "hands", "mat" ], "confidence": "high", "mask_relation": "subject matches the supplied selected-person mask", "temporal_status": "not_observed_in_sampled_frames", "intervals": [], "evidence": "No sampled clean frame directly shows her hands contacting the mat; this phase falls between sparse samples." }, { "id": "action_5", "mentions": [ { "text": "propels", "ranges": [ [ 118, 125 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "type": "simple", "subject_ref": "girl", "object_refs": [ "body" ], "confidence": "high", "mask_relation": "subject matches the supplied selected-person mask", "temporal_status": "not_observed_in_sampled_frames", "intervals": [], "evidence": "The over-the-body propulsion phase is not directly visible in any sampled clean frame." }, { "id": "action_6", "mentions": [ { "text": "landing", "ranges": [ [ 141, 148 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "type": "simple", "subject_ref": "girl", "object_refs": [ "feet" ], "confidence": "high", "mask_relation": "subject matches the supplied selected-person mask", "temporal_status": "localized_sparse", "intervals": [ { "clip_relative_s": [ 13.0, 14.0 ], "source_video_s": [ 82.71, 83.71 ] } ], "evidence": "At 13 seconds, the masked girl is upright with both feet planted immediately after the maneuver." }, { "id": "action_7", "mentions": [ { "text": "facing", "ranges": [ [ 162, 168 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "type": "posture", "subject_ref": "girl", "object_refs": [], "confidence": "high", "mask_relation": "subject matches the supplied selected-person mask", "temporal_status": "localized_sparse", "intervals": [ { "clip_relative_s": [ 13.0, 15.21 ], "source_video_s": [ 82.71, 84.92 ] } ], "evidence": "From 13 seconds through the clip boundary, the masked girl maintains her post-landing body orientation." }, { "id": "action_8", "mentions": [ { "text": "finishes", "ranges": [ [ 197, 205 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "type": "simple", "subject_ref": "girl", "object_refs": [ "arms" ], "confidence": "medium", "mask_relation": "subject matches the supplied selected-person mask", "temporal_status": "localized_sparse", "intervals": [ { "clip_relative_s": [ 14.0, 15.0 ], "source_video_s": [ 83.71, 84.71 ] } ], "evidence": "At 14 seconds, the masked girl holds the concluding pose with both arms overhead; by 15 seconds she is lowering them." } ], "notes": [ "The supplied masklet consistently identifies the girl described by the caption, not the unmentioned adult visible in early frames.", "No separate masks were supplied for the girl's hands, body, feet, arms, or the mat, so those entity groups have null source-mask links.", "The hand placement and inverted propulsion phases occur between the 12- and 13-second samples and therefore have no direct sampled-frame interval.", "The descriptive participle \"raised\" was not added as a separate action because it states the arms' resulting held configuration and the arms have no separate supplied mask." ] }, "normalization_repairs": 0 }, { "source_video_id": "v_50MmJszF2XM", "caption_count_category": 11, "clip": { "caption_index": 7, "caption": "She turns around and walks back to her starting spot, where an adult adjusts her posture. The adult lifts the girl's arms straight up, positioning them in a high V shape, which she then holds.", "source_interval_sec": [ 84.92, 100.71 ], "loom_clip_id": "v_50MmJszF2XM_2545_3018", "clip_frame_interval_4fps": [ 2545, 3018 ], "sample_count_1fps": 16, "sampled_frames": [ { "frame": "00000", "timestamp_sec": 84.92, "offset_sec": 0, "rgb_path": "assets/cat_11/v_50MmJszF2XM_2545_3018/0000s_00000_rgb.jpg", "mask_path": "assets/cat_11/v_50MmJszF2XM_2545_3018/0000s_00000_mask.png", "mask_status": "available" }, { "frame": "00028", "timestamp_sec": 85.92, "offset_sec": 1, "rgb_path": "assets/cat_11/v_50MmJszF2XM_2545_3018/0001s_00028_rgb.jpg", "mask_path": "assets/cat_11/v_50MmJszF2XM_2545_3018/0001s_00028_mask.png", "mask_status": "available" }, { "frame": "00060", "timestamp_sec": 86.92, "offset_sec": 2, "rgb_path": "assets/cat_11/v_50MmJszF2XM_2545_3018/0002s_00060_rgb.jpg", "mask_path": "assets/cat_11/v_50MmJszF2XM_2545_3018/0002s_00060_mask.png", "mask_status": "available" }, { "frame": "00088", "timestamp_sec": 87.92, "offset_sec": 3, "rgb_path": "assets/cat_11/v_50MmJszF2XM_2545_3018/0003s_00088_rgb.jpg", "mask_path": "assets/cat_11/v_50MmJszF2XM_2545_3018/0003s_00088_mask.png", "mask_status": "available" }, { "frame": "00120", "timestamp_sec": 88.92, "offset_sec": 4, "rgb_path": "assets/cat_11/v_50MmJszF2XM_2545_3018/0004s_00120_rgb.jpg", "mask_path": "assets/cat_11/v_50MmJszF2XM_2545_3018/0004s_00120_mask.png", "mask_status": "available" }, { "frame": "00148", "timestamp_sec": 89.92, "offset_sec": 5, "rgb_path": "assets/cat_11/v_50MmJszF2XM_2545_3018/0005s_00148_rgb.jpg", "mask_path": "assets/cat_11/v_50MmJszF2XM_2545_3018/0005s_00148_mask.png", "mask_status": "available" }, { "frame": "00180", "timestamp_sec": 90.92, "offset_sec": 6, "rgb_path": "assets/cat_11/v_50MmJszF2XM_2545_3018/0006s_00180_rgb.jpg", "mask_path": "assets/cat_11/v_50MmJszF2XM_2545_3018/0006s_00180_mask.png", "mask_status": "available" }, { "frame": "00208", "timestamp_sec": 91.92, "offset_sec": 7, "rgb_path": "assets/cat_11/v_50MmJszF2XM_2545_3018/0007s_00208_rgb.jpg", "mask_path": "assets/cat_11/v_50MmJszF2XM_2545_3018/0007s_00208_mask.png", "mask_status": "available" }, { "frame": "00240", "timestamp_sec": 92.92, "offset_sec": 8, "rgb_path": "assets/cat_11/v_50MmJszF2XM_2545_3018/0008s_00240_rgb.jpg", "mask_path": "assets/cat_11/v_50MmJszF2XM_2545_3018/0008s_00240_mask.png", "mask_status": "available" }, { "frame": "00268", "timestamp_sec": 93.92, "offset_sec": 9, "rgb_path": "assets/cat_11/v_50MmJszF2XM_2545_3018/0009s_00268_rgb.jpg", "mask_path": "assets/cat_11/v_50MmJszF2XM_2545_3018/0009s_00268_mask.png", "mask_status": "available" }, { "frame": "00300", "timestamp_sec": 94.92, "offset_sec": 10, "rgb_path": "assets/cat_11/v_50MmJszF2XM_2545_3018/0010s_00300_rgb.jpg", "mask_path": "assets/cat_11/v_50MmJszF2XM_2545_3018/0010s_00300_mask.png", "mask_status": "available" }, { "frame": "00328", "timestamp_sec": 95.92, "offset_sec": 11, "rgb_path": "assets/cat_11/v_50MmJszF2XM_2545_3018/0011s_00328_rgb.jpg", "mask_path": "assets/cat_11/v_50MmJszF2XM_2545_3018/0011s_00328_mask.png", "mask_status": "available" }, { "frame": "00360", "timestamp_sec": 96.92, "offset_sec": 12, "rgb_path": "assets/cat_11/v_50MmJszF2XM_2545_3018/0012s_00360_rgb.jpg", "mask_path": "assets/cat_11/v_50MmJszF2XM_2545_3018/0012s_00360_mask.png", "mask_status": "available" }, { "frame": "00388", "timestamp_sec": 97.92, "offset_sec": 13, "rgb_path": "assets/cat_11/v_50MmJszF2XM_2545_3018/0013s_00388_rgb.jpg", "mask_path": "assets/cat_11/v_50MmJszF2XM_2545_3018/0013s_00388_mask.png", "mask_status": "available" }, { "frame": "00420", "timestamp_sec": 98.92, "offset_sec": 14, "rgb_path": "assets/cat_11/v_50MmJszF2XM_2545_3018/0014s_00420_rgb.jpg", "mask_path": "assets/cat_11/v_50MmJszF2XM_2545_3018/0014s_00420_mask.png", "mask_status": "available" }, { "frame": "00448", "timestamp_sec": 99.92, "offset_sec": 15, "rgb_path": "assets/cat_11/v_50MmJszF2XM_2545_3018/0015s_00448_rgb.jpg", "mask_path": "assets/cat_11/v_50MmJszF2XM_2545_3018/0015s_00448_mask.png", "mask_status": "available" } ], "obj_id": [ 0 ], "anno_id": [ 726 ], "fps_inferred_from_clip_interval": 29.9557 }, "usage": { "input_tokens": 51538, "input_tokens_details": { "cache_write_tokens": 51535, "cached_tokens": 0 }, "output_tokens": 3496, "output_tokens_details": { "reasoning_tokens": 2387 }, "total_tokens": 55034 }, "model": "gpt-5.6-sol", "effort": "medium", "custom_id": "cat11_cap07_v_50MmJszF2XM_2545_3018", "annotation": { "caption": "She turns around and walks back to her starting spot, where an adult adjusts her posture. The adult lifts the girl's arms straight up, positioning them in a high V shape, which she then holds.", "entity_groups": [ { "id": "girl_1", "label": "girl", "spans": [ { "text": "She", "range": [ 0, 3 ] }, { "text": "her", "range": [ 35, 38 ] }, { "text": "her", "range": [ 77, 80 ] }, { "text": "girl's", "range": [ 110, 116 ] }, { "text": "she", "range": [ 177, 180 ] } ], "source_mask_link": "selected_person" }, { "id": "adult_1", "label": "adult", "spans": [ { "text": "an adult", "range": [ 60, 68 ] }, { "text": "The adult", "range": [ 90, 99 ] } ], "source_mask_link": null }, { "id": "arms_1", "label": "girl's arms", "spans": [ { "text": "the girl's arms", "range": [ 106, 121 ] }, { "text": "them", "range": [ 147, 151 ] } ], "source_mask_link": null } ], "actions": [ { "id": "action_1", "mentions": [ { "text": "turns around", "ranges": [ [ 4, 16 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "type": "phrasal", "subject_refs": [ "girl_1" ], "object_refs": [], "confidence": "high", "temporal_status": "localized", "intervals": [ { "start_s": 0.0, "end_s": 2.0, "source_start_s": 84.92, "source_end_s": 86.92 } ], "evidence": "The selected girl faces forward at 0.000 seconds and has turned away while stepping toward the adult by 1.000 seconds.", "mask_relation": "subject matches the supplied selected-person mask", "source_subject_refs": "girl_1" }, { "id": "action_2", "mentions": [ { "text": "walks", "ranges": [ [ 21, 26 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "type": "simple", "subject_refs": [ "girl_1" ], "object_refs": [], "confidence": "high", "temporal_status": "localized", "intervals": [ { "start_s": 1.0, "end_s": 5.0, "source_start_s": 85.92, "source_end_s": 89.92 } ], "evidence": "At 1.000–4.000 seconds, the masked girl takes successive steps back toward the waiting adult.", "mask_relation": "subject matches the supplied selected-person mask", "source_subject_refs": "girl_1" }, { "id": "action_3", "mentions": [ { "text": "holds", "ranges": [ [ 186, 191 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "type": "simple", "subject_refs": [ "girl_1" ], "object_refs": [ "arms_1" ], "confidence": "high", "temporal_status": "localized", "intervals": [ { "start_s": 7.0, "end_s": 15.789999999999992, "source_start_s": 91.92, "source_end_s": 100.71 } ], "evidence": "From 7.000 seconds through the final 15.000-second sample, the masked girl maintains both arms overhead in the described high-V position, with assistance from the adult.", "mask_relation": "subject matches the supplied selected-person mask", "source_subject_refs": "girl_1" } ], "notes": [ "The supplied masklet consistently identifies the girl, not the adult.", "The adult has no supplied instance mask, so the adult-subject actions “adjusts,” “lifts,” and “positioning” are omitted under the mask-linked action rule.", "The girl's arms are text-groundable body-part entities, but no independent arm mask was supplied; the person mask includes them only as part of the girl's silhouette.", "“starting spot,” “posture,” and “a high V shape” describe a location or bodily configuration rather than independently segmentable concrete entities.", "The source-video clip interval is [84.920000, 100.710000), corresponding to the clip-relative interval [0.000000, 15.790000)." ] }, "model_output_raw": { "caption": "She turns around and walks back to her starting spot, where an adult adjusts her posture. The adult lifts the girl's arms straight up, positioning them in a high V shape, which she then holds.", "entity_groups": [ { "id": "girl_1", "label": "girl", "spans": [ { "text": "She", "range": [ 0, 3 ] }, { "text": "her", "range": [ 35, 38 ] }, { "text": "her", "range": [ 77, 80 ] }, { "text": "girl's", "range": [ 110, 116 ] }, { "text": "she", "range": [ 177, 180 ] } ], "source_mask": { "track_id": "selected_person_masklet" } }, { "id": "adult_1", "label": "adult", "spans": [ { "text": "an adult", "range": [ 60, 68 ] }, { "text": "The adult", "range": [ 90, 99 ] } ], "source_mask": null }, { "id": "arms_1", "label": "girl's arms", "spans": [ { "text": "the girl's arms", "range": [ 106, 121 ] }, { "text": "them", "range": [ 147, 151 ] } ], "source_mask": null } ], "actions": [ { "mentions": [ { "text": "turns around", "ranges": [ [ 4, 16 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "type": "phrasal", "subject": "girl_1", "objects": [], "confidence": "high", "temporal_status": "localized", "intervals": [ { "start_s": 0.0, "end_s": 2.0 } ], "evidence": "The selected girl faces forward at 0.000 seconds and has turned away while stepping toward the adult by 1.000 seconds.", "mask_relation": "subject matches the supplied selected-person mask" }, { "mentions": [ { "text": "walks", "ranges": [ [ 21, 26 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "type": "simple", "subject": "girl_1", "objects": [], "confidence": "high", "temporal_status": "localized", "intervals": [ { "start_s": 1.0, "end_s": 5.0 } ], "evidence": "At 1.000–4.000 seconds, the masked girl takes successive steps back toward the waiting adult.", "mask_relation": "subject matches the supplied selected-person mask" }, { "mentions": [ { "text": "holds", "ranges": [ [ 186, 191 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "type": "simple", "subject": "girl_1", "objects": [ "arms_1" ], "confidence": "high", "temporal_status": "localized", "intervals": [ { "start_s": 7.0, "end_s": 15.79 } ], "evidence": "From 7.000 seconds through the final 15.000-second sample, the masked girl maintains both arms overhead in the described high-V position, with assistance from the adult.", "mask_relation": "subject matches the supplied selected-person mask" } ], "notes": [ "The supplied masklet consistently identifies the girl, not the adult.", "The adult has no supplied instance mask, so the adult-subject actions “adjusts,” “lifts,” and “positioning” are omitted under the mask-linked action rule.", "The girl's arms are text-groundable body-part entities, but no independent arm mask was supplied; the person mask includes them only as part of the girl's silhouette.", "“starting spot,” “posture,” and “a high V shape” describe a location or bodily configuration rather than independently segmentable concrete entities.", "The source-video clip interval is [84.920000, 100.710000), corresponding to the clip-relative interval [0.000000, 15.790000)." ] }, "normalization_repairs": 0 }, { "source_video_id": "v_50MmJszF2XM", "caption_count_category": 11, "clip": { "caption_index": 8, "caption": "While standing, she practices arm movements, transitioning from a high V to a position with bent elbows and hands near her head. She then extends her arms back up into the high V and holds the pose.", "source_interval_sec": [ 100.71, 110.71 ], "loom_clip_id": "v_50MmJszF2XM_3018_3318", "clip_frame_interval_4fps": [ 3018, 3318 ], "sample_count_1fps": 10, "sampled_frames": [ { "frame": "00000", "timestamp_sec": 100.71, "offset_sec": 0, "rgb_path": "assets/cat_11/v_50MmJszF2XM_3018_3318/0000s_00000_rgb.jpg", "mask_path": "assets/cat_11/v_50MmJszF2XM_3018_3318/0000s_00000_mask.png", "mask_status": "available" }, { "frame": "00028", "timestamp_sec": 101.71, "offset_sec": 1, "rgb_path": "assets/cat_11/v_50MmJszF2XM_3018_3318/0001s_00028_rgb.jpg", "mask_path": "assets/cat_11/v_50MmJszF2XM_3018_3318/0001s_00028_mask.png", "mask_status": "available" }, { "frame": "00060", "timestamp_sec": 102.71, "offset_sec": 2, "rgb_path": "assets/cat_11/v_50MmJszF2XM_3018_3318/0002s_00060_rgb.jpg", "mask_path": "assets/cat_11/v_50MmJszF2XM_3018_3318/0002s_00060_mask.png", "mask_status": "available" }, { "frame": "00088", "timestamp_sec": 103.71, "offset_sec": 3, "rgb_path": "assets/cat_11/v_50MmJszF2XM_3018_3318/0003s_00088_rgb.jpg", "mask_path": "assets/cat_11/v_50MmJszF2XM_3018_3318/0003s_00088_mask.png", "mask_status": "available" }, { "frame": "00120", "timestamp_sec": 104.71, "offset_sec": 4, "rgb_path": "assets/cat_11/v_50MmJszF2XM_3018_3318/0004s_00120_rgb.jpg", "mask_path": "assets/cat_11/v_50MmJszF2XM_3018_3318/0004s_00120_mask.png", "mask_status": "available" }, { "frame": "00148", "timestamp_sec": 105.71, "offset_sec": 5, "rgb_path": "assets/cat_11/v_50MmJszF2XM_3018_3318/0005s_00148_rgb.jpg", "mask_path": "assets/cat_11/v_50MmJszF2XM_3018_3318/0005s_00148_mask.png", "mask_status": "available" }, { "frame": "00180", "timestamp_sec": 106.71, "offset_sec": 6, "rgb_path": "assets/cat_11/v_50MmJszF2XM_3018_3318/0006s_00180_rgb.jpg", "mask_path": "assets/cat_11/v_50MmJszF2XM_3018_3318/0006s_00180_mask.png", "mask_status": "available" }, { "frame": "00208", "timestamp_sec": 107.71, "offset_sec": 7, "rgb_path": "assets/cat_11/v_50MmJszF2XM_3018_3318/0007s_00208_rgb.jpg", "mask_path": "assets/cat_11/v_50MmJszF2XM_3018_3318/0007s_00208_mask.png", "mask_status": "available" }, { "frame": "00240", "timestamp_sec": 108.71, "offset_sec": 8, "rgb_path": "assets/cat_11/v_50MmJszF2XM_3018_3318/0008s_00240_rgb.jpg", "mask_path": "assets/cat_11/v_50MmJszF2XM_3018_3318/0008s_00240_mask.png", "mask_status": "available" }, { "frame": "00268", "timestamp_sec": 109.71, "offset_sec": 9, "rgb_path": "assets/cat_11/v_50MmJszF2XM_3018_3318/0009s_00268_rgb.jpg", "mask_path": "assets/cat_11/v_50MmJszF2XM_3018_3318/0009s_00268_mask.png", "mask_status": "available" } ], "obj_id": [ 0 ], "anno_id": [ 727 ], "fps_inferred_from_clip_interval": 30.0 }, "usage": { "input_tokens": 37961, "input_tokens_details": { "cache_write_tokens": 37958, "cached_tokens": 0 }, "output_tokens": 4155, "output_tokens_details": { "reasoning_tokens": 2588 }, "total_tokens": 42116 }, "model": "gpt-5.6-sol", "effort": "medium", "custom_id": "cat11_cap08_v_50MmJszF2XM_3018_3318", "annotation": { "caption": "While standing, she practices arm movements, transitioning from a high V to a position with bent elbows and hands near her head. She then extends her arms back up into the high V and holds the pose.", "entity_groups": [ { "id": "E1", "label": "person", "spans": [ { "text": "she", "range": [ 16, 19 ] }, { "text": "her", "range": [ 119, 122 ] }, { "text": "She", "range": [ 129, 132 ] }, { "text": "her", "range": [ 146, 149 ] } ], "source_mask_link": "selected_person" }, { "id": "E2", "label": "elbows", "spans": [ { "text": "bent elbows", "range": [ 92, 103 ] } ], "source_mask_link": null }, { "id": "E3", "label": "hands", "spans": [ { "text": "hands", "range": [ 108, 113 ] } ], "source_mask_link": null }, { "id": "E4", "label": "head", "spans": [ { "text": "her head", "range": [ 119, 127 ] } ], "source_mask_link": null }, { "id": "E5", "label": "arms", "spans": [ { "text": "her arms", "range": [ 146, 154 ] } ], "source_mask_link": null } ], "actions": [ { "id": "A1", "mentions": [ { "text": "standing", "ranges": [ [ 6, 14 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "type": "posture", "subject_refs": [ "E1" ], "object_refs": [], "confidence": "high", "temporal_status": "localized", "intervals": [ { "start_s": 0.0, "end_s": 10.0, "source_start_s": 100.71, "source_end_s": 110.71 } ], "evidence": "The masked foreground child remains upright on her feet in every sampled frame from 0.000 through 9.000 seconds.", "mask_relation": "subject matches the supplied selected-person mask", "source_subject_refs": "E1" }, { "id": "A2", "mentions": [ { "text": "practices arm movements", "ranges": [ [ 20, 43 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "type": "light_verb", "subject_refs": [ "E1" ], "object_refs": [ "E5" ], "confidence": "high", "temporal_status": "localized", "intervals": [ { "start_s": 0.0, "end_s": 4.0, "source_start_s": 100.71, "source_end_s": 104.71 } ], "evidence": "Across samples 0.000–3.000, the masked child changes her arm configuration from raised arms to bent elbows with hands near her head.", "mask_relation": "subject matches the supplied selected-person mask", "source_subject_refs": "E1" }, { "id": "A3", "mentions": [ { "text": "transitioning", "ranges": [ [ 45, 58 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "type": "prepositional", "subject_refs": [ "E1" ], "object_refs": [], "confidence": "high", "temporal_status": "localized", "intervals": [ { "start_s": 0.0, "end_s": 4.0, "source_start_s": 100.71, "source_end_s": 104.71 } ], "evidence": "The masked child's arms progressively move downward and bend between samples 0.000 and 3.000.", "mask_relation": "subject matches the supplied selected-person mask", "source_subject_refs": "E1" }, { "id": "A4", "mentions": [ { "text": "extends", "ranges": [ [ 138, 145 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "type": "simple", "subject_refs": [ "E1" ], "object_refs": [ "E5" ], "confidence": "high", "temporal_status": "localized", "intervals": [ { "start_s": 3.0, "end_s": 6.0, "source_start_s": 103.71, "source_end_s": 106.71 } ], "evidence": "The masked child's arms move from the bent position at 3.000 seconds to a fully raised position by 5.000 seconds.", "mask_relation": "subject matches the supplied selected-person mask", "source_subject_refs": "E1" }, { "id": "A5", "mentions": [ { "text": "holds the pose", "ranges": [ [ 183, 197 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "type": "light_verb", "subject_refs": [ "E1" ], "object_refs": [], "confidence": "high", "temporal_status": "localized", "intervals": [ { "start_s": 5.0, "end_s": 10.0, "source_start_s": 105.71, "source_end_s": 110.71 } ], "evidence": "The masked child maintains both arms raised in the high-V configuration in samples 5.000 through 9.000, continuing to the clip boundary.", "mask_relation": "subject matches the supplied selected-person mask", "source_subject_refs": "E1" } ], "notes": [ "The supplied mask consistently identifies the foreground child, who is the caption's 'she'; the adult behind her is visible but is not mentioned by the caption.", "Only the person has a supplied mask. The elbows, hands, head, and arms are required text entities but have no independent source masks.", "The phrases 'a high V', 'a position', and 'the pose' denote configurations or abstract posture descriptions rather than separately segmentable entities." ] }, "model_output_raw": { "caption": "While standing, she practices arm movements, transitioning from a high V to a position with bent elbows and hands near her head. She then extends her arms back up into the high V and holds the pose.", "entity_groups": [ { "id": "E1", "label": "person", "spans": [ { "text": "she", "range": [ 16, 19 ] }, { "text": "her", "range": [ 119, 122 ] }, { "text": "She", "range": [ 129, 132 ] }, { "text": "her", "range": [ 146, 149 ] } ], "source_mask": "selected_person_mask" }, { "id": "E2", "label": "elbows", "spans": [ { "text": "bent elbows", "range": [ 92, 103 ] } ], "source_mask": null }, { "id": "E3", "label": "hands", "spans": [ { "text": "hands", "range": [ 108, 113 ] } ], "source_mask": null }, { "id": "E4", "label": "head", "spans": [ { "text": "her head", "range": [ 119, 127 ] } ], "source_mask": null }, { "id": "E5", "label": "arms", "spans": [ { "text": "her arms", "range": [ 146, 154 ] } ], "source_mask": null } ], "actions": [ { "id": "A1", "mentions": [ { "text": "standing", "ranges": [ [ 6, 14 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "type": "posture", "subject": "E1", "objects": [], "confidence": "high", "mask_relation": "subject matches the supplied selected-person mask", "temporal_status": "localized", "intervals": [ { "start_s": 0.0, "end_s": 10.0, "source_start_s": 100.71, "source_end_s": 110.71 } ], "evidence": "The masked foreground child remains upright on her feet in every sampled frame from 0.000 through 9.000 seconds." }, { "id": "A2", "mentions": [ { "text": "practices arm movements", "ranges": [ [ 20, 43 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "type": "light_verb", "subject": "E1", "objects": [ "E5" ], "confidence": "high", "mask_relation": "subject matches the supplied selected-person mask", "temporal_status": "localized", "intervals": [ { "start_s": 0.0, "end_s": 4.0, "source_start_s": 100.71, "source_end_s": 104.71 } ], "evidence": "Across samples 0.000–3.000, the masked child changes her arm configuration from raised arms to bent elbows with hands near her head." }, { "id": "A3", "mentions": [ { "text": "transitioning", "ranges": [ [ 45, 58 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "type": "prepositional", "subject": "E1", "objects": [], "confidence": "high", "mask_relation": "subject matches the supplied selected-person mask", "temporal_status": "localized", "intervals": [ { "start_s": 0.0, "end_s": 4.0, "source_start_s": 100.71, "source_end_s": 104.71 } ], "evidence": "The masked child's arms progressively move downward and bend between samples 0.000 and 3.000." }, { "id": "A4", "mentions": [ { "text": "extends", "ranges": [ [ 138, 145 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "type": "simple", "subject": "E1", "objects": [ "E5" ], "confidence": "high", "mask_relation": "subject matches the supplied selected-person mask", "temporal_status": "localized", "intervals": [ { "start_s": 3.0, "end_s": 6.0, "source_start_s": 103.71, "source_end_s": 106.71 } ], "evidence": "The masked child's arms move from the bent position at 3.000 seconds to a fully raised position by 5.000 seconds." }, { "id": "A5", "mentions": [ { "text": "holds the pose", "ranges": [ [ 183, 197 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "type": "light_verb", "subject": "E1", "objects": [], "confidence": "high", "mask_relation": "subject matches the supplied selected-person mask", "temporal_status": "localized", "intervals": [ { "start_s": 5.0, "end_s": 10.0, "source_start_s": 105.71, "source_end_s": 110.71 } ], "evidence": "The masked child maintains both arms raised in the high-V configuration in samples 5.000 through 9.000, continuing to the clip boundary." } ], "notes": [ "The supplied mask consistently identifies the foreground child, who is the caption's 'she'; the adult behind her is visible but is not mentioned by the caption.", "Only the person has a supplied mask. The elbows, hands, head, and arms are required text entities but have no independent source masks.", "The phrases 'a high V', 'a position', and 'the pose' denote configurations or abstract posture descriptions rather than separately segmentable entities." ] }, "normalization_repairs": 0 }, { "source_video_id": "v_50MmJszF2XM", "caption_count_category": 11, "clip": { "caption_index": 9, "caption": "The girl performs a back walkover while being assisted by an adult kneeling beside her. The adult provides support at her back and leg as she arches backward, places her hands on the floor, and kicks over. She stands up and finishes with her arms raised.", "source_interval_sec": [ 110.71, 128.71 ], "loom_clip_id": "v_50MmJszF2XM_3318_3857", "clip_frame_interval_4fps": [ 3318, 3857 ], "sample_count_1fps": 19, "sampled_frames": [ { "frame": "00000", "timestamp_sec": 110.71, "offset_sec": 0, "rgb_path": "assets/cat_11/v_50MmJszF2XM_3318_3857/0000s_00000_rgb.jpg", "mask_path": "assets/cat_11/v_50MmJszF2XM_3318_3857/0000s_00000_mask.png", "mask_status": "available" }, { "frame": "00028", "timestamp_sec": 111.71, "offset_sec": 1, "rgb_path": "assets/cat_11/v_50MmJszF2XM_3318_3857/0001s_00028_rgb.jpg", "mask_path": "assets/cat_11/v_50MmJszF2XM_3318_3857/0001s_00028_mask.png", "mask_status": "available" }, { "frame": "00060", "timestamp_sec": 112.71, "offset_sec": 2, "rgb_path": "assets/cat_11/v_50MmJszF2XM_3318_3857/0002s_00060_rgb.jpg", "mask_path": "assets/cat_11/v_50MmJszF2XM_3318_3857/0002s_00060_mask.png", "mask_status": "available" }, { "frame": "00088", "timestamp_sec": 113.71, "offset_sec": 3, "rgb_path": "assets/cat_11/v_50MmJszF2XM_3318_3857/0003s_00088_rgb.jpg", "mask_path": "assets/cat_11/v_50MmJszF2XM_3318_3857/0003s_00088_mask.png", "mask_status": "available" }, { "frame": "00120", "timestamp_sec": 114.71, "offset_sec": 4, "rgb_path": "assets/cat_11/v_50MmJszF2XM_3318_3857/0004s_00120_rgb.jpg", "mask_path": "assets/cat_11/v_50MmJszF2XM_3318_3857/0004s_00120_mask.png", "mask_status": "available" }, { "frame": "00148", "timestamp_sec": 115.71, "offset_sec": 5, "rgb_path": "assets/cat_11/v_50MmJszF2XM_3318_3857/0005s_00148_rgb.jpg", "mask_path": "assets/cat_11/v_50MmJszF2XM_3318_3857/0005s_00148_mask.png", "mask_status": "available" }, { "frame": "00180", "timestamp_sec": 116.71, "offset_sec": 6, "rgb_path": "assets/cat_11/v_50MmJszF2XM_3318_3857/0006s_00180_rgb.jpg", "mask_path": "assets/cat_11/v_50MmJszF2XM_3318_3857/0006s_00180_mask.png", "mask_status": "available" }, { "frame": "00208", "timestamp_sec": 117.71, "offset_sec": 7, "rgb_path": "assets/cat_11/v_50MmJszF2XM_3318_3857/0007s_00208_rgb.jpg", "mask_path": "assets/cat_11/v_50MmJszF2XM_3318_3857/0007s_00208_mask.png", "mask_status": "available" }, { "frame": "00240", "timestamp_sec": 118.71, "offset_sec": 8, "rgb_path": "assets/cat_11/v_50MmJszF2XM_3318_3857/0008s_00240_rgb.jpg", "mask_path": "assets/cat_11/v_50MmJszF2XM_3318_3857/0008s_00240_mask.png", "mask_status": "available" }, { "frame": "00268", "timestamp_sec": 119.71, "offset_sec": 9, "rgb_path": "assets/cat_11/v_50MmJszF2XM_3318_3857/0009s_00268_rgb.jpg", "mask_path": "assets/cat_11/v_50MmJszF2XM_3318_3857/0009s_00268_mask.png", "mask_status": "available" }, { "frame": "00300", "timestamp_sec": 120.71, "offset_sec": 10, "rgb_path": "assets/cat_11/v_50MmJszF2XM_3318_3857/0010s_00300_rgb.jpg", "mask_path": "assets/cat_11/v_50MmJszF2XM_3318_3857/0010s_00300_mask.png", "mask_status": "available" }, { "frame": "00328", "timestamp_sec": 121.71, "offset_sec": 11, "rgb_path": "assets/cat_11/v_50MmJszF2XM_3318_3857/0011s_00328_rgb.jpg", "mask_path": "assets/cat_11/v_50MmJszF2XM_3318_3857/0011s_00328_mask.png", "mask_status": "available" }, { "frame": "00360", "timestamp_sec": 122.71, "offset_sec": 12, "rgb_path": "assets/cat_11/v_50MmJszF2XM_3318_3857/0012s_00360_rgb.jpg", "mask_path": "assets/cat_11/v_50MmJszF2XM_3318_3857/0012s_00360_mask.png", "mask_status": "available" }, { "frame": "00388", "timestamp_sec": 123.71, "offset_sec": 13, "rgb_path": "assets/cat_11/v_50MmJszF2XM_3318_3857/0013s_00388_rgb.jpg", "mask_path": "assets/cat_11/v_50MmJszF2XM_3318_3857/0013s_00388_mask.png", "mask_status": "available" }, { "frame": "00420", "timestamp_sec": 124.71, "offset_sec": 14, "rgb_path": "assets/cat_11/v_50MmJszF2XM_3318_3857/0014s_00420_rgb.jpg", "mask_path": "assets/cat_11/v_50MmJszF2XM_3318_3857/0014s_00420_mask.png", "mask_status": "available" }, { "frame": "00448", "timestamp_sec": 125.71, "offset_sec": 15, "rgb_path": "assets/cat_11/v_50MmJszF2XM_3318_3857/0015s_00448_rgb.jpg", "mask_path": "assets/cat_11/v_50MmJszF2XM_3318_3857/0015s_00448_mask.png", "mask_status": "available" }, { "frame": "00480", "timestamp_sec": 126.71, "offset_sec": 16, "rgb_path": "assets/cat_11/v_50MmJszF2XM_3318_3857/0016s_00480_rgb.jpg", "mask_path": "assets/cat_11/v_50MmJszF2XM_3318_3857/0016s_00480_mask.png", "mask_status": "available" }, { "frame": "00508", "timestamp_sec": 127.71, "offset_sec": 17, "rgb_path": "assets/cat_11/v_50MmJszF2XM_3318_3857/0017s_00508_rgb.jpg", "mask_path": "assets/cat_11/v_50MmJszF2XM_3318_3857/0017s_00508_mask.png", "mask_status": "available" }, { "frame": "00536", "timestamp_sec": 128.71, "offset_sec": 18.0, "rgb_path": "assets/cat_11/v_50MmJszF2XM_3318_3857/0018s_00536_rgb.jpg", "mask_path": "assets/cat_11/v_50MmJszF2XM_3318_3857/0018s_00536_mask.png", "mask_status": "available" } ], "obj_id": [ 0 ], "anno_id": [ 728 ], "fps_inferred_from_clip_interval": 29.9444 }, "usage": { "input_tokens": 58340, "input_tokens_details": { "cache_write_tokens": 58337, "cached_tokens": 0 }, "output_tokens": 5000, "output_tokens_details": { "reasoning_tokens": 3011 }, "total_tokens": 63340 }, "model": "gpt-5.6-sol", "effort": "medium", "custom_id": "cat11_cap09_v_50MmJszF2XM_3318_3857", "annotation": { "caption": "The girl performs a back walkover while being assisted by an adult kneeling beside her. The adult provides support at her back and leg as she arches backward, places her hands on the floor, and kicks over. She stands up and finishes with her arms raised.", "entity_groups": [ { "id": "E1", "label": "girl", "spans": [ { "text": "The girl", "range": [ 0, 8 ] }, { "text": "her", "range": [ 83, 86 ] }, { "text": "her", "range": [ 118, 121 ] }, { "text": "she", "range": [ 138, 141 ] }, { "text": "her", "range": [ 166, 169 ] }, { "text": "She", "range": [ 206, 209 ] }, { "text": "her", "range": [ 238, 241 ] } ], "source_mask_link": "selected_person" }, { "id": "E2", "label": "adult", "spans": [ { "text": "an adult", "range": [ 58, 66 ] }, { "text": "The adult", "range": [ 88, 97 ] } ], "source_mask_link": null }, { "id": "E3", "label": "back", "spans": [ { "text": "her back", "range": [ 118, 126 ] } ], "source_mask_link": null }, { "id": "E4", "label": "leg", "spans": [ { "text": "leg", "range": [ 131, 134 ] } ], "source_mask_link": null }, { "id": "E5", "label": "hands", "spans": [ { "text": "her hands", "range": [ 166, 175 ] } ], "source_mask_link": null }, { "id": "E6", "label": "floor", "spans": [ { "text": "the floor", "range": [ 179, 188 ] } ], "source_mask_link": null }, { "id": "E7", "label": "arms", "spans": [ { "text": "her arms", "range": [ 238, 246 ] } ], "source_mask_link": null } ], "actions": [ { "id": "A1", "mentions": [ { "text": "performs a back walkover", "ranges": [ [ 9, 33 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "type": "light_verb", "subject_refs": [ "E1" ], "object_refs": [], "confidence": "high", "temporal_status": "localized", "intervals": [ { "start_s": 7.0, "end_s": 18.000000000000014, "source_start_s": 117.71, "source_end_s": 128.71 } ], "evidence": "The masked girl bends into the maneuver at 7–9 seconds, arches and becomes supported horizontally at 10–11 seconds, then passes her leg overhead while her hands are down through the clip boundary.", "mask_relation": "Action subject matches the supplied selected-person mask.", "source_subject_refs": [ "E1" ] }, { "id": "A2", "mentions": [ { "text": "being assisted", "ranges": [ [ 40, 54 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "type": "passive", "subject_refs": [ "E1" ], "object_refs": [], "confidence": "high", "temporal_status": "localized", "intervals": [ { "start_s": 0.0, "end_s": 18.000000000000014, "source_start_s": 110.71, "source_end_s": 128.71 } ], "evidence": "Across the sampled clip, the adult visibly holds or guides the masked girl at her waist, back, or leg.", "mask_relation": "Action subject matches the supplied selected-person mask.", "source_subject_refs": [ "E1" ] }, { "id": "A3", "mentions": [ { "text": "arches", "ranges": [ [ 142, 148 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "type": "simple", "subject_refs": [ "E1" ], "object_refs": [], "confidence": "high", "temporal_status": "localized", "intervals": [ { "start_s": 10.0, "end_s": 12.0, "source_start_s": 120.71, "source_end_s": 122.71 } ], "evidence": "At 10–11 seconds, the masked girl's torso is visibly curved backward as the adult supports her.", "mask_relation": "Action subject matches the supplied selected-person mask.", "source_subject_refs": [ "E1" ] }, { "id": "A4", "mentions": [ { "text": "places", "ranges": [ [ 159, 165 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "type": "simple", "subject_refs": [ "E1" ], "object_refs": [ "E5" ], "confidence": "high", "temporal_status": "localized", "intervals": [ { "start_s": 11.0, "end_s": 13.0, "source_start_s": 121.71, "source_end_s": 123.71 } ], "evidence": "The masked girl's hands descend toward the floor at 11 seconds and are visibly planted by 12 seconds.", "mask_relation": "Action subject matches the supplied selected-person mask.", "source_subject_refs": [ "E1" ] }, { "id": "A5", "mentions": [ { "text": "kicks over", "ranges": [ [ 194, 204 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "type": "phrasal", "subject_refs": [ "E1" ], "object_refs": [], "confidence": "high", "temporal_status": "localized", "intervals": [ { "start_s": 10.0, "end_s": 18.000000000000014, "source_start_s": 120.71, "source_end_s": 128.71 } ], "evidence": "From 10 seconds onward, the masked girl's legs swing upward and over her supported arched body.", "mask_relation": "Action subject matches the supplied selected-person mask.", "source_subject_refs": [ "E1" ] }, { "id": "A6", "mentions": [ { "text": "stands up", "ranges": [ [ 210, 219 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "type": "phrasal", "subject_refs": [ "E1" ], "object_refs": [], "confidence": "high", "temporal_status": "uncertain", "intervals": [], "evidence": "The frame at the 18-second right boundary shows the masked girl bent and beginning to recover, but no interior sampled frame clearly shows the captioned final rise to standing.", "mask_relation": "Action subject matches the supplied selected-person mask.", "source_subject_refs": [ "E1" ] } ], "notes": [ "The supplied masklet tracks the girl, including during overlap with the adult; it does not provide a separate mask for the adult.", "The adult's actions \"kneeling\" and \"provides\" were omitted because the adult has no corresponding supplied mask.", "The body parts and floor are text-groundable entities but have no separate supplied masks.", "\"finishes\" was excluded as a completion/aspectual predicate, and \"raised\" describes a resulting arm state rather than asserting a separate raising event.", "The sampled frame at clip-relative 18.0 seconds coincides with the half-open clip endpoint and therefore cannot support a positive-duration interval beyond the source interval." ] }, "model_output_raw": { "caption": "The girl performs a back walkover while being assisted by an adult kneeling beside her. The adult provides support at her back and leg as she arches backward, places her hands on the floor, and kicks over. She stands up and finishes with her arms raised.", "entity_groups": [ { "id": "E1", "entity": "girl", "spans": [ { "text": "The girl", "range": [ 0, 8 ] }, { "text": "her", "range": [ 83, 86 ] }, { "text": "her", "range": [ 118, 121 ] }, { "text": "she", "range": [ 138, 141 ] }, { "text": "her", "range": [ 166, 169 ] }, { "text": "She", "range": [ 206, 209 ] }, { "text": "her", "range": [ 238, 241 ] } ], "source_mask": "selected_person_masklet" }, { "id": "E2", "entity": "adult", "spans": [ { "text": "an adult", "range": [ 58, 66 ] }, { "text": "The adult", "range": [ 88, 97 ] } ], "source_mask": null }, { "id": "E3", "entity": "back", "spans": [ { "text": "her back", "range": [ 118, 126 ] } ], "source_mask": null }, { "id": "E4", "entity": "leg", "spans": [ { "text": "leg", "range": [ 131, 134 ] } ], "source_mask": null }, { "id": "E5", "entity": "hands", "spans": [ { "text": "her hands", "range": [ 166, 175 ] } ], "source_mask": null }, { "id": "E6", "entity": "floor", "spans": [ { "text": "the floor", "range": [ 179, 188 ] } ], "source_mask": null }, { "id": "E7", "entity": "arms", "spans": [ { "text": "her arms", "range": [ 238, 246 ] } ], "source_mask": null } ], "actions": [ { "id": "A1", "mentions": [ { "text": "performs a back walkover", "ranges": [ [ 9, 33 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "type": "light_verb", "subject_refs": [ "E1" ], "object_refs": [], "confidence": "high", "temporal_status": "localized", "intervals": [ { "clip_relative_s": [ 7.0, 18.0 ], "source_video_s": [ 117.71, 128.71 ] } ], "evidence": "The masked girl bends into the maneuver at 7–9 seconds, arches and becomes supported horizontally at 10–11 seconds, then passes her leg overhead while her hands are down through the clip boundary.", "mask_relation": "Action subject matches the supplied selected-person mask." }, { "id": "A2", "mentions": [ { "text": "being assisted", "ranges": [ [ 40, 54 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "type": "passive", "subject_refs": [ "E1" ], "object_refs": [], "agent_refs": [ "E2" ], "confidence": "high", "temporal_status": "localized", "intervals": [ { "clip_relative_s": [ 0.0, 18.0 ], "source_video_s": [ 110.71, 128.71 ] } ], "evidence": "Across the sampled clip, the adult visibly holds or guides the masked girl at her waist, back, or leg.", "mask_relation": "Action subject matches the supplied selected-person mask." }, { "id": "A3", "mentions": [ { "text": "arches", "ranges": [ [ 142, 148 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "type": "simple", "subject_refs": [ "E1" ], "object_refs": [], "confidence": "high", "temporal_status": "localized", "intervals": [ { "clip_relative_s": [ 10.0, 12.0 ], "source_video_s": [ 120.71, 122.71 ] } ], "evidence": "At 10–11 seconds, the masked girl's torso is visibly curved backward as the adult supports her.", "mask_relation": "Action subject matches the supplied selected-person mask." }, { "id": "A4", "mentions": [ { "text": "places", "ranges": [ [ 159, 165 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "type": "simple", "subject_refs": [ "E1" ], "object_refs": [ "E5" ], "location_refs": [ "E6" ], "confidence": "high", "temporal_status": "localized", "intervals": [ { "clip_relative_s": [ 11.0, 13.0 ], "source_video_s": [ 121.71, 123.71 ] } ], "evidence": "The masked girl's hands descend toward the floor at 11 seconds and are visibly planted by 12 seconds.", "mask_relation": "Action subject matches the supplied selected-person mask." }, { "id": "A5", "mentions": [ { "text": "kicks over", "ranges": [ [ 194, 204 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "type": "phrasal", "subject_refs": [ "E1" ], "object_refs": [], "confidence": "high", "temporal_status": "localized", "intervals": [ { "clip_relative_s": [ 10.0, 18.0 ], "source_video_s": [ 120.71, 128.71 ] } ], "evidence": "From 10 seconds onward, the masked girl's legs swing upward and over her supported arched body.", "mask_relation": "Action subject matches the supplied selected-person mask." }, { "id": "A6", "mentions": [ { "text": "stands up", "ranges": [ [ 210, 219 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "type": "phrasal", "subject_refs": [ "E1" ], "object_refs": [], "confidence": "high", "temporal_status": "uncertain", "intervals": [], "evidence": "The frame at the 18-second right boundary shows the masked girl bent and beginning to recover, but no interior sampled frame clearly shows the captioned final rise to standing.", "mask_relation": "Action subject matches the supplied selected-person mask." } ], "notes": [ "The supplied masklet tracks the girl, including during overlap with the adult; it does not provide a separate mask for the adult.", "The adult's actions \"kneeling\" and \"provides\" were omitted because the adult has no corresponding supplied mask.", "The body parts and floor are text-groundable entities but have no separate supplied masks.", "\"finishes\" was excluded as a completion/aspectual predicate, and \"raised\" describes a resulting arm state rather than asserting a separate raising event.", "The sampled frame at clip-relative 18.0 seconds coincides with the half-open clip endpoint and therefore cannot support a positive-duration interval beyond the source interval." ] }, "normalization_repairs": 0 }, { "source_video_id": "v_50MmJszF2XM", "caption_count_category": 11, "clip": { "caption_index": 10, "caption": "After finishing the assisted move, she lowers her arms and walks a few steps away. She turns to face the adult and stands still.", "source_interval_sec": [ 128.71, 136.71 ], "loom_clip_id": "v_50MmJszF2XM_3857_4097", "clip_frame_interval_4fps": [ 3857, 4097 ], "sample_count_1fps": 8, "sampled_frames": [ { "frame": "00000", "timestamp_sec": 128.71, "offset_sec": 0, "rgb_path": "assets/cat_11/v_50MmJszF2XM_3857_4097/0000s_00000_rgb.jpg", "mask_path": "assets/cat_11/v_50MmJszF2XM_3857_4097/0000s_00000_mask.png", "mask_status": "available" }, { "frame": "00028", "timestamp_sec": 129.71, "offset_sec": 1, "rgb_path": "assets/cat_11/v_50MmJszF2XM_3857_4097/0001s_00028_rgb.jpg", "mask_path": "assets/cat_11/v_50MmJszF2XM_3857_4097/0001s_00028_mask.png", "mask_status": "available" }, { "frame": "00060", "timestamp_sec": 130.71, "offset_sec": 2, "rgb_path": "assets/cat_11/v_50MmJszF2XM_3857_4097/0002s_00060_rgb.jpg", "mask_path": "assets/cat_11/v_50MmJszF2XM_3857_4097/0002s_00060_mask.png", "mask_status": "available" }, { "frame": "00088", "timestamp_sec": 131.71, "offset_sec": 3, "rgb_path": "assets/cat_11/v_50MmJszF2XM_3857_4097/0003s_00088_rgb.jpg", "mask_path": "assets/cat_11/v_50MmJszF2XM_3857_4097/0003s_00088_mask.png", "mask_status": "available" }, { "frame": "00120", "timestamp_sec": 132.71, "offset_sec": 4, "rgb_path": "assets/cat_11/v_50MmJszF2XM_3857_4097/0004s_00120_rgb.jpg", "mask_path": "assets/cat_11/v_50MmJszF2XM_3857_4097/0004s_00120_mask.png", "mask_status": "available" }, { "frame": "00148", "timestamp_sec": 133.71, "offset_sec": 5, "rgb_path": "assets/cat_11/v_50MmJszF2XM_3857_4097/0005s_00148_rgb.jpg", "mask_path": "assets/cat_11/v_50MmJszF2XM_3857_4097/0005s_00148_mask.png", "mask_status": "available" }, { "frame": "00180", "timestamp_sec": 134.71, "offset_sec": 6, "rgb_path": "assets/cat_11/v_50MmJszF2XM_3857_4097/0006s_00180_rgb.jpg", "mask_path": "assets/cat_11/v_50MmJszF2XM_3857_4097/0006s_00180_mask.png", "mask_status": "available" }, { "frame": "00208", "timestamp_sec": 135.71, "offset_sec": 7, "rgb_path": "assets/cat_11/v_50MmJszF2XM_3857_4097/0007s_00208_rgb.jpg", "mask_path": "assets/cat_11/v_50MmJszF2XM_3857_4097/0007s_00208_mask.png", "mask_status": "available" } ], "obj_id": [ 0 ], "anno_id": [ 729 ], "fps_inferred_from_clip_interval": 30.0 }, "usage": { "input_tokens": 33421, "input_tokens_details": { "cache_write_tokens": 33418, "cached_tokens": 0 }, "output_tokens": 4124, "output_tokens_details": { "reasoning_tokens": 2588 }, "total_tokens": 37545 }, "model": "gpt-5.6-sol", "effort": "medium", "custom_id": "cat11_cap10_v_50MmJszF2XM_3857_4097", "annotation": { "caption": "After finishing the assisted move, she lowers her arms and walks a few steps away. She turns to face the adult and stands still.", "entity_groups": [ { "id": "person_1", "label": "girl", "spans": [ { "text": "she", "range": [ 35, 38 ] }, { "text": "her", "range": [ 46, 49 ] }, { "text": "She", "range": [ 83, 86 ] } ], "source_mask_link": "selected_person" }, { "id": "arms_1", "label": "arms", "spans": [ { "text": "her arms", "range": [ 46, 54 ] } ], "source_mask_link": null }, { "id": "person_2", "label": "adult", "spans": [ { "text": "the adult", "range": [ 101, 110 ] } ], "source_mask_link": null } ], "actions": [ { "id": "action_1", "mentions": [ { "text": "lowers", "ranges": [ [ 39, 45 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "type": "simple", "subject_refs": [ "person_1" ], "object_refs": [ "arms_1" ], "confidence": "high", "temporal_status": "supported", "intervals": [ { "start_s": 1.0, "end_s": 4.0, "source_start_s": 129.71, "source_end_s": 132.71 } ], "evidence": "At local 1.0 s the masked girl has both arms overhead; at 2.0–3.0 s they descend toward shoulder and side level.", "mask_relation": "The action subject matches the supplied selected-person mask.", "source_subject_refs": [ "person_1" ] }, { "id": "action_2", "mentions": [ { "text": "walks", "ranges": [ [ 59, 64 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "type": "simple", "subject_refs": [ "person_1" ], "object_refs": [], "confidence": "medium", "temporal_status": "supported", "intervals": [ { "start_s": 1.0, "end_s": 3.0, "source_start_s": 129.71, "source_end_s": 131.71 } ], "evidence": "At local 1.0–2.0 s the masked girl's crossed and changing leg positions support a short stepping movement.", "mask_relation": "The action subject matches the supplied selected-person mask.", "source_subject_refs": [ "person_1" ] }, { "id": "action_3", "mentions": [ { "text": "turns", "ranges": [ [ 87, 92 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "type": "simple", "subject_refs": [ "person_1" ], "object_refs": [], "confidence": "medium", "temporal_status": "supported", "intervals": [ { "start_s": 2.0, "end_s": 4.0, "source_start_s": 130.71, "source_end_s": 132.71 } ], "evidence": "Across local 2.0–3.0 s the masked girl's body and head orientation change toward the adult on her right.", "mask_relation": "The action subject matches the supplied selected-person mask.", "source_subject_refs": [ "person_1" ] }, { "id": "action_4", "mentions": [ { "text": "face", "ranges": [ [ 96, 100 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "type": "posture", "subject_refs": [ "person_1" ], "object_refs": [ "person_2" ], "confidence": "high", "temporal_status": "supported", "intervals": [ { "start_s": 3.0, "end_s": 4.0, "source_start_s": 131.71, "source_end_s": 132.71 }, { "start_s": 6.0, "end_s": 8.0, "source_start_s": 134.71, "source_end_s": 136.71 } ], "evidence": "At local 3.0 s and again at 6.0–7.0 s, the masked girl is oriented toward the adult standing to her right.", "mask_relation": "The action subject matches the supplied selected-person mask.", "source_subject_refs": [ "person_1" ] }, { "id": "action_5", "mentions": [ { "text": "stands", "ranges": [ [ 115, 121 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "type": "posture", "subject_refs": [ "person_1" ], "object_refs": [], "confidence": "high", "temporal_status": "supported", "intervals": [ { "start_s": 6.0, "end_s": 8.0, "source_start_s": 134.71, "source_end_s": 136.71 } ], "evidence": "At local 6.0 and 7.0 s the masked girl remains upright in essentially the same location; support reaches the clip boundary.", "mask_relation": "The action subject matches the supplied selected-person mask.", "source_subject_refs": [ "person_1" ] } ], "notes": [ "The supplied mask consistently identifies the younger person on the left, corresponding to both occurrences of “she” and the possessive “her.”", "The adult is visible but has no supplied instance mask, so her entity group has a null source-mask link.", "The arms are separately annotated as a body-part entity but have no independent supplied mask.", "“finishing” is omitted as an aspectual predicate framing the nominal event “the assisted move”; the generic event noun is not a concrete entity.", "The walking and turning intervals are medium-confidence sparse support inferred from changes across sampled clean frames." ] }, "model_output_raw": { "caption": "After finishing the assisted move, she lowers her arms and walks a few steps away. She turns to face the adult and stands still.", "entity_groups": [ { "id": "person_1", "label": "girl", "spans": [ { "text": "she", "range": [ 35, 38 ] }, { "text": "her", "range": [ 46, 49 ] }, { "text": "She", "range": [ 83, 86 ] } ], "source_mask": "selected_person_masklet" }, { "id": "arms_1", "label": "arms", "spans": [ { "text": "her arms", "range": [ 46, 54 ] } ], "source_mask": null }, { "id": "person_2", "label": "adult", "spans": [ { "text": "the adult", "range": [ 101, 110 ] } ], "source_mask": null } ], "actions": [ { "id": "action_1", "mentions": [ { "text": "lowers", "ranges": [ [ 39, 45 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "type": "simple", "subject_refs": [ "person_1" ], "object_refs": [ "arms_1" ], "confidence": "high", "temporal_status": "supported", "intervals": [ { "start_s": 1.0, "end_s": 4.0, "source_start_s": 129.71, "source_end_s": 132.71 } ], "evidence": "At local 1.0 s the masked girl has both arms overhead; at 2.0–3.0 s they descend toward shoulder and side level.", "mask_relation": "The action subject matches the supplied selected-person mask." }, { "id": "action_2", "mentions": [ { "text": "walks", "ranges": [ [ 59, 64 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "type": "simple", "subject_refs": [ "person_1" ], "object_refs": [], "confidence": "medium", "temporal_status": "supported", "intervals": [ { "start_s": 1.0, "end_s": 3.0, "source_start_s": 129.71, "source_end_s": 131.71 } ], "evidence": "At local 1.0–2.0 s the masked girl's crossed and changing leg positions support a short stepping movement.", "mask_relation": "The action subject matches the supplied selected-person mask." }, { "id": "action_3", "mentions": [ { "text": "turns", "ranges": [ [ 87, 92 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "type": "simple", "subject_refs": [ "person_1" ], "object_refs": [], "confidence": "medium", "temporal_status": "supported", "intervals": [ { "start_s": 2.0, "end_s": 4.0, "source_start_s": 130.71, "source_end_s": 132.71 } ], "evidence": "Across local 2.0–3.0 s the masked girl's body and head orientation change toward the adult on her right.", "mask_relation": "The action subject matches the supplied selected-person mask." }, { "id": "action_4", "mentions": [ { "text": "face", "ranges": [ [ 96, 100 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "type": "posture", "subject_refs": [ "person_1" ], "object_refs": [ "person_2" ], "confidence": "high", "temporal_status": "supported", "intervals": [ { "start_s": 3.0, "end_s": 4.0, "source_start_s": 131.71, "source_end_s": 132.71 }, { "start_s": 6.0, "end_s": 8.0, "source_start_s": 134.71, "source_end_s": 136.71 } ], "evidence": "At local 3.0 s and again at 6.0–7.0 s, the masked girl is oriented toward the adult standing to her right.", "mask_relation": "The action subject matches the supplied selected-person mask." }, { "id": "action_5", "mentions": [ { "text": "stands", "ranges": [ [ 115, 121 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "type": "posture", "subject_refs": [ "person_1" ], "object_refs": [], "confidence": "high", "temporal_status": "supported", "intervals": [ { "start_s": 6.0, "end_s": 8.0, "source_start_s": 134.71, "source_end_s": 136.71 } ], "evidence": "At local 6.0 and 7.0 s the masked girl remains upright in essentially the same location; support reaches the clip boundary.", "mask_relation": "The action subject matches the supplied selected-person mask." } ], "notes": [ "The supplied mask consistently identifies the younger person on the left, corresponding to both occurrences of “she” and the possessive “her.”", "The adult is visible but has no supplied instance mask, so her entity group has a null source-mask link.", "The arms are separately annotated as a body-part entity but have no independent supplied mask.", "“finishing” is omitted as an aspectual predicate framing the nominal event “the assisted move”; the generic event noun is not a concrete entity.", "The walking and turning intervals are medium-confidence sparse support inferred from changes across sampled clean frames." ] }, "normalization_repairs": 0 }, { "source_video_id": "v_50MmJszF2XM", "caption_count_category": 11, "clip": { "caption_index": 11, "caption": "The girl performs an unassisted back walkover, starting from a lunge position. She arches her back, places her hands on the floor, kicks her legs over, and lands on her feet. She concludes by standing with her arms raised before lowering them.", "source_interval_sec": [ 136.71, 149.68 ], "loom_clip_id": "v_50MmJszF2XM_4097_4486", "clip_frame_interval_4fps": [ 4097, 4486 ], "sample_count_1fps": 13, "sampled_frames": [ { "frame": "00000", "timestamp_sec": 136.71, "offset_sec": 0, "rgb_path": "assets/cat_11/v_50MmJszF2XM_4097_4486/0000s_00000_rgb.jpg", "mask_path": "assets/cat_11/v_50MmJszF2XM_4097_4486/0000s_00000_mask.png", "mask_status": "available" }, { "frame": "00028", "timestamp_sec": 137.71, "offset_sec": 1, "rgb_path": "assets/cat_11/v_50MmJszF2XM_4097_4486/0001s_00028_rgb.jpg", "mask_path": "assets/cat_11/v_50MmJszF2XM_4097_4486/0001s_00028_mask.png", "mask_status": "available" }, { "frame": "00060", "timestamp_sec": 138.71, "offset_sec": 2, "rgb_path": "assets/cat_11/v_50MmJszF2XM_4097_4486/0002s_00060_rgb.jpg", "mask_path": "assets/cat_11/v_50MmJszF2XM_4097_4486/0002s_00060_mask.png", "mask_status": "available" }, { "frame": "00088", "timestamp_sec": 139.71, "offset_sec": 3, "rgb_path": "assets/cat_11/v_50MmJszF2XM_4097_4486/0003s_00088_rgb.jpg", "mask_path": "assets/cat_11/v_50MmJszF2XM_4097_4486/0003s_00088_mask.png", "mask_status": "available" }, { "frame": "00120", "timestamp_sec": 140.71, "offset_sec": 4, "rgb_path": "assets/cat_11/v_50MmJszF2XM_4097_4486/0004s_00120_rgb.jpg", "mask_path": "assets/cat_11/v_50MmJszF2XM_4097_4486/0004s_00120_mask.png", "mask_status": "available" }, { "frame": "00148", "timestamp_sec": 141.71, "offset_sec": 5, "rgb_path": "assets/cat_11/v_50MmJszF2XM_4097_4486/0005s_00148_rgb.jpg", "mask_path": "assets/cat_11/v_50MmJszF2XM_4097_4486/0005s_00148_mask.png", "mask_status": "available" }, { "frame": "00180", "timestamp_sec": 142.71, "offset_sec": 6, "rgb_path": "assets/cat_11/v_50MmJszF2XM_4097_4486/0006s_00180_rgb.jpg", "mask_path": "assets/cat_11/v_50MmJszF2XM_4097_4486/0006s_00180_mask.png", "mask_status": "available" }, { "frame": "00208", "timestamp_sec": 143.71, "offset_sec": 7, "rgb_path": "assets/cat_11/v_50MmJszF2XM_4097_4486/0007s_00208_rgb.jpg", "mask_path": "assets/cat_11/v_50MmJszF2XM_4097_4486/0007s_00208_mask.png", "mask_status": "available" }, { "frame": "00240", "timestamp_sec": 144.71, "offset_sec": 8, "rgb_path": "assets/cat_11/v_50MmJszF2XM_4097_4486/0008s_00240_rgb.jpg", "mask_path": "assets/cat_11/v_50MmJszF2XM_4097_4486/0008s_00240_mask.png", "mask_status": "available" }, { "frame": "00268", "timestamp_sec": 145.71, "offset_sec": 9, "rgb_path": "assets/cat_11/v_50MmJszF2XM_4097_4486/0009s_00268_rgb.jpg", "mask_path": "assets/cat_11/v_50MmJszF2XM_4097_4486/0009s_00268_mask.png", "mask_status": "available" }, { "frame": "00300", "timestamp_sec": 146.71, "offset_sec": 10, "rgb_path": "assets/cat_11/v_50MmJszF2XM_4097_4486/0010s_00300_rgb.jpg", "mask_path": "assets/cat_11/v_50MmJszF2XM_4097_4486/0010s_00300_mask.png", "mask_status": "available" }, { "frame": "00328", "timestamp_sec": 147.71, "offset_sec": 11, "rgb_path": "assets/cat_11/v_50MmJszF2XM_4097_4486/0011s_00328_rgb.jpg", "mask_path": "assets/cat_11/v_50MmJszF2XM_4097_4486/0011s_00328_mask.png", "mask_status": "available" }, { "frame": "00360", "timestamp_sec": 148.71, "offset_sec": 12, "rgb_path": "assets/cat_11/v_50MmJszF2XM_4097_4486/0012s_00360_rgb.jpg", "mask_path": "assets/cat_11/v_50MmJszF2XM_4097_4486/0012s_00360_mask.png", "mask_status": "available" } ], "obj_id": [ 0 ], "anno_id": [ 730 ], "fps_inferred_from_clip_interval": 29.9923 }, "usage": { "input_tokens": 44762, "input_tokens_details": { "cache_write_tokens": 44759, "cached_tokens": 0 }, "output_tokens": 4649, "output_tokens_details": { "reasoning_tokens": 2588 }, "total_tokens": 49411 }, "model": "gpt-5.6-sol", "effort": "medium", "custom_id": "cat11_cap11_v_50MmJszF2XM_4097_4486", "annotation": { "caption": "The girl performs an unassisted back walkover, starting from a lunge position. She arches her back, places her hands on the floor, kicks her legs over, and lands on her feet. She concludes by standing with her arms raised before lowering them.", "entity_groups": [ { "id": "girl", "label": "girl", "spans": [ { "text": "The girl", "range": [ 0, 8 ] }, { "text": "She", "range": [ 79, 82 ] }, { "text": "her", "range": [ 90, 93 ] }, { "text": "her", "range": [ 107, 110 ] }, { "text": "her", "range": [ 137, 140 ] }, { "text": "her", "range": [ 165, 168 ] }, { "text": "She", "range": [ 175, 178 ] }, { "text": "her", "range": [ 206, 209 ] } ], "source_mask_link": "selected_person" }, { "id": "back", "label": "back", "spans": [ { "text": "her back", "range": [ 90, 98 ] } ], "source_mask_link": null }, { "id": "hands", "label": "hands", "spans": [ { "text": "her hands", "range": [ 107, 116 ] } ], "source_mask_link": null }, { "id": "floor", "label": "floor", "spans": [ { "text": "the floor", "range": [ 120, 129 ] } ], "source_mask_link": null }, { "id": "legs", "label": "legs", "spans": [ { "text": "her legs", "range": [ 137, 145 ] } ], "source_mask_link": null }, { "id": "feet", "label": "feet", "spans": [ { "text": "her feet", "range": [ 165, 173 ] } ], "source_mask_link": null }, { "id": "arms", "label": "arms", "spans": [ { "text": "her arms", "range": [ 206, 214 ] }, { "text": "them", "range": [ 238, 242 ] } ], "source_mask_link": null } ], "actions": [ { "id": "action_1", "mentions": [ { "text": "performs an unassisted back walkover", "ranges": [ [ 9, 45 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "type": "light_verb", "subject_refs": [ "girl" ], "object_refs": [], "confidence": "high", "temporal_status": "localized", "intervals": [ { "start_s": 9.0, "end_s": 12.0, "source_start_s": 145.71, "source_end_s": 148.71 } ], "evidence": "The masked girl enters a lunge at 9.0 s, moves through the maneuver at 10.0–11.0 s, and is upright again by 12.0 s.", "mask_relation": "subject matches the supplied selected-person mask", "source_subject_refs": "girl" }, { "id": "action_2", "mentions": [ { "text": "arches", "ranges": [ [ 83, 89 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "type": "simple", "subject_refs": [ "girl" ], "object_refs": [ "back" ], "confidence": "high", "temporal_status": "localized", "intervals": [ { "start_s": 11.0, "end_s": 12.0, "source_start_s": 147.71, "source_end_s": 148.71 } ], "evidence": "At 11.0 s the masked girl is visibly leaning backward with her back arched.", "mask_relation": "subject matches the supplied selected-person mask", "source_subject_refs": "girl" }, { "id": "action_3", "mentions": [ { "text": "places", "ranges": [ [ 100, 106 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "type": "simple", "subject_refs": [ "girl" ], "object_refs": [ "hands" ], "confidence": "high", "temporal_status": "uncertain", "intervals": [], "evidence": "No sampled clean frame directly shows her hands contacting the floor; this phase likely occurs between samples.", "mask_relation": "subject matches the supplied selected-person mask", "source_subject_refs": "girl" }, { "id": "action_4", "mentions": [ { "text": "kicks", "ranges": [ [ 131, 136 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "type": "simple", "subject_refs": [ "girl" ], "object_refs": [ "legs" ], "confidence": "high", "temporal_status": "uncertain", "intervals": [], "evidence": "The sampled frames do not directly capture the leg kick-over phase.", "mask_relation": "subject matches the supplied selected-person mask", "source_subject_refs": "girl" }, { "id": "action_5", "mentions": [ { "text": "lands", "ranges": [ [ 156, 161 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "type": "simple", "subject_refs": [ "girl" ], "object_refs": [], "confidence": "high", "temporal_status": "uncertain", "intervals": [], "evidence": "The girl is upright on her feet at 12.0 s, but no sampled frame directly shows the landing event.", "mask_relation": "subject matches the supplied selected-person mask", "source_subject_refs": "girl" }, { "id": "action_6", "mentions": [ { "text": "standing", "ranges": [ [ 192, 200 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "type": "posture", "subject_refs": [ "girl" ], "object_refs": [], "confidence": "high", "temporal_status": "localized", "intervals": [ { "start_s": 12.0, "end_s": 12.969999999999999, "source_start_s": 148.71, "source_end_s": 149.68 } ], "evidence": "The final sampled frame at 12.0 s shows the masked girl standing upright, with support continuing to the clip boundary.", "mask_relation": "subject matches the supplied selected-person mask", "source_subject_refs": "girl" }, { "id": "action_7", "mentions": [ { "text": "lowering", "ranges": [ [ 229, 237 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "type": "simple", "subject_refs": [ "girl" ], "object_refs": [ "arms" ], "confidence": "high", "temporal_status": "localized", "intervals": [ { "start_s": 11.0, "end_s": 12.0, "source_start_s": 147.71, "source_end_s": 148.71 } ], "evidence": "Her arms are overhead at 11.0 s and lowered by 12.0 s.", "mask_relation": "subject matches the supplied selected-person mask", "source_subject_refs": "girl" } ], "notes": [ "The supplied mask consistently identifies the foreground girl in black and purple, not the adult woman standing behind her.", "No separate masks were supplied for the girl's body parts or the floor, so those entity groups have null source-mask links.", "\"starting\" and \"concludes\" are excluded as aspectual/control predicates. \"raised\" describes the resulting held state of the arms and is not separately annotated as an event.", "The hand placement, leg kick-over, and landing phases are not directly visible in the sparse one-frame-per-second samples, so their temporal intervals are left empty rather than inferred." ] }, "model_output_raw": { "caption": "The girl performs an unassisted back walkover, starting from a lunge position. She arches her back, places her hands on the floor, kicks her legs over, and lands on her feet. She concludes by standing with her arms raised before lowering them.", "entity_groups": [ { "id": "girl", "label": "girl", "spans": [ { "text": "The girl", "range": [ 0, 8 ] }, { "text": "She", "range": [ 79, 82 ] }, { "text": "her", "range": [ 90, 93 ] }, { "text": "her", "range": [ 107, 110 ] }, { "text": "her", "range": [ 137, 140 ] }, { "text": "her", "range": [ 165, 168 ] }, { "text": "She", "range": [ 175, 178 ] }, { "text": "her", "range": [ 206, 209 ] } ], "source_mask": "selected_person_masklet" }, { "id": "back", "label": "back", "spans": [ { "text": "her back", "range": [ 90, 98 ] } ], "source_mask": null }, { "id": "hands", "label": "hands", "spans": [ { "text": "her hands", "range": [ 107, 116 ] } ], "source_mask": null }, { "id": "floor", "label": "floor", "spans": [ { "text": "the floor", "range": [ 120, 129 ] } ], "source_mask": null }, { "id": "legs", "label": "legs", "spans": [ { "text": "her legs", "range": [ 137, 145 ] } ], "source_mask": null }, { "id": "feet", "label": "feet", "spans": [ { "text": "her feet", "range": [ 165, 173 ] } ], "source_mask": null }, { "id": "arms", "label": "arms", "spans": [ { "text": "her arms", "range": [ 206, 214 ] }, { "text": "them", "range": [ 238, 242 ] } ], "source_mask": null } ], "actions": [ { "id": "action_1", "mentions": [ { "text": "performs an unassisted back walkover", "ranges": [ [ 9, 45 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "type": "light_verb", "subject_entity_group_id": "girl", "object_entity_group_ids": [], "confidence": "high", "mask_relation": "subject matches the supplied selected-person mask", "temporal_status": "localized", "intervals": [ { "start_s": 9.0, "end_s": 12.0 } ], "evidence": "The masked girl enters a lunge at 9.0 s, moves through the maneuver at 10.0–11.0 s, and is upright again by 12.0 s." }, { "id": "action_2", "mentions": [ { "text": "arches", "ranges": [ [ 83, 89 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "type": "simple", "subject_entity_group_id": "girl", "object_entity_group_ids": [ "back" ], "confidence": "high", "mask_relation": "subject matches the supplied selected-person mask", "temporal_status": "localized", "intervals": [ { "start_s": 11.0, "end_s": 12.0 } ], "evidence": "At 11.0 s the masked girl is visibly leaning backward with her back arched." }, { "id": "action_3", "mentions": [ { "text": "places", "ranges": [ [ 100, 106 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "type": "simple", "subject_entity_group_id": "girl", "object_entity_group_ids": [ "hands" ], "confidence": "high", "mask_relation": "subject matches the supplied selected-person mask", "temporal_status": "uncertain", "intervals": [], "evidence": "No sampled clean frame directly shows her hands contacting the floor; this phase likely occurs between samples." }, { "id": "action_4", "mentions": [ { "text": "kicks", "ranges": [ [ 131, 136 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "type": "simple", "subject_entity_group_id": "girl", "object_entity_group_ids": [ "legs" ], "confidence": "high", "mask_relation": "subject matches the supplied selected-person mask", "temporal_status": "uncertain", "intervals": [], "evidence": "The sampled frames do not directly capture the leg kick-over phase." }, { "id": "action_5", "mentions": [ { "text": "lands", "ranges": [ [ 156, 161 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "type": "simple", "subject_entity_group_id": "girl", "object_entity_group_ids": [], "confidence": "high", "mask_relation": "subject matches the supplied selected-person mask", "temporal_status": "uncertain", "intervals": [], "evidence": "The girl is upright on her feet at 12.0 s, but no sampled frame directly shows the landing event." }, { "id": "action_6", "mentions": [ { "text": "standing", "ranges": [ [ 192, 200 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "type": "posture", "subject_entity_group_id": "girl", "object_entity_group_ids": [], "confidence": "high", "mask_relation": "subject matches the supplied selected-person mask", "temporal_status": "localized", "intervals": [ { "start_s": 12.0, "end_s": 12.97 } ], "evidence": "The final sampled frame at 12.0 s shows the masked girl standing upright, with support continuing to the clip boundary." }, { "id": "action_7", "mentions": [ { "text": "lowering", "ranges": [ [ 229, 237 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "type": "simple", "subject_entity_group_id": "girl", "object_entity_group_ids": [ "arms" ], "confidence": "high", "mask_relation": "subject matches the supplied selected-person mask", "temporal_status": "localized", "intervals": [ { "start_s": 11.0, "end_s": 12.0 } ], "evidence": "Her arms are overhead at 11.0 s and lowered by 12.0 s." } ], "notes": [ "The supplied mask consistently identifies the foreground girl in black and purple, not the adult woman standing behind her.", "No separate masks were supplied for the girl's body parts or the floor, so those entity groups have null source-mask links.", "\"starting\" and \"concludes\" are excluded as aspectual/control predicates. \"raised\" describes the resulting held state of the arms and is not separately annotated as an event.", "The hand placement, leg kick-over, and landing phases are not directly visible in the sparse one-frame-per-second samples, so their temporal intervals are left empty rather than inferred." ] }, "normalization_repairs": 0 }, { "source_video_id": "v_-YMpwZkNc2A", "caption_count_category": 12, "clip": { "caption_index": 1, "caption": "A person with a beard and wearing a blue t-shirt appears and looks down. They run their hand through their hair before looking forward and talking. They then gesture with their left hand while speaking.", "source_interval_sec": [ 2.07, 16.57 ], "loom_clip_id": "v_-YMpwZkNc2A_62_497", "clip_frame_interval_4fps": [ 62, 497 ], "sample_count_1fps": 15, "sampled_frames": [ { "frame": "00000", "timestamp_sec": 2.07, "offset_sec": 0, "rgb_path": "assets/cat_12/v_-YMpwZkNc2A_62_497/0000s_00000_rgb.jpg", "mask_path": "assets/cat_12/v_-YMpwZkNc2A_62_497/0000s_00000_mask.png", "mask_status": "available" }, { "frame": "00028", "timestamp_sec": 3.07, "offset_sec": 1, "rgb_path": "assets/cat_12/v_-YMpwZkNc2A_62_497/0001s_00028_rgb.jpg", "mask_path": "assets/cat_12/v_-YMpwZkNc2A_62_497/0001s_00028_mask.png", "mask_status": "available" }, { "frame": "00060", "timestamp_sec": 4.07, "offset_sec": 2, "rgb_path": "assets/cat_12/v_-YMpwZkNc2A_62_497/0002s_00060_rgb.jpg", "mask_path": "assets/cat_12/v_-YMpwZkNc2A_62_497/0002s_00060_mask.png", "mask_status": "available" }, { "frame": "00088", "timestamp_sec": 5.07, "offset_sec": 3, "rgb_path": "assets/cat_12/v_-YMpwZkNc2A_62_497/0003s_00088_rgb.jpg", "mask_path": "assets/cat_12/v_-YMpwZkNc2A_62_497/0003s_00088_mask.png", "mask_status": "available" }, { "frame": "00120", "timestamp_sec": 6.07, "offset_sec": 4, "rgb_path": "assets/cat_12/v_-YMpwZkNc2A_62_497/0004s_00120_rgb.jpg", "mask_path": "assets/cat_12/v_-YMpwZkNc2A_62_497/0004s_00120_mask.png", "mask_status": "available" }, { "frame": "00148", "timestamp_sec": 7.07, "offset_sec": 5, "rgb_path": "assets/cat_12/v_-YMpwZkNc2A_62_497/0005s_00148_rgb.jpg", "mask_path": "assets/cat_12/v_-YMpwZkNc2A_62_497/0005s_00148_mask.png", "mask_status": "available" }, { "frame": "00180", "timestamp_sec": 8.07, "offset_sec": 6, "rgb_path": "assets/cat_12/v_-YMpwZkNc2A_62_497/0006s_00180_rgb.jpg", "mask_path": "assets/cat_12/v_-YMpwZkNc2A_62_497/0006s_00180_mask.png", "mask_status": "available" }, { "frame": "00208", "timestamp_sec": 9.07, "offset_sec": 7, "rgb_path": "assets/cat_12/v_-YMpwZkNc2A_62_497/0007s_00208_rgb.jpg", "mask_path": "assets/cat_12/v_-YMpwZkNc2A_62_497/0007s_00208_mask.png", "mask_status": "available" }, { "frame": "00240", "timestamp_sec": 10.07, "offset_sec": 8, "rgb_path": "assets/cat_12/v_-YMpwZkNc2A_62_497/0008s_00240_rgb.jpg", "mask_path": "assets/cat_12/v_-YMpwZkNc2A_62_497/0008s_00240_mask.png", "mask_status": "available" }, { "frame": "00268", "timestamp_sec": 11.07, "offset_sec": 9, "rgb_path": "assets/cat_12/v_-YMpwZkNc2A_62_497/0009s_00268_rgb.jpg", "mask_path": "assets/cat_12/v_-YMpwZkNc2A_62_497/0009s_00268_mask.png", "mask_status": "available" }, { "frame": "00300", "timestamp_sec": 12.07, "offset_sec": 10, "rgb_path": "assets/cat_12/v_-YMpwZkNc2A_62_497/0010s_00300_rgb.jpg", "mask_path": "assets/cat_12/v_-YMpwZkNc2A_62_497/0010s_00300_mask.png", "mask_status": "available" }, { "frame": "00328", "timestamp_sec": 13.07, "offset_sec": 11, "rgb_path": "assets/cat_12/v_-YMpwZkNc2A_62_497/0011s_00328_rgb.jpg", "mask_path": "assets/cat_12/v_-YMpwZkNc2A_62_497/0011s_00328_mask.png", "mask_status": "available" }, { "frame": "00360", "timestamp_sec": 14.07, "offset_sec": 12, "rgb_path": "assets/cat_12/v_-YMpwZkNc2A_62_497/0012s_00360_rgb.jpg", "mask_path": "assets/cat_12/v_-YMpwZkNc2A_62_497/0012s_00360_mask.png", "mask_status": "available" }, { "frame": "00388", "timestamp_sec": 15.07, "offset_sec": 13, "rgb_path": "assets/cat_12/v_-YMpwZkNc2A_62_497/0013s_00388_rgb.jpg", "mask_path": "assets/cat_12/v_-YMpwZkNc2A_62_497/0013s_00388_mask.png", "mask_status": "available" }, { "frame": "00420", "timestamp_sec": 16.07, "offset_sec": 14, "rgb_path": "assets/cat_12/v_-YMpwZkNc2A_62_497/0014s_00420_rgb.jpg", "mask_path": "assets/cat_12/v_-YMpwZkNc2A_62_497/0014s_00420_mask.png", "mask_status": "available" } ], "obj_id": [ 0 ], "anno_id": [ 91 ], "fps_inferred_from_clip_interval": 30.0 }, "usage": { "input_tokens": 49274, "input_tokens_details": { "cache_write_tokens": 49271, "cached_tokens": 0 }, "output_tokens": 4678, "output_tokens_details": { "reasoning_tokens": 2588 }, "total_tokens": 53952 }, "model": "gpt-5.6-sol", "effort": "medium", "custom_id": "cat12_cap01_v_-YMpwZkNc2A_62_497", "annotation": { "caption": "A person with a beard and wearing a blue t-shirt appears and looks down. They run their hand through their hair before looking forward and talking. They then gesture with their left hand while speaking.", "entity_groups": [ { "id": "person_1", "label": "person", "spans": [ { "text": "A person", "range": [ 0, 8 ] }, { "text": "They", "range": [ 73, 77 ] }, { "text": "their", "range": [ 82, 87 ] }, { "text": "their", "range": [ 101, 106 ] }, { "text": "They", "range": [ 148, 152 ] }, { "text": "their", "range": [ 171, 176 ] } ], "source_mask_link": "selected_person" }, { "id": "beard_1", "label": "beard", "spans": [ { "text": "a beard", "range": [ 14, 21 ] } ], "source_mask_link": null }, { "id": "tshirt_1", "label": "t-shirt", "spans": [ { "text": "a blue t-shirt", "range": [ 34, 48 ] } ], "source_mask_link": null }, { "id": "hand_1", "label": "hand", "spans": [ { "text": "their hand", "range": [ 82, 92 ] } ], "source_mask_link": null }, { "id": "hair_1", "label": "hair", "spans": [ { "text": "their hair", "range": [ 101, 111 ] } ], "source_mask_link": null }, { "id": "left_hand_1", "label": "left hand", "spans": [ { "text": "their left hand", "range": [ 171, 186 ] } ], "source_mask_link": null } ], "actions": [ { "id": "action_1", "mentions": [ { "text": "wearing", "ranges": [ [ 26, 33 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "type": "simple", "subject_refs": [ "person_1" ], "object_refs": [ "tshirt_1" ], "confidence": "high", "temporal_status": "localized_sparse", "intervals": [ { "start_s": 1.0, "end_s": 14.5, "source_start_s": 3.07, "source_end_s": 16.57 } ], "evidence": "The mask-linked person visibly wears the blue t-shirt in every sampled frame from 1.0 seconds through the clip boundary.", "mask_relation": "subject matches the supplied selected-person mask", "source_subject_refs": "person_1" }, { "id": "action_2", "mentions": [ { "text": "looks", "ranges": [ [ 61, 66 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "type": "simple", "subject_refs": [ "person_1" ], "object_refs": [], "confidence": "high", "temporal_status": "localized_sparse", "intervals": [ { "start_s": 1.0, "end_s": 3.0, "source_start_s": 3.07, "source_end_s": 5.07 }, { "start_s": 4.0, "end_s": 8.0, "source_start_s": 6.07, "source_end_s": 10.07 }, { "start_s": 9.0, "end_s": 11.0, "source_start_s": 11.07, "source_end_s": 13.07 } ], "evidence": "The selected person repeatedly holds his head and gaze downward in the cited sampled frames; the directional word “down” is excluded from the span.", "mask_relation": "subject matches the supplied selected-person mask", "source_subject_refs": "person_1" }, { "id": "action_3", "mentions": [ { "text": "run", "ranges": [ [ 78, 81 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "type": "simple", "subject_refs": [ "person_1" ], "object_refs": [ "hand_1", "hair_1" ], "confidence": "high", "temporal_status": "localized_sparse", "intervals": [ { "start_s": 1.0, "end_s": 2.0, "source_start_s": 3.07, "source_end_s": 4.07 } ], "evidence": "At 1.0 seconds, the selected person has a hand pressed into and moving across the front of his hair.", "mask_relation": "subject matches the supplied selected-person mask", "source_subject_refs": "person_1" }, { "id": "action_4", "mentions": [ { "text": "looking", "ranges": [ [ 119, 126 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "type": "simple", "subject_refs": [ "person_1" ], "object_refs": [], "confidence": "high", "temporal_status": "localized_sparse", "intervals": [ { "start_s": 3.0, "end_s": 4.0, "source_start_s": 5.07, "source_end_s": 6.07 }, { "start_s": 8.0, "end_s": 9.0, "source_start_s": 10.07, "source_end_s": 11.07 } ], "evidence": "At 3.0 and 8.0 seconds, the mask-linked person raises his gaze toward the forward-facing viewpoint; “forward” is excluded as a directional modifier.", "mask_relation": "subject matches the supplied selected-person mask", "source_subject_refs": "person_1" }, { "id": "action_5", "mentions": [ { "text": "talking", "ranges": [ [ 139, 146 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } }, { "text": "speaking", "ranges": [ [ 193, 201 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "type": "simple", "subject_refs": [ "person_1" ], "object_refs": [], "confidence": "medium", "temporal_status": "localized_sparse", "intervals": [ { "start_s": 4.0, "end_s": 14.5, "source_start_s": 6.07, "source_end_s": 16.57 } ], "evidence": "Across the later samples, the selected person shows changing open-mouth speech configurations; “speaking” continues the same talking event during the gesture.", "mask_relation": "subject matches the supplied selected-person mask", "source_subject_refs": "person_1" }, { "id": "action_6", "mentions": [ { "text": "gesture", "ranges": [ [ 158, 165 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "type": "simple", "subject_refs": [ "person_1" ], "object_refs": [ "left_hand_1" ], "confidence": "high", "temporal_status": "localized_sparse", "intervals": [ { "start_s": 11.0, "end_s": 14.5, "source_start_s": 13.07, "source_end_s": 16.57 } ], "evidence": "From 11.0 seconds through the clip boundary, the selected person raises and changes the position of the captioned left hand while speaking.", "mask_relation": "subject matches the supplied selected-person mask", "source_subject_refs": "person_1" } ], "notes": [ "The supplied masklet identifies the single captioned person from 1.0 seconds onward; the 0.0-second mask is empty because the person is not yet visible.", "No separate source masks were supplied for the beard, t-shirt, hands, or hair, so those entity groups have null source-mask links.", "The unspecified earlier hand and the later explicitly named left hand are kept as separate entity groups because the caption does not establish that they are the same hand.", "“appears” is excluded as a visibility/presence predicate rather than a physical action." ] }, "model_output_raw": { "caption": "A person with a beard and wearing a blue t-shirt appears and looks down. They run their hand through their hair before looking forward and talking. They then gesture with their left hand while speaking.", "entity_groups": [ { "id": "person_1", "label": "person", "spans": [ { "text": "A person", "range": [ 0, 8 ] }, { "text": "They", "range": [ 73, 77 ] }, { "text": "their", "range": [ 82, 87 ] }, { "text": "their", "range": [ 101, 106 ] }, { "text": "They", "range": [ 148, 152 ] }, { "text": "their", "range": [ 171, 176 ] } ], "source_mask": "selected_person_masklet" }, { "id": "beard_1", "label": "beard", "spans": [ { "text": "a beard", "range": [ 14, 21 ] } ], "source_mask": null }, { "id": "tshirt_1", "label": "t-shirt", "spans": [ { "text": "a blue t-shirt", "range": [ 34, 48 ] } ], "source_mask": null }, { "id": "hand_1", "label": "hand", "spans": [ { "text": "their hand", "range": [ 82, 92 ] } ], "source_mask": null }, { "id": "hair_1", "label": "hair", "spans": [ { "text": "their hair", "range": [ 101, 111 ] } ], "source_mask": null }, { "id": "left_hand_1", "label": "left hand", "spans": [ { "text": "their left hand", "range": [ 171, 186 ] } ], "source_mask": null } ], "actions": [ { "id": "action_1", "mentions": [ { "text": "wearing", "ranges": [ [ 26, 33 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "type": "simple", "subject_ref": "person_1", "object_refs": [ "tshirt_1" ], "confidence": "high", "mask_relation": "subject matches the supplied selected-person mask", "temporal_status": "localized_sparse", "intervals": [ { "clip_relative_s": [ 1.0, 14.5 ], "source_video_s": [ 3.07, 16.57 ] } ], "evidence": "The mask-linked person visibly wears the blue t-shirt in every sampled frame from 1.0 seconds through the clip boundary." }, { "id": "action_2", "mentions": [ { "text": "looks", "ranges": [ [ 61, 66 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "type": "simple", "subject_ref": "person_1", "object_refs": [], "confidence": "high", "mask_relation": "subject matches the supplied selected-person mask", "temporal_status": "localized_sparse", "intervals": [ { "clip_relative_s": [ 1.0, 3.0 ], "source_video_s": [ 3.07, 5.07 ] }, { "clip_relative_s": [ 4.0, 8.0 ], "source_video_s": [ 6.07, 10.07 ] }, { "clip_relative_s": [ 9.0, 11.0 ], "source_video_s": [ 11.07, 13.07 ] } ], "evidence": "The selected person repeatedly holds his head and gaze downward in the cited sampled frames; the directional word “down” is excluded from the span." }, { "id": "action_3", "mentions": [ { "text": "run", "ranges": [ [ 78, 81 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "type": "simple", "subject_ref": "person_1", "object_refs": [ "hand_1", "hair_1" ], "confidence": "high", "mask_relation": "subject matches the supplied selected-person mask", "temporal_status": "localized_sparse", "intervals": [ { "clip_relative_s": [ 1.0, 2.0 ], "source_video_s": [ 3.07, 4.07 ] } ], "evidence": "At 1.0 seconds, the selected person has a hand pressed into and moving across the front of his hair." }, { "id": "action_4", "mentions": [ { "text": "looking", "ranges": [ [ 119, 126 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "type": "simple", "subject_ref": "person_1", "object_refs": [], "confidence": "high", "mask_relation": "subject matches the supplied selected-person mask", "temporal_status": "localized_sparse", "intervals": [ { "clip_relative_s": [ 3.0, 4.0 ], "source_video_s": [ 5.07, 6.07 ] }, { "clip_relative_s": [ 8.0, 9.0 ], "source_video_s": [ 10.07, 11.07 ] } ], "evidence": "At 3.0 and 8.0 seconds, the mask-linked person raises his gaze toward the forward-facing viewpoint; “forward” is excluded as a directional modifier." }, { "id": "action_5", "mentions": [ { "text": "talking", "ranges": [ [ 139, 146 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } }, { "text": "speaking", "ranges": [ [ 193, 201 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "type": "simple", "subject_ref": "person_1", "object_refs": [], "confidence": "medium", "mask_relation": "subject matches the supplied selected-person mask", "temporal_status": "localized_sparse", "intervals": [ { "clip_relative_s": [ 4.0, 14.5 ], "source_video_s": [ 6.07, 16.57 ] } ], "evidence": "Across the later samples, the selected person shows changing open-mouth speech configurations; “speaking” continues the same talking event during the gesture." }, { "id": "action_6", "mentions": [ { "text": "gesture", "ranges": [ [ 158, 165 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "type": "simple", "subject_ref": "person_1", "object_refs": [ "left_hand_1" ], "confidence": "high", "mask_relation": "subject matches the supplied selected-person mask", "temporal_status": "localized_sparse", "intervals": [ { "clip_relative_s": [ 11.0, 14.5 ], "source_video_s": [ 13.07, 16.57 ] } ], "evidence": "From 11.0 seconds through the clip boundary, the selected person raises and changes the position of the captioned left hand while speaking." } ], "notes": [ "The supplied masklet identifies the single captioned person from 1.0 seconds onward; the 0.0-second mask is empty because the person is not yet visible.", "No separate source masks were supplied for the beard, t-shirt, hands, or hair, so those entity groups have null source-mask links.", "The unspecified earlier hand and the later explicitly named left hand are kept as separate entity groups because the caption does not establish that they are the same hand.", "“appears” is excluded as a visibility/presence predicate rather than a physical action." ] }, "normalization_repairs": 0 }, { "source_video_id": "v_-YMpwZkNc2A", "caption_count_category": 12, "clip": { "caption_index": 2, "caption": "The person picks up a pack of razor blades in their right hand and a razor in their left, holding them up. They speak while displaying the items before putting them down. They then move out of view.", "source_interval_sec": [ 16.57, 26.07 ], "loom_clip_id": "v_-YMpwZkNc2A_497_781", "clip_frame_interval_4fps": [ 497, 781 ], "sample_count_1fps": 10, "sampled_frames": [ { "frame": "00000", "timestamp_sec": 16.57, "offset_sec": 0, "rgb_path": "assets/cat_12/v_-YMpwZkNc2A_497_781/0000s_00000_rgb.jpg", "mask_path": "assets/cat_12/v_-YMpwZkNc2A_497_781/0000s_00000_mask.png", "mask_status": "available" }, { "frame": "00028", "timestamp_sec": 17.57, "offset_sec": 1, "rgb_path": "assets/cat_12/v_-YMpwZkNc2A_497_781/0001s_00028_rgb.jpg", "mask_path": "assets/cat_12/v_-YMpwZkNc2A_497_781/0001s_00028_mask.png", "mask_status": "available" }, { "frame": "00060", "timestamp_sec": 18.57, "offset_sec": 2, "rgb_path": "assets/cat_12/v_-YMpwZkNc2A_497_781/0002s_00060_rgb.jpg", "mask_path": "assets/cat_12/v_-YMpwZkNc2A_497_781/0002s_00060_mask.png", "mask_status": "available" }, { "frame": "00088", "timestamp_sec": 19.57, "offset_sec": 3, "rgb_path": "assets/cat_12/v_-YMpwZkNc2A_497_781/0003s_00088_rgb.jpg", "mask_path": "assets/cat_12/v_-YMpwZkNc2A_497_781/0003s_00088_mask.png", "mask_status": "available" }, { "frame": "00120", "timestamp_sec": 20.57, "offset_sec": 4, "rgb_path": "assets/cat_12/v_-YMpwZkNc2A_497_781/0004s_00120_rgb.jpg", "mask_path": "assets/cat_12/v_-YMpwZkNc2A_497_781/0004s_00120_mask.png", "mask_status": "available" }, { "frame": "00148", "timestamp_sec": 21.57, "offset_sec": 5, "rgb_path": "assets/cat_12/v_-YMpwZkNc2A_497_781/0005s_00148_rgb.jpg", "mask_path": "assets/cat_12/v_-YMpwZkNc2A_497_781/0005s_00148_mask.png", "mask_status": "available" }, { "frame": "00180", "timestamp_sec": 22.57, "offset_sec": 6, "rgb_path": "assets/cat_12/v_-YMpwZkNc2A_497_781/0006s_00180_rgb.jpg", "mask_path": "assets/cat_12/v_-YMpwZkNc2A_497_781/0006s_00180_mask.png", "mask_status": "available" }, { "frame": "00208", "timestamp_sec": 23.57, "offset_sec": 7, "rgb_path": "assets/cat_12/v_-YMpwZkNc2A_497_781/0007s_00208_rgb.jpg", "mask_path": "assets/cat_12/v_-YMpwZkNc2A_497_781/0007s_00208_mask.png", "mask_status": "available" }, { "frame": "00240", "timestamp_sec": 24.57, "offset_sec": 8, "rgb_path": "assets/cat_12/v_-YMpwZkNc2A_497_781/0008s_00240_rgb.jpg", "mask_path": "assets/cat_12/v_-YMpwZkNc2A_497_781/0008s_00240_mask.png", "mask_status": "available" }, { "frame": "00268", "timestamp_sec": 25.57, "offset_sec": 9, "rgb_path": "assets/cat_12/v_-YMpwZkNc2A_497_781/0009s_00268_rgb.jpg", "mask_path": "assets/cat_12/v_-YMpwZkNc2A_497_781/0009s_00268_mask.png", "mask_status": "available" } ], "obj_id": [ 0 ], "anno_id": [ 90 ], "fps_inferred_from_clip_interval": 29.8947 }, "usage": { "input_tokens": 37962, "input_tokens_details": { "cache_write_tokens": 37959, "cached_tokens": 0 }, "output_tokens": 5406, "output_tokens_details": { "reasoning_tokens": 3419 }, "total_tokens": 43368 }, "model": "gpt-5.6-sol", "effort": "medium", "custom_id": "cat12_cap02_v_-YMpwZkNc2A_497_781", "annotation": { "caption": "The person picks up a pack of razor blades in their right hand and a razor in their left, holding them up. They speak while displaying the items before putting them down. They then move out of view.", "entity_groups": [ { "id": "person_1", "label": "person", "spans": [ { "text": "The person", "range": [ 0, 10 ] }, { "text": "their", "range": [ 46, 51 ] }, { "text": "their", "range": [ 78, 83 ] }, { "text": "They", "range": [ 107, 111 ] }, { "text": "They", "range": [ 171, 175 ] } ], "source_mask_link": "selected_person" }, { "id": "pack_1", "label": "pack", "spans": [ { "text": "a pack of razor blades", "range": [ 20, 42 ] }, { "text": "them", "range": [ 98, 102 ] }, { "text": "the items", "range": [ 135, 144 ] }, { "text": "them", "range": [ 160, 164 ] } ], "source_mask_link": null }, { "id": "razor_blades_1", "label": "razor blades", "spans": [ { "text": "a pack of razor blades", "range": [ 20, 42 ] } ], "source_mask_link": null }, { "id": "right_hand_1", "label": "right hand", "spans": [ { "text": "their right hand", "range": [ 46, 62 ] } ], "source_mask_link": null }, { "id": "razor_1", "label": "razor", "spans": [ { "text": "a razor", "range": [ 67, 74 ] }, { "text": "them", "range": [ 98, 102 ] }, { "text": "the items", "range": [ 135, 144 ] }, { "text": "them", "range": [ 160, 164 ] } ], "source_mask_link": null }, { "id": "left_hand_1", "label": "left hand", "spans": [ { "text": "their left", "range": [ 78, 88 ] } ], "source_mask_link": null } ], "actions": [ { "id": "action_1", "mentions": [ { "text": "picks up", "ranges": [ [ 11, 19 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "type": "phrasal", "subject_refs": [ "person_1" ], "object_refs": [ "pack_1", "razor_1" ], "confidence": "high", "temporal_status": "supported_by_sparse_samples", "intervals": [ { "start_s": 0.0, "end_s": 2.0, "source_start_s": 16.57, "source_end_s": 18.57 } ], "evidence": "At 0.000 seconds the masked person has no clearly raised items; by 1.000–2.000 seconds he has brought the razor package and razor into his hands.", "mask_relation": "subject_matches_selected_person_mask", "source_subject_refs": "person_1" }, { "id": "action_2", "mentions": [ { "text": "holding", "ranges": [ [ 90, 97 ] ], "occurrence": 1, "particle": { "text": "up", "range": [ 103, 105 ], "occurrence": 2 } }, { "text": "displaying", "ranges": [ [ 124, 134 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "type": "phrasal", "subject_refs": [ "person_1" ], "object_refs": [ "pack_1", "razor_1" ], "confidence": "high", "temporal_status": "supported_by_sparse_samples", "intervals": [ { "start_s": 2.0, "end_s": 6.0, "source_start_s": 18.57, "source_end_s": 22.57 } ], "evidence": "From 2.000 through 5.000 seconds, the masked person visibly holds the razor and blade package up toward the viewer.", "mask_relation": "subject_matches_selected_person_mask", "source_subject_refs": "person_1" }, { "id": "action_3", "mentions": [ { "text": "speak", "ranges": [ [ 112, 117 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "type": "simple", "subject_refs": [ "person_1" ], "object_refs": [], "confidence": "medium", "temporal_status": "supported_by_sparse_samples", "intervals": [ { "start_s": 2.0, "end_s": 6.0, "source_start_s": 18.57, "source_end_s": 22.57 } ], "evidence": "While displaying the items at 2.000–5.000 seconds, the masked person's mouth appears in changing open speaking shapes; still frames provide only medium-confidence visual evidence.", "mask_relation": "subject_matches_selected_person_mask", "source_subject_refs": "person_1" }, { "id": "action_4", "mentions": [ { "text": "putting", "ranges": [ [ 152, 159 ] ], "occurrence": 1, "particle": { "text": "down", "range": [ 165, 169 ], "occurrence": 1 } } ], "type": "phrasal", "subject_refs": [ "person_1" ], "object_refs": [ "pack_1", "razor_1" ], "confidence": "medium", "temporal_status": "supported_by_sparse_transition", "intervals": [ { "start_s": 6.0, "end_s": 7.0, "source_start_s": 22.57, "source_end_s": 23.57 } ], "evidence": "The masked person still holds the package at 6.000 seconds, then looks downward with the items no longer raised at 7.000 seconds, supporting a put-down transition.", "mask_relation": "subject_matches_selected_person_mask", "source_subject_refs": "person_1" }, { "id": "action_5", "mentions": [ { "text": "move", "ranges": [ [ 181, 185 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "type": "simple", "subject_refs": [ "person_1" ], "object_refs": [], "confidence": "high", "temporal_status": "not_observed_in_samples", "intervals": [], "evidence": "The selected person remains substantially visible through the final sampled frame at 9.000 seconds, so no sampled frame supports movement out of view.", "mask_relation": "subject_matches_selected_person_mask", "source_subject_refs": "person_1" } ], "notes": [ "The supplied mask consistently identifies the caption's person and no other entity.", "No separate masks were supplied for the pack, razor blades, razor, or hands; their source-mask links are therefore null.", "The shared span \"a pack of razor blades\" is used for both the visible pack and its blade contents under the container/content grounding rule.", "The caption's final movement out of view is not visually supported by the sampled frames." ] }, "model_output_raw": { "caption": "The person picks up a pack of razor blades in their right hand and a razor in their left, holding them up. They speak while displaying the items before putting them down. They then move out of view.", "entity_groups": [ { "id": "person_1", "label": "person", "spans": [ { "text": "The person", "range": [ 0, 10 ] }, { "text": "their", "range": [ 46, 51 ] }, { "text": "their", "range": [ 78, 83 ] }, { "text": "They", "range": [ 107, 111 ] }, { "text": "They", "range": [ 171, 175 ] } ], "source_mask": { "track": "selected_person" } }, { "id": "pack_1", "label": "pack", "spans": [ { "text": "a pack of razor blades", "range": [ 20, 42 ] }, { "text": "them", "range": [ 98, 102 ] }, { "text": "the items", "range": [ 135, 144 ] }, { "text": "them", "range": [ 160, 164 ] } ], "source_mask": null }, { "id": "razor_blades_1", "label": "razor blades", "spans": [ { "text": "a pack of razor blades", "range": [ 20, 42 ] } ], "source_mask": null }, { "id": "right_hand_1", "label": "right hand", "spans": [ { "text": "their right hand", "range": [ 46, 62 ] } ], "source_mask": null }, { "id": "razor_1", "label": "razor", "spans": [ { "text": "a razor", "range": [ 67, 74 ] }, { "text": "them", "range": [ 98, 102 ] }, { "text": "the items", "range": [ 135, 144 ] }, { "text": "them", "range": [ 160, 164 ] } ], "source_mask": null }, { "id": "left_hand_1", "label": "left hand", "spans": [ { "text": "their left", "range": [ 78, 88 ] } ], "source_mask": null } ], "actions": [ { "id": "action_1", "mentions": [ { "text": "picks up", "ranges": [ [ 11, 19 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "type": "phrasal", "subject": "person_1", "objects": [ "pack_1", "razor_1" ], "confidence": "high", "mask_relation": "subject_matches_selected_person_mask", "temporal_status": "supported_by_sparse_samples", "intervals": [ { "clip_relative": { "start_s": 0.0, "end_s": 2.0 }, "source_video": { "start_s": 16.57, "end_s": 18.57 } } ], "evidence": "At 0.000 seconds the masked person has no clearly raised items; by 1.000–2.000 seconds he has brought the razor package and razor into his hands." }, { "id": "action_2", "mentions": [ { "text": "holding", "ranges": [ [ 90, 97 ] ], "occurrence": 1, "particle": { "text": "up", "range": [ 103, 105 ], "occurrence": 2 } }, { "text": "displaying", "ranges": [ [ 124, 134 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "type": "phrasal", "subject": "person_1", "objects": [ "pack_1", "razor_1" ], "confidence": "high", "mask_relation": "subject_matches_selected_person_mask", "temporal_status": "supported_by_sparse_samples", "intervals": [ { "clip_relative": { "start_s": 2.0, "end_s": 6.0 }, "source_video": { "start_s": 18.57, "end_s": 22.57 } } ], "evidence": "From 2.000 through 5.000 seconds, the masked person visibly holds the razor and blade package up toward the viewer." }, { "id": "action_3", "mentions": [ { "text": "speak", "ranges": [ [ 112, 117 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "type": "simple", "subject": "person_1", "objects": [], "confidence": "medium", "mask_relation": "subject_matches_selected_person_mask", "temporal_status": "supported_by_sparse_samples", "intervals": [ { "clip_relative": { "start_s": 2.0, "end_s": 6.0 }, "source_video": { "start_s": 18.57, "end_s": 22.57 } } ], "evidence": "While displaying the items at 2.000–5.000 seconds, the masked person's mouth appears in changing open speaking shapes; still frames provide only medium-confidence visual evidence." }, { "id": "action_4", "mentions": [ { "text": "putting", "ranges": [ [ 152, 159 ] ], "occurrence": 1, "particle": { "text": "down", "range": [ 165, 169 ], "occurrence": 1 } } ], "type": "phrasal", "subject": "person_1", "objects": [ "pack_1", "razor_1" ], "confidence": "medium", "mask_relation": "subject_matches_selected_person_mask", "temporal_status": "supported_by_sparse_transition", "intervals": [ { "clip_relative": { "start_s": 6.0, "end_s": 7.0 }, "source_video": { "start_s": 22.57, "end_s": 23.57 } } ], "evidence": "The masked person still holds the package at 6.000 seconds, then looks downward with the items no longer raised at 7.000 seconds, supporting a put-down transition." }, { "id": "action_5", "mentions": [ { "text": "move", "ranges": [ [ 181, 185 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "type": "simple", "subject": "person_1", "objects": [], "confidence": "high", "mask_relation": "subject_matches_selected_person_mask", "temporal_status": "not_observed_in_samples", "intervals": [], "evidence": "The selected person remains substantially visible through the final sampled frame at 9.000 seconds, so no sampled frame supports movement out of view." } ], "notes": [ "The supplied mask consistently identifies the caption's person and no other entity.", "No separate masks were supplied for the pack, razor blades, razor, or hands; their source-mask links are therefore null.", "The shared span \"a pack of razor blades\" is used for both the visible pack and its blade contents under the container/content grounding rule.", "The caption's final movement out of view is not visually supported by the sampled frames." ] }, "normalization_repairs": 0 }, { "source_video_id": "v_-YMpwZkNc2A", "caption_count_category": 12, "clip": { "caption_index": 3, "caption": "The person reappears and begins to apply white shaving cream to their face. They spread the cream over their cheeks, chin, and neck with their hands.", "source_interval_sec": [ 26.07, 34.57 ], "loom_clip_id": "v_-YMpwZkNc2A_781_1036", "clip_frame_interval_4fps": [ 781, 1036 ], "sample_count_1fps": 9, "sampled_frames": [ { "frame": "00000", "timestamp_sec": 26.07, "offset_sec": 0, "rgb_path": "assets/cat_12/v_-YMpwZkNc2A_781_1036/0000s_00000_rgb.jpg", "mask_path": "assets/cat_12/v_-YMpwZkNc2A_781_1036/0000s_00000_mask.png", "mask_status": "available" }, { "frame": "00028", "timestamp_sec": 27.07, "offset_sec": 1, "rgb_path": "assets/cat_12/v_-YMpwZkNc2A_781_1036/0001s_00028_rgb.jpg", "mask_path": "assets/cat_12/v_-YMpwZkNc2A_781_1036/0001s_00028_mask.png", "mask_status": "available" }, { "frame": "00060", "timestamp_sec": 28.07, "offset_sec": 2, "rgb_path": "assets/cat_12/v_-YMpwZkNc2A_781_1036/0002s_00060_rgb.jpg", "mask_path": "assets/cat_12/v_-YMpwZkNc2A_781_1036/0002s_00060_mask.png", "mask_status": "available" }, { "frame": "00088", "timestamp_sec": 29.07, "offset_sec": 3, "rgb_path": "assets/cat_12/v_-YMpwZkNc2A_781_1036/0003s_00088_rgb.jpg", "mask_path": "assets/cat_12/v_-YMpwZkNc2A_781_1036/0003s_00088_mask.png", "mask_status": "available" }, { "frame": "00120", "timestamp_sec": 30.07, "offset_sec": 4, "rgb_path": "assets/cat_12/v_-YMpwZkNc2A_781_1036/0004s_00120_rgb.jpg", "mask_path": "assets/cat_12/v_-YMpwZkNc2A_781_1036/0004s_00120_mask.png", "mask_status": "available" }, { "frame": "00148", "timestamp_sec": 31.07, "offset_sec": 5, "rgb_path": "assets/cat_12/v_-YMpwZkNc2A_781_1036/0005s_00148_rgb.jpg", "mask_path": "assets/cat_12/v_-YMpwZkNc2A_781_1036/0005s_00148_mask.png", "mask_status": "available" }, { "frame": "00180", "timestamp_sec": 32.07, "offset_sec": 6, "rgb_path": "assets/cat_12/v_-YMpwZkNc2A_781_1036/0006s_00180_rgb.jpg", "mask_path": "assets/cat_12/v_-YMpwZkNc2A_781_1036/0006s_00180_mask.png", "mask_status": "available" }, { "frame": "00208", "timestamp_sec": 33.07, "offset_sec": 7, "rgb_path": "assets/cat_12/v_-YMpwZkNc2A_781_1036/0007s_00208_rgb.jpg", "mask_path": "assets/cat_12/v_-YMpwZkNc2A_781_1036/0007s_00208_mask.png", "mask_status": "available" }, { "frame": "00240", "timestamp_sec": 34.07, "offset_sec": 8, "rgb_path": "assets/cat_12/v_-YMpwZkNc2A_781_1036/0008s_00240_rgb.jpg", "mask_path": "assets/cat_12/v_-YMpwZkNc2A_781_1036/0008s_00240_mask.png", "mask_status": "available" } ], "obj_id": [ 0 ], "anno_id": [ 92 ], "fps_inferred_from_clip_interval": 30.0 }, "usage": { "input_tokens": 35689, "input_tokens_details": { "cache_write_tokens": 35686, "cached_tokens": 0 }, "output_tokens": 3421, "output_tokens_details": { "reasoning_tokens": 2111 }, "total_tokens": 39110 }, "model": "gpt-5.6-sol", "effort": "medium", "custom_id": "cat12_cap03_v_-YMpwZkNc2A_781_1036", "annotation": { "caption": "The person reappears and begins to apply white shaving cream to their face. They spread the cream over their cheeks, chin, and neck with their hands.", "entity_groups": [ { "id": "E1", "label": "person", "spans": [ { "text": "The person", "range": [ 0, 10 ] }, { "text": "their", "range": [ 64, 69 ] }, { "text": "They", "range": [ 76, 80 ] }, { "text": "their", "range": [ 103, 108 ] }, { "text": "their", "range": [ 137, 142 ] } ], "source_mask_link": "selected_person" }, { "id": "E2", "label": "shaving cream", "spans": [ { "text": "white shaving cream", "range": [ 41, 60 ] }, { "text": "the cream", "range": [ 88, 97 ] } ], "source_mask_link": null }, { "id": "E3", "label": "face", "spans": [ { "text": "their face", "range": [ 64, 74 ] } ], "source_mask_link": null }, { "id": "E4", "label": "cheeks", "spans": [ { "text": "their cheeks", "range": [ 103, 115 ] } ], "source_mask_link": null }, { "id": "E5", "label": "chin", "spans": [ { "text": "chin", "range": [ 117, 121 ] } ], "source_mask_link": null }, { "id": "E6", "label": "neck", "spans": [ { "text": "neck", "range": [ 127, 131 ] } ], "source_mask_link": null }, { "id": "E7", "label": "hands", "spans": [ { "text": "their hands", "range": [ 137, 148 ] } ], "source_mask_link": null } ], "actions": [ { "id": "A1", "mentions": [ { "text": "reappears", "ranges": [ [ 11, 20 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "type": "simple", "subject_refs": [ "E1" ], "object_refs": [], "confidence": "medium", "temporal_status": "sparse_localized", "intervals": [ { "start_s": 1.0, "end_s": 3.0, "source_start_s": 27.07, "source_end_s": 29.07 } ], "evidence": "At 1.0 seconds only a small portion of the masked person remains at the right edge; by 2.0 seconds the person is fully back in view.", "mask_relation": "subject matches the supplied selected-person mask", "source_subject_refs": "E1" }, { "id": "A2", "mentions": [ { "text": "apply", "ranges": [ [ 35, 40 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } }, { "text": "spread", "ranges": [ [ 81, 87 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "type": "simple", "subject_refs": [ "E1" ], "object_refs": [ "E2" ], "confidence": "high", "temporal_status": "sparse_localized", "intervals": [ { "start_s": 4.0, "end_s": 5.0, "source_start_s": 30.07, "source_end_s": 31.07 }, { "start_s": 6.0, "end_s": 8.5, "source_start_s": 32.07, "source_end_s": 34.57 } ], "evidence": "At 4.0 seconds the masked person rubs the lower cheek and jaw. At 6.0, 7.0, and 8.0 seconds the person visibly spreads white cream across the cheek, chin, and neck with a hand; the action remains active in the final sampled frame.", "mask_relation": "subject matches the supplied selected-person mask", "source_subject_refs": "E1" } ], "notes": [ "The supplied masklet consistently identifies the foreground person described by the caption.", "The shaving cream and named body parts have no separate supplied masks, so their source-mask links are null.", "The inset person visible in the RGB frames is not the selected masked person and is not separately mentioned by the caption.", "\"begins\" is excluded as an aspectual verb; \"apply\" is retained as the embedded physical action. \"apply\" and \"spread\" are grouped as two mentions of the same cream-application event." ] }, "model_output_raw": { "caption": "The person reappears and begins to apply white shaving cream to their face. They spread the cream over their cheeks, chin, and neck with their hands.", "entity_groups": [ { "id": "E1", "entity": "person", "spans": [ { "text": "The person", "range": [ 0, 10 ] }, { "text": "their", "range": [ 64, 69 ] }, { "text": "They", "range": [ 76, 80 ] }, { "text": "their", "range": [ 103, 108 ] }, { "text": "their", "range": [ 137, 142 ] } ], "source_mask": "selected_person_masklet" }, { "id": "E2", "entity": "shaving cream", "spans": [ { "text": "white shaving cream", "range": [ 41, 60 ] }, { "text": "the cream", "range": [ 88, 97 ] } ], "source_mask": null }, { "id": "E3", "entity": "face", "spans": [ { "text": "their face", "range": [ 64, 74 ] } ], "source_mask": null }, { "id": "E4", "entity": "cheeks", "spans": [ { "text": "their cheeks", "range": [ 103, 115 ] } ], "source_mask": null }, { "id": "E5", "entity": "chin", "spans": [ { "text": "chin", "range": [ 117, 121 ] } ], "source_mask": null }, { "id": "E6", "entity": "neck", "spans": [ { "text": "neck", "range": [ 127, 131 ] } ], "source_mask": null }, { "id": "E7", "entity": "hands", "spans": [ { "text": "their hands", "range": [ 137, 148 ] } ], "source_mask": null } ], "actions": [ { "id": "A1", "mentions": [ { "text": "reappears", "ranges": [ [ 11, 20 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "type": "simple", "subject_entity_group": "E1", "object_entity_groups": [], "confidence": "medium", "mask_relation": "subject matches the supplied selected-person mask", "temporal_status": "sparse_localized", "intervals": [ { "start_s": 1.0, "end_s": 3.0, "source_start_s": 27.07, "source_end_s": 29.07 } ], "evidence": "At 1.0 seconds only a small portion of the masked person remains at the right edge; by 2.0 seconds the person is fully back in view." }, { "id": "A2", "mentions": [ { "text": "apply", "ranges": [ [ 35, 40 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } }, { "text": "spread", "ranges": [ [ 81, 87 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "type": "simple", "subject_entity_group": "E1", "object_entity_groups": [ "E2" ], "target_entity_groups": [ "E3", "E4", "E5", "E6" ], "instrument_entity_groups": [ "E7" ], "confidence": "high", "mask_relation": "subject matches the supplied selected-person mask", "temporal_status": "sparse_localized", "intervals": [ { "start_s": 4.0, "end_s": 5.0, "source_start_s": 30.07, "source_end_s": 31.07 }, { "start_s": 6.0, "end_s": 8.5, "source_start_s": 32.07, "source_end_s": 34.57 } ], "evidence": "At 4.0 seconds the masked person rubs the lower cheek and jaw. At 6.0, 7.0, and 8.0 seconds the person visibly spreads white cream across the cheek, chin, and neck with a hand; the action remains active in the final sampled frame." } ], "notes": [ "The supplied masklet consistently identifies the foreground person described by the caption.", "The shaving cream and named body parts have no separate supplied masks, so their source-mask links are null.", "The inset person visible in the RGB frames is not the selected masked person and is not separately mentioned by the caption.", "\"begins\" is excluded as an aspectual verb; \"apply\" is retained as the embedded physical action. \"apply\" and \"spread\" are grouped as two mentions of the same cream-application event." ] }, "normalization_repairs": 0 }, { "source_video_id": "v_-YMpwZkNc2A", "caption_count_category": 12, "clip": { "caption_index": 4, "caption": "The person starts shaving with the razor, beginning with their right cheek. They move on to their neck, shaving in downward strokes. They then shave their left cheek.", "source_interval_sec": [ 34.57, 52.57 ], "loom_clip_id": "v_-YMpwZkNc2A_1036_1576", "clip_frame_interval_4fps": [ 1036, 1576 ], "sample_count_1fps": 18, "sampled_frames": [ { "frame": "00000", "timestamp_sec": 34.57, "offset_sec": 0, "rgb_path": "assets/cat_12/v_-YMpwZkNc2A_1036_1576/0000s_00000_rgb.jpg", "mask_path": "assets/cat_12/v_-YMpwZkNc2A_1036_1576/0000s_00000_mask.png", "mask_status": "available" }, { "frame": "00028", "timestamp_sec": 35.57, "offset_sec": 1, "rgb_path": "assets/cat_12/v_-YMpwZkNc2A_1036_1576/0001s_00028_rgb.jpg", "mask_path": "assets/cat_12/v_-YMpwZkNc2A_1036_1576/0001s_00028_mask.png", "mask_status": "available" }, { "frame": "00060", "timestamp_sec": 36.57, "offset_sec": 2, "rgb_path": "assets/cat_12/v_-YMpwZkNc2A_1036_1576/0002s_00060_rgb.jpg", "mask_path": "assets/cat_12/v_-YMpwZkNc2A_1036_1576/0002s_00060_mask.png", "mask_status": "available" }, { "frame": "00088", "timestamp_sec": 37.57, "offset_sec": 3, "rgb_path": "assets/cat_12/v_-YMpwZkNc2A_1036_1576/0003s_00088_rgb.jpg", "mask_path": "assets/cat_12/v_-YMpwZkNc2A_1036_1576/0003s_00088_mask.png", "mask_status": "available" }, { "frame": "00120", "timestamp_sec": 38.57, "offset_sec": 4, "rgb_path": "assets/cat_12/v_-YMpwZkNc2A_1036_1576/0004s_00120_rgb.jpg", "mask_path": "assets/cat_12/v_-YMpwZkNc2A_1036_1576/0004s_00120_mask.png", "mask_status": "available" }, { "frame": "00148", "timestamp_sec": 39.57, "offset_sec": 5, "rgb_path": "assets/cat_12/v_-YMpwZkNc2A_1036_1576/0005s_00148_rgb.jpg", "mask_path": "assets/cat_12/v_-YMpwZkNc2A_1036_1576/0005s_00148_mask.png", "mask_status": "available" }, { "frame": "00180", "timestamp_sec": 40.57, "offset_sec": 6, "rgb_path": "assets/cat_12/v_-YMpwZkNc2A_1036_1576/0006s_00180_rgb.jpg", "mask_path": "assets/cat_12/v_-YMpwZkNc2A_1036_1576/0006s_00180_mask.png", "mask_status": "available" }, { "frame": "00208", "timestamp_sec": 41.57, "offset_sec": 7, "rgb_path": "assets/cat_12/v_-YMpwZkNc2A_1036_1576/0007s_00208_rgb.jpg", "mask_path": "assets/cat_12/v_-YMpwZkNc2A_1036_1576/0007s_00208_mask.png", "mask_status": "available" }, { "frame": "00240", "timestamp_sec": 42.57, "offset_sec": 8, "rgb_path": "assets/cat_12/v_-YMpwZkNc2A_1036_1576/0008s_00240_rgb.jpg", "mask_path": "assets/cat_12/v_-YMpwZkNc2A_1036_1576/0008s_00240_mask.png", "mask_status": "available" }, { "frame": "00268", "timestamp_sec": 43.57, "offset_sec": 9, "rgb_path": "assets/cat_12/v_-YMpwZkNc2A_1036_1576/0009s_00268_rgb.jpg", "mask_path": "assets/cat_12/v_-YMpwZkNc2A_1036_1576/0009s_00268_mask.png", "mask_status": "available" }, { "frame": "00300", "timestamp_sec": 44.57, "offset_sec": 10, "rgb_path": "assets/cat_12/v_-YMpwZkNc2A_1036_1576/0010s_00300_rgb.jpg", "mask_path": "assets/cat_12/v_-YMpwZkNc2A_1036_1576/0010s_00300_mask.png", "mask_status": "available" }, { "frame": "00328", "timestamp_sec": 45.57, "offset_sec": 11, "rgb_path": "assets/cat_12/v_-YMpwZkNc2A_1036_1576/0011s_00328_rgb.jpg", "mask_path": "assets/cat_12/v_-YMpwZkNc2A_1036_1576/0011s_00328_mask.png", "mask_status": "available" }, { "frame": "00360", "timestamp_sec": 46.57, "offset_sec": 12, "rgb_path": "assets/cat_12/v_-YMpwZkNc2A_1036_1576/0012s_00360_rgb.jpg", "mask_path": "assets/cat_12/v_-YMpwZkNc2A_1036_1576/0012s_00360_mask.png", "mask_status": "available" }, { "frame": "00388", "timestamp_sec": 47.57, "offset_sec": 13, "rgb_path": "assets/cat_12/v_-YMpwZkNc2A_1036_1576/0013s_00388_rgb.jpg", "mask_path": "assets/cat_12/v_-YMpwZkNc2A_1036_1576/0013s_00388_mask.png", "mask_status": "available" }, { "frame": "00420", "timestamp_sec": 48.57, "offset_sec": 14, "rgb_path": "assets/cat_12/v_-YMpwZkNc2A_1036_1576/0014s_00420_rgb.jpg", "mask_path": "assets/cat_12/v_-YMpwZkNc2A_1036_1576/0014s_00420_mask.png", "mask_status": "available" }, { "frame": "00448", "timestamp_sec": 49.57, "offset_sec": 15, "rgb_path": "assets/cat_12/v_-YMpwZkNc2A_1036_1576/0015s_00448_rgb.jpg", "mask_path": "assets/cat_12/v_-YMpwZkNc2A_1036_1576/0015s_00448_mask.png", "mask_status": "available" }, { "frame": "00480", "timestamp_sec": 50.57, "offset_sec": 16, "rgb_path": "assets/cat_12/v_-YMpwZkNc2A_1036_1576/0016s_00480_rgb.jpg", "mask_path": "assets/cat_12/v_-YMpwZkNc2A_1036_1576/0016s_00480_mask.png", "mask_status": "available" }, { "frame": "00508", "timestamp_sec": 51.57, "offset_sec": 17, "rgb_path": "assets/cat_12/v_-YMpwZkNc2A_1036_1576/0017s_00508_rgb.jpg", "mask_path": "assets/cat_12/v_-YMpwZkNc2A_1036_1576/0017s_00508_mask.png", "mask_status": "available" } ], "obj_id": [ 0 ], "anno_id": [ 81 ], "fps_inferred_from_clip_interval": 30.0 }, "usage": { "input_tokens": 56057, "input_tokens_details": { "cache_write_tokens": 56054, "cached_tokens": 0 }, "output_tokens": 4214, "output_tokens_details": { "reasoning_tokens": 2805 }, "total_tokens": 60271 }, "model": "gpt-5.6-sol", "effort": "medium", "custom_id": "cat12_cap04_v_-YMpwZkNc2A_1036_1576", "annotation": { "caption": "The person starts shaving with the razor, beginning with their right cheek. They move on to their neck, shaving in downward strokes. They then shave their left cheek.", "entity_groups": [ { "id": "entity_person_1", "label": "person", "spans": [ { "text": "The person", "range": [ 0, 10 ] }, { "text": "their", "range": [ 57, 62 ] }, { "text": "They", "range": [ 76, 80 ] }, { "text": "their", "range": [ 92, 97 ] }, { "text": "They", "range": [ 133, 137 ] }, { "text": "their", "range": [ 149, 154 ] } ], "source_mask_link": "selected_person" }, { "id": "entity_razor_1", "label": "razor", "spans": [ { "text": "the razor", "range": [ 31, 40 ] } ], "source_mask_link": null }, { "id": "entity_right_cheek_1", "label": "right cheek", "spans": [ { "text": "their right cheek", "range": [ 57, 74 ] } ], "source_mask_link": null }, { "id": "entity_neck_1", "label": "neck", "spans": [ { "text": "their neck", "range": [ 92, 102 ] } ], "source_mask_link": null }, { "id": "entity_left_cheek_1", "label": "left cheek", "spans": [ { "text": "their left cheek", "range": [ 149, 165 ] } ], "source_mask_link": null } ], "actions": [ { "id": "action_shave_right_cheek", "mentions": [ { "text": "shaving", "ranges": [ [ 18, 25 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "type": "simple", "subject_refs": [ "entity_person_1" ], "object_refs": [ "entity_right_cheek_1" ], "confidence": "medium", "temporal_status": "sparse_supported", "intervals": [ { "start_s": 0.0, "end_s": 1.0, "source_start_s": 34.57, "source_end_s": 35.57 }, { "start_s": 3.0, "end_s": 10.0, "source_start_s": 37.57, "source_end_s": 44.57 } ], "evidence": "The mask-linked foreground person visibly holds a razor against a lathered cheek or jaw area in the supported samples; exact anatomical side is somewhat uncertain from the mirrored viewpoint.", "mask_relation": "subject_matches_selected_person_mask", "source_subject_refs": "entity_person_1" }, { "id": "action_shave_neck", "mentions": [ { "text": "shaving", "ranges": [ [ 104, 111 ] ], "occurrence": 2, "particle": { "text": null, "occurrence": 1 } } ], "type": "simple", "subject_refs": [ "entity_person_1" ], "object_refs": [ "entity_neck_1" ], "confidence": "medium", "temporal_status": "sparse_supported", "intervals": [ { "start_s": 11.0, "end_s": 12.0, "source_start_s": 45.57, "source_end_s": 46.57 } ], "evidence": "At 11.0 seconds the mask-linked person visibly applies the razor along the lathered side of the neck.", "mask_relation": "subject_matches_selected_person_mask", "source_subject_refs": "entity_person_1" }, { "id": "action_shave_left_cheek", "mentions": [ { "text": "shave", "ranges": [ [ 143, 148 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "type": "simple", "subject_refs": [ "entity_person_1" ], "object_refs": [ "entity_left_cheek_1" ], "confidence": "medium", "temporal_status": "sparse_supported", "intervals": [ { "start_s": 12.0, "end_s": 15.0, "source_start_s": 46.57, "source_end_s": 49.57 }, { "start_s": 16.0, "end_s": 17.0, "source_start_s": 50.57, "source_end_s": 51.57 } ], "evidence": "Later samples show the mask-linked person making additional razor passes over the opposite cheek and adjacent jaw area; side identification follows the caption's stated sequence.", "mask_relation": "subject_matches_selected_person_mask", "source_subject_refs": "entity_person_1" } ], "notes": [ "The supplied masklet consistently identifies the foreground person who is shaving, not the people visible in the inset screen.", "No separate masks were supplied for the razor, right cheek, neck, or left cheek, so those entity groups have null source-mask links.", "The aspectual predicates \"starts\" and \"beginning\" were excluded; \"move on\" was treated as a sequencing expression rather than a separate physical action.", "Temporal intervals are sparse sampled-frame support rather than exact continuous-video boundaries. Anatomical left/right assignment has medium confidence because of head rotation and a potentially mirrored view." ] }, "model_output_raw": { "caption": "The person starts shaving with the razor, beginning with their right cheek. They move on to their neck, shaving in downward strokes. They then shave their left cheek.", "entity_groups": [ { "id": "entity_person_1", "label": "person", "spans": [ { "text": "The person", "range": [ 0, 10 ] }, { "text": "their", "range": [ 57, 62 ] }, { "text": "They", "range": [ 76, 80 ] }, { "text": "their", "range": [ 92, 97 ] }, { "text": "They", "range": [ 133, 137 ] }, { "text": "their", "range": [ 149, 154 ] } ], "source_mask": "selected_person_masklet" }, { "id": "entity_razor_1", "label": "razor", "spans": [ { "text": "the razor", "range": [ 31, 40 ] } ], "source_mask": null }, { "id": "entity_right_cheek_1", "label": "right cheek", "spans": [ { "text": "their right cheek", "range": [ 57, 74 ] } ], "source_mask": null }, { "id": "entity_neck_1", "label": "neck", "spans": [ { "text": "their neck", "range": [ 92, 102 ] } ], "source_mask": null }, { "id": "entity_left_cheek_1", "label": "left cheek", "spans": [ { "text": "their left cheek", "range": [ 149, 165 ] } ], "source_mask": null } ], "actions": [ { "id": "action_shave_right_cheek", "mentions": [ { "text": "shaving", "ranges": [ [ 18, 25 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "type": "simple", "subject_entity_group_id": "entity_person_1", "object_entity_group_ids": [ "entity_right_cheek_1" ], "instrument_entity_group_ids": [ "entity_razor_1" ], "confidence": "medium", "mask_relation": "subject_matches_selected_person_mask", "temporal_status": "sparse_supported", "intervals": [ { "clip_relative_s": [ 0.0, 1.0 ], "source_video_s": [ 34.57, 35.57 ] }, { "clip_relative_s": [ 3.0, 10.0 ], "source_video_s": [ 37.57, 44.57 ] } ], "evidence": "The mask-linked foreground person visibly holds a razor against a lathered cheek or jaw area in the supported samples; exact anatomical side is somewhat uncertain from the mirrored viewpoint." }, { "id": "action_shave_neck", "mentions": [ { "text": "shaving", "ranges": [ [ 104, 111 ] ], "occurrence": 2, "particle": { "text": null, "occurrence": 1 } } ], "type": "simple", "subject_entity_group_id": "entity_person_1", "object_entity_group_ids": [ "entity_neck_1" ], "instrument_entity_group_ids": [ "entity_razor_1" ], "confidence": "medium", "mask_relation": "subject_matches_selected_person_mask", "temporal_status": "sparse_supported", "intervals": [ { "clip_relative_s": [ 11.0, 12.0 ], "source_video_s": [ 45.57, 46.57 ] } ], "evidence": "At 11.0 seconds the mask-linked person visibly applies the razor along the lathered side of the neck." }, { "id": "action_shave_left_cheek", "mentions": [ { "text": "shave", "ranges": [ [ 143, 148 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "type": "simple", "subject_entity_group_id": "entity_person_1", "object_entity_group_ids": [ "entity_left_cheek_1" ], "instrument_entity_group_ids": [ "entity_razor_1" ], "confidence": "medium", "mask_relation": "subject_matches_selected_person_mask", "temporal_status": "sparse_supported", "intervals": [ { "clip_relative_s": [ 12.0, 15.0 ], "source_video_s": [ 46.57, 49.57 ] }, { "clip_relative_s": [ 16.0, 17.0 ], "source_video_s": [ 50.57, 51.57 ] } ], "evidence": "Later samples show the mask-linked person making additional razor passes over the opposite cheek and adjacent jaw area; side identification follows the caption's stated sequence." } ], "notes": [ "The supplied masklet consistently identifies the foreground person who is shaving, not the people visible in the inset screen.", "No separate masks were supplied for the razor, right cheek, neck, or left cheek, so those entity groups have null source-mask links.", "The aspectual predicates \"starts\" and \"beginning\" were excluded; \"move on\" was treated as a sequencing expression rather than a separate physical action.", "Temporal intervals are sparse sampled-frame support rather than exact continuous-video boundaries. Anatomical left/right assignment has medium confidence because of head rotation and a potentially mirrored view." ] }, "normalization_repairs": 0 }, { "source_video_id": "v_-YMpwZkNc2A", "caption_count_category": 12, "clip": { "caption_index": 5, "caption": "The person continues to shave the shaving cream from their neck and chin area. They tilt their head back to shave under their chin, frequently bending down to rinse the razor.", "source_interval_sec": [ 52.57, 80.07 ], "loom_clip_id": "v_-YMpwZkNc2A_1576_2400", "clip_frame_interval_4fps": [ 1576, 2400 ], "sample_count_1fps": 28, "sampled_frames": [ { "frame": "00000", "timestamp_sec": 52.57, "offset_sec": 0, "rgb_path": "assets/cat_12/v_-YMpwZkNc2A_1576_2400/0000s_00000_rgb.jpg", "mask_path": "assets/cat_12/v_-YMpwZkNc2A_1576_2400/0000s_00000_mask.png", "mask_status": "available" }, { "frame": "00028", "timestamp_sec": 53.57, "offset_sec": 1, "rgb_path": "assets/cat_12/v_-YMpwZkNc2A_1576_2400/0001s_00028_rgb.jpg", "mask_path": "assets/cat_12/v_-YMpwZkNc2A_1576_2400/0001s_00028_mask.png", "mask_status": "available" }, { "frame": "00060", "timestamp_sec": 54.57, "offset_sec": 2, "rgb_path": "assets/cat_12/v_-YMpwZkNc2A_1576_2400/0002s_00060_rgb.jpg", "mask_path": "assets/cat_12/v_-YMpwZkNc2A_1576_2400/0002s_00060_mask.png", "mask_status": "available" }, { "frame": "00088", "timestamp_sec": 55.57, "offset_sec": 3, "rgb_path": "assets/cat_12/v_-YMpwZkNc2A_1576_2400/0003s_00088_rgb.jpg", "mask_path": "assets/cat_12/v_-YMpwZkNc2A_1576_2400/0003s_00088_mask.png", "mask_status": "available" }, { "frame": "00120", "timestamp_sec": 56.57, "offset_sec": 4, "rgb_path": "assets/cat_12/v_-YMpwZkNc2A_1576_2400/0004s_00120_rgb.jpg", "mask_path": "assets/cat_12/v_-YMpwZkNc2A_1576_2400/0004s_00120_mask.png", "mask_status": "available" }, { "frame": "00148", "timestamp_sec": 57.57, "offset_sec": 5, "rgb_path": "assets/cat_12/v_-YMpwZkNc2A_1576_2400/0005s_00148_rgb.jpg", "mask_path": "assets/cat_12/v_-YMpwZkNc2A_1576_2400/0005s_00148_mask.png", "mask_status": "available" }, { "frame": "00180", "timestamp_sec": 58.57, "offset_sec": 6, "rgb_path": "assets/cat_12/v_-YMpwZkNc2A_1576_2400/0006s_00180_rgb.jpg", "mask_path": "assets/cat_12/v_-YMpwZkNc2A_1576_2400/0006s_00180_mask.png", "mask_status": "available" }, { "frame": "00208", "timestamp_sec": 59.57, "offset_sec": 7, "rgb_path": "assets/cat_12/v_-YMpwZkNc2A_1576_2400/0007s_00208_rgb.jpg", "mask_path": "assets/cat_12/v_-YMpwZkNc2A_1576_2400/0007s_00208_mask.png", "mask_status": "available" }, { "frame": "00240", "timestamp_sec": 60.57, "offset_sec": 8, "rgb_path": "assets/cat_12/v_-YMpwZkNc2A_1576_2400/0008s_00240_rgb.jpg", "mask_path": "assets/cat_12/v_-YMpwZkNc2A_1576_2400/0008s_00240_mask.png", "mask_status": "available" }, { "frame": "00268", "timestamp_sec": 61.57, "offset_sec": 9, "rgb_path": "assets/cat_12/v_-YMpwZkNc2A_1576_2400/0009s_00268_rgb.jpg", "mask_path": "assets/cat_12/v_-YMpwZkNc2A_1576_2400/0009s_00268_mask.png", "mask_status": "available" }, { "frame": "00300", "timestamp_sec": 62.57, "offset_sec": 10, "rgb_path": "assets/cat_12/v_-YMpwZkNc2A_1576_2400/0010s_00300_rgb.jpg", "mask_path": "assets/cat_12/v_-YMpwZkNc2A_1576_2400/0010s_00300_mask.png", "mask_status": "available" }, { "frame": "00328", "timestamp_sec": 63.57, "offset_sec": 11, "rgb_path": "assets/cat_12/v_-YMpwZkNc2A_1576_2400/0011s_00328_rgb.jpg", "mask_path": "assets/cat_12/v_-YMpwZkNc2A_1576_2400/0011s_00328_mask.png", "mask_status": "available" }, { "frame": "00360", "timestamp_sec": 64.57, "offset_sec": 12, "rgb_path": "assets/cat_12/v_-YMpwZkNc2A_1576_2400/0012s_00360_rgb.jpg", "mask_path": "assets/cat_12/v_-YMpwZkNc2A_1576_2400/0012s_00360_mask.png", "mask_status": "available" }, { "frame": "00388", "timestamp_sec": 65.57, "offset_sec": 13, "rgb_path": "assets/cat_12/v_-YMpwZkNc2A_1576_2400/0013s_00388_rgb.jpg", "mask_path": "assets/cat_12/v_-YMpwZkNc2A_1576_2400/0013s_00388_mask.png", "mask_status": "available" }, { "frame": "00420", "timestamp_sec": 66.57, "offset_sec": 14, "rgb_path": "assets/cat_12/v_-YMpwZkNc2A_1576_2400/0014s_00420_rgb.jpg", "mask_path": "assets/cat_12/v_-YMpwZkNc2A_1576_2400/0014s_00420_mask.png", "mask_status": "available" }, { "frame": "00448", "timestamp_sec": 67.57, "offset_sec": 15, "rgb_path": "assets/cat_12/v_-YMpwZkNc2A_1576_2400/0015s_00448_rgb.jpg", "mask_path": "assets/cat_12/v_-YMpwZkNc2A_1576_2400/0015s_00448_mask.png", "mask_status": "available" }, { "frame": "00480", "timestamp_sec": 68.57, "offset_sec": 16, "rgb_path": "assets/cat_12/v_-YMpwZkNc2A_1576_2400/0016s_00480_rgb.jpg", "mask_path": "assets/cat_12/v_-YMpwZkNc2A_1576_2400/0016s_00480_mask.png", "mask_status": "available" }, { "frame": "00508", "timestamp_sec": 69.57, "offset_sec": 17, "rgb_path": "assets/cat_12/v_-YMpwZkNc2A_1576_2400/0017s_00508_rgb.jpg", "mask_path": "assets/cat_12/v_-YMpwZkNc2A_1576_2400/0017s_00508_mask.png", "mask_status": "available" }, { "frame": "00540", "timestamp_sec": 70.57, "offset_sec": 18, "rgb_path": "assets/cat_12/v_-YMpwZkNc2A_1576_2400/0018s_00540_rgb.jpg", "mask_path": "assets/cat_12/v_-YMpwZkNc2A_1576_2400/0018s_00540_mask.png", "mask_status": "available" }, { "frame": "00568", "timestamp_sec": 71.57, "offset_sec": 19, "rgb_path": "assets/cat_12/v_-YMpwZkNc2A_1576_2400/0019s_00568_rgb.jpg", "mask_path": "assets/cat_12/v_-YMpwZkNc2A_1576_2400/0019s_00568_mask.png", "mask_status": "available" }, { "frame": "00600", "timestamp_sec": 72.57, "offset_sec": 20, "rgb_path": "assets/cat_12/v_-YMpwZkNc2A_1576_2400/0020s_00600_rgb.jpg", "mask_path": "assets/cat_12/v_-YMpwZkNc2A_1576_2400/0020s_00600_mask.png", "mask_status": "available" }, { "frame": "00628", "timestamp_sec": 73.57, "offset_sec": 21, "rgb_path": "assets/cat_12/v_-YMpwZkNc2A_1576_2400/0021s_00628_rgb.jpg", "mask_path": "assets/cat_12/v_-YMpwZkNc2A_1576_2400/0021s_00628_mask.png", "mask_status": "available" }, { "frame": "00660", "timestamp_sec": 74.57, "offset_sec": 22, "rgb_path": "assets/cat_12/v_-YMpwZkNc2A_1576_2400/0022s_00660_rgb.jpg", "mask_path": "assets/cat_12/v_-YMpwZkNc2A_1576_2400/0022s_00660_mask.png", "mask_status": "available" }, { "frame": "00688", "timestamp_sec": 75.57, "offset_sec": 23, "rgb_path": "assets/cat_12/v_-YMpwZkNc2A_1576_2400/0023s_00688_rgb.jpg", "mask_path": "assets/cat_12/v_-YMpwZkNc2A_1576_2400/0023s_00688_mask.png", "mask_status": "available" }, { "frame": "00720", "timestamp_sec": 76.57, "offset_sec": 24, "rgb_path": "assets/cat_12/v_-YMpwZkNc2A_1576_2400/0024s_00720_rgb.jpg", "mask_path": "assets/cat_12/v_-YMpwZkNc2A_1576_2400/0024s_00720_mask.png", "mask_status": "available" }, { "frame": "00748", "timestamp_sec": 77.57, "offset_sec": 25, "rgb_path": "assets/cat_12/v_-YMpwZkNc2A_1576_2400/0025s_00748_rgb.jpg", "mask_path": "assets/cat_12/v_-YMpwZkNc2A_1576_2400/0025s_00748_mask.png", "mask_status": "available" }, { "frame": "00780", "timestamp_sec": 78.57, "offset_sec": 26, "rgb_path": "assets/cat_12/v_-YMpwZkNc2A_1576_2400/0026s_00780_rgb.jpg", "mask_path": "assets/cat_12/v_-YMpwZkNc2A_1576_2400/0026s_00780_mask.png", "mask_status": "available" }, { "frame": "00808", "timestamp_sec": 79.57, "offset_sec": 27, "rgb_path": "assets/cat_12/v_-YMpwZkNc2A_1576_2400/0027s_00808_rgb.jpg", "mask_path": "assets/cat_12/v_-YMpwZkNc2A_1576_2400/0027s_00808_mask.png", "mask_status": "available" } ], "obj_id": [ 0 ], "anno_id": [ 82 ], "fps_inferred_from_clip_interval": 29.9636 }, "usage": { "input_tokens": 78688, "input_tokens_details": { "cache_write_tokens": 78685, "cached_tokens": 0 }, "output_tokens": 5172, "output_tokens_details": { "reasoning_tokens": 2852 }, "total_tokens": 83860 }, "model": "gpt-5.6-sol", "effort": "medium", "custom_id": "cat12_cap05_v_-YMpwZkNc2A_1576_2400", "annotation": { "caption": "The person continues to shave the shaving cream from their neck and chin area. They tilt their head back to shave under their chin, frequently bending down to rinse the razor.", "entity_groups": [ { "id": "person_1", "label": "person", "spans": [ { "text": "The person", "range": [ 0, 10 ] }, { "text": "their", "range": [ 53, 58 ] }, { "text": "They", "range": [ 79, 83 ] }, { "text": "their", "range": [ 89, 94 ] }, { "text": "their", "range": [ 120, 125 ] } ], "source_mask_link": "selected_person" }, { "id": "shaving_cream_1", "label": "shaving cream", "spans": [ { "text": "the shaving cream", "range": [ 30, 47 ] } ], "source_mask_link": null }, { "id": "neck_1", "label": "neck", "spans": [ { "text": "their neck", "range": [ 53, 63 ] } ], "source_mask_link": null }, { "id": "chin_1", "label": "chin", "spans": [ { "text": "chin area", "range": [ 68, 77 ] }, { "text": "their chin", "range": [ 120, 130 ] } ], "source_mask_link": null }, { "id": "head_1", "label": "head", "spans": [ { "text": "their head", "range": [ 89, 99 ] } ], "source_mask_link": null }, { "id": "razor_1", "label": "razor", "spans": [ { "text": "the razor", "range": [ 165, 174 ] } ], "source_mask_link": null } ], "actions": [ { "id": "action_shave_1", "mentions": [ { "text": "shave", "ranges": [ [ 24, 29 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } }, { "text": "shave", "ranges": [ [ 108, 113 ] ], "occurrence": 2, "particle": { "text": null, "occurrence": 1 } } ], "type": "simple", "subject_refs": [ "person_1" ], "object_refs": [ "shaving_cream_1" ], "confidence": "high", "temporal_status": "supported", "intervals": [ { "start_s": 0.0, "end_s": 2.0, "source_start_s": 52.57, "source_end_s": 54.57 }, { "start_s": 3.0, "end_s": 5.0, "source_start_s": 55.57, "source_end_s": 57.57 }, { "start_s": 6.0, "end_s": 8.0, "source_start_s": 58.57, "source_end_s": 60.57 }, { "start_s": 9.0, "end_s": 10.0, "source_start_s": 61.57, "source_end_s": 62.57 }, { "start_s": 12.0, "end_s": 14.0, "source_start_s": 64.57, "source_end_s": 66.57 }, { "start_s": 16.0, "end_s": 17.0, "source_start_s": 68.57, "source_end_s": 69.57 }, { "start_s": 19.0, "end_s": 21.0, "source_start_s": 71.57, "source_end_s": 73.57 }, { "start_s": 22.0, "end_s": 25.0, "source_start_s": 74.57, "source_end_s": 77.57 }, { "start_s": 26.0, "end_s": 27.499999999999993, "source_start_s": 78.57, "source_end_s": 80.07 } ], "evidence": "Across the supported frames, the mask-linked foreground person repeatedly draws a razor over shaving-cream-covered portions of his neck and chin.", "mask_relation": "subject matches the supplied selected-person mask", "source_subject_refs": "person_1" }, { "id": "action_tilt_1", "mentions": [ { "text": "tilt", "ranges": [ [ 84, 88 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "type": "simple", "subject_refs": [ "person_1" ], "object_refs": [ "head_1" ], "confidence": "high", "temporal_status": "supported", "intervals": [ { "start_s": 1.0, "end_s": 2.0, "source_start_s": 53.57, "source_end_s": 54.57 }, { "start_s": 3.0, "end_s": 5.0, "source_start_s": 55.57, "source_end_s": 57.57 }, { "start_s": 6.0, "end_s": 10.0, "source_start_s": 58.57, "source_end_s": 62.57 }, { "start_s": 12.0, "end_s": 14.0, "source_start_s": 64.57, "source_end_s": 66.57 }, { "start_s": 16.0, "end_s": 18.0, "source_start_s": 68.57, "source_end_s": 70.57 }, { "start_s": 19.0, "end_s": 21.0, "source_start_s": 71.57, "source_end_s": 73.57 }, { "start_s": 22.0, "end_s": 25.0, "source_start_s": 74.57, "source_end_s": 77.57 }, { "start_s": 26.0, "end_s": 27.499999999999993, "source_start_s": 78.57, "source_end_s": 80.07 } ], "evidence": "The selected person repeatedly raises and holds his head back, exposing the underside of his chin and neck.", "mask_relation": "subject matches the supplied selected-person mask", "source_subject_refs": "person_1" }, { "id": "action_bend_1", "mentions": [ { "text": "bending", "ranges": [ [ 143, 150 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "type": "simple", "subject_refs": [ "person_1" ], "object_refs": [], "confidence": "high", "temporal_status": "supported", "intervals": [ { "start_s": 2.0, "end_s": 3.0, "source_start_s": 54.57, "source_end_s": 55.57 }, { "start_s": 5.0, "end_s": 6.0, "source_start_s": 57.57, "source_end_s": 58.57 }, { "start_s": 10.0, "end_s": 12.0, "source_start_s": 62.57, "source_end_s": 64.57 }, { "start_s": 14.0, "end_s": 15.0, "source_start_s": 66.57, "source_end_s": 67.57 }, { "start_s": 18.0, "end_s": 19.0, "source_start_s": 70.57, "source_end_s": 71.57 }, { "start_s": 21.0, "end_s": 22.0, "source_start_s": 73.57, "source_end_s": 74.57 }, { "start_s": 25.0, "end_s": 26.0, "source_start_s": 77.57, "source_end_s": 78.57 } ], "evidence": "The mask-linked person repeatedly lowers his head and upper torso out of the central shaving position.", "mask_relation": "subject matches the supplied selected-person mask", "source_subject_refs": "person_1" }, { "id": "action_rinse_1", "mentions": [ { "text": "rinse", "ranges": [ [ 159, 164 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "type": "simple", "subject_refs": [ "person_1" ], "object_refs": [ "razor_1" ], "confidence": "medium", "temporal_status": "not_directly_visible", "intervals": [], "evidence": "The person bends below the visible shaving area several times, but the razor, water, and rinsing contact are occluded or outside the sampled frame.", "mask_relation": "subject matches the supplied selected-person mask", "source_subject_refs": "person_1" } ], "notes": [ "The supplied masklet consistently identifies the foreground person who is shaving; the unrelated people visible in the inset video are not mask-linked caption subjects.", "No separate source masks were supplied for the shaving cream, neck, chin, head, or razor.", "The aspectual verb \"continues\" and the directional word \"down\" are excluded from action spans under RDCap V5.", "The two occurrences of \"shave\" describe the same continuing shaving event and are grouped as two mentions.", "The final shaving and head-tilt support reaches the clip boundary, so those intervals end at 27.5 clip-relative seconds, corresponding to source time 80.07 seconds." ] }, "model_output_raw": { "caption": "The person continues to shave the shaving cream from their neck and chin area. They tilt their head back to shave under their chin, frequently bending down to rinse the razor.", "entity_groups": [ { "id": "person_1", "label": "person", "spans": [ { "text": "The person", "range": [ 0, 10 ] }, { "text": "their", "range": [ 53, 58 ] }, { "text": "They", "range": [ 79, 83 ] }, { "text": "their", "range": [ 89, 94 ] }, { "text": "their", "range": [ 120, 125 ] } ], "source_mask": "selected_person_masklet" }, { "id": "shaving_cream_1", "label": "shaving cream", "spans": [ { "text": "the shaving cream", "range": [ 30, 47 ] } ], "source_mask": null }, { "id": "neck_1", "label": "neck", "spans": [ { "text": "their neck", "range": [ 53, 63 ] } ], "source_mask": null }, { "id": "chin_1", "label": "chin", "spans": [ { "text": "chin area", "range": [ 68, 77 ] }, { "text": "their chin", "range": [ 120, 130 ] } ], "source_mask": null }, { "id": "head_1", "label": "head", "spans": [ { "text": "their head", "range": [ 89, 99 ] } ], "source_mask": null }, { "id": "razor_1", "label": "razor", "spans": [ { "text": "the razor", "range": [ 165, 174 ] } ], "source_mask": null } ], "actions": [ { "id": "action_shave_1", "mentions": [ { "text": "shave", "ranges": [ [ 24, 29 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } }, { "text": "shave", "ranges": [ [ 108, 113 ] ], "occurrence": 2, "particle": { "text": null, "occurrence": 1 } } ], "type": "simple", "subject_ref": "person_1", "object_refs": [ "shaving_cream_1" ], "related_entity_refs": [ "neck_1", "chin_1", "razor_1" ], "confidence": "high", "mask_relation": "subject matches the supplied selected-person mask", "temporal_status": "supported", "intervals": [ { "clip_relative_s": [ 0.0, 2.0 ], "source_video_s": [ 52.57, 54.57 ] }, { "clip_relative_s": [ 3.0, 5.0 ], "source_video_s": [ 55.57, 57.57 ] }, { "clip_relative_s": [ 6.0, 8.0 ], "source_video_s": [ 58.57, 60.57 ] }, { "clip_relative_s": [ 9.0, 10.0 ], "source_video_s": [ 61.57, 62.57 ] }, { "clip_relative_s": [ 12.0, 14.0 ], "source_video_s": [ 64.57, 66.57 ] }, { "clip_relative_s": [ 16.0, 17.0 ], "source_video_s": [ 68.57, 69.57 ] }, { "clip_relative_s": [ 19.0, 21.0 ], "source_video_s": [ 71.57, 73.57 ] }, { "clip_relative_s": [ 22.0, 25.0 ], "source_video_s": [ 74.57, 77.57 ] }, { "clip_relative_s": [ 26.0, 27.5 ], "source_video_s": [ 78.57, 80.07 ] } ], "evidence": "Across the supported frames, the mask-linked foreground person repeatedly draws a razor over shaving-cream-covered portions of his neck and chin." }, { "id": "action_tilt_1", "mentions": [ { "text": "tilt", "ranges": [ [ 84, 88 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "type": "simple", "subject_ref": "person_1", "object_refs": [ "head_1" ], "confidence": "high", "mask_relation": "subject matches the supplied selected-person mask", "temporal_status": "supported", "intervals": [ { "clip_relative_s": [ 1.0, 2.0 ], "source_video_s": [ 53.57, 54.57 ] }, { "clip_relative_s": [ 3.0, 5.0 ], "source_video_s": [ 55.57, 57.57 ] }, { "clip_relative_s": [ 6.0, 10.0 ], "source_video_s": [ 58.57, 62.57 ] }, { "clip_relative_s": [ 12.0, 14.0 ], "source_video_s": [ 64.57, 66.57 ] }, { "clip_relative_s": [ 16.0, 18.0 ], "source_video_s": [ 68.57, 70.57 ] }, { "clip_relative_s": [ 19.0, 21.0 ], "source_video_s": [ 71.57, 73.57 ] }, { "clip_relative_s": [ 22.0, 25.0 ], "source_video_s": [ 74.57, 77.57 ] }, { "clip_relative_s": [ 26.0, 27.5 ], "source_video_s": [ 78.57, 80.07 ] } ], "evidence": "The selected person repeatedly raises and holds his head back, exposing the underside of his chin and neck." }, { "id": "action_bend_1", "mentions": [ { "text": "bending", "ranges": [ [ 143, 150 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "type": "simple", "subject_ref": "person_1", "object_refs": [], "confidence": "high", "mask_relation": "subject matches the supplied selected-person mask", "temporal_status": "supported", "intervals": [ { "clip_relative_s": [ 2.0, 3.0 ], "source_video_s": [ 54.57, 55.57 ] }, { "clip_relative_s": [ 5.0, 6.0 ], "source_video_s": [ 57.57, 58.57 ] }, { "clip_relative_s": [ 10.0, 12.0 ], "source_video_s": [ 62.57, 64.57 ] }, { "clip_relative_s": [ 14.0, 15.0 ], "source_video_s": [ 66.57, 67.57 ] }, { "clip_relative_s": [ 18.0, 19.0 ], "source_video_s": [ 70.57, 71.57 ] }, { "clip_relative_s": [ 21.0, 22.0 ], "source_video_s": [ 73.57, 74.57 ] }, { "clip_relative_s": [ 25.0, 26.0 ], "source_video_s": [ 77.57, 78.57 ] } ], "evidence": "The mask-linked person repeatedly lowers his head and upper torso out of the central shaving position." }, { "id": "action_rinse_1", "mentions": [ { "text": "rinse", "ranges": [ [ 159, 164 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "type": "simple", "subject_ref": "person_1", "object_refs": [ "razor_1" ], "confidence": "medium", "mask_relation": "subject matches the supplied selected-person mask", "temporal_status": "not_directly_visible", "intervals": [], "evidence": "The person bends below the visible shaving area several times, but the razor, water, and rinsing contact are occluded or outside the sampled frame." } ], "notes": [ "The supplied masklet consistently identifies the foreground person who is shaving; the unrelated people visible in the inset video are not mask-linked caption subjects.", "No separate source masks were supplied for the shaving cream, neck, chin, head, or razor.", "The aspectual verb \"continues\" and the directional word \"down\" are excluded from action spans under RDCap V5.", "The two occurrences of \"shave\" describe the same continuing shaving event and are grouped as two mentions.", "The final shaving and head-tilt support reaches the clip boundary, so those intervals end at 27.5 clip-relative seconds, corresponding to source time 80.07 seconds." ] }, "normalization_repairs": 0 }, { "source_video_id": "v_-YMpwZkNc2A", "caption_count_category": 12, "clip": { "caption_index": 6, "caption": "The person continues shaving their face, focusing on the remaining patches of shaving cream on their cheeks and neck. After several strokes, they inspect the razor. They then continue to shave their neck.", "source_interval_sec": [ 80.07, 94.57 ], "loom_clip_id": "v_-YMpwZkNc2A_2400_2834", "clip_frame_interval_4fps": [ 2400, 2834 ], "sample_count_1fps": 15, "sampled_frames": [ { "frame": "00000", "timestamp_sec": 80.07, "offset_sec": 0, "rgb_path": "assets/cat_12/v_-YMpwZkNc2A_2400_2834/0000s_00000_rgb.jpg", "mask_path": "assets/cat_12/v_-YMpwZkNc2A_2400_2834/0000s_00000_mask.png", "mask_status": "available" }, { "frame": "00028", "timestamp_sec": 81.07, "offset_sec": 1, "rgb_path": "assets/cat_12/v_-YMpwZkNc2A_2400_2834/0001s_00028_rgb.jpg", "mask_path": "assets/cat_12/v_-YMpwZkNc2A_2400_2834/0001s_00028_mask.png", "mask_status": "available" }, { "frame": "00060", "timestamp_sec": 82.07, "offset_sec": 2, "rgb_path": "assets/cat_12/v_-YMpwZkNc2A_2400_2834/0002s_00060_rgb.jpg", "mask_path": "assets/cat_12/v_-YMpwZkNc2A_2400_2834/0002s_00060_mask.png", "mask_status": "available" }, { "frame": "00088", "timestamp_sec": 83.07, "offset_sec": 3, "rgb_path": "assets/cat_12/v_-YMpwZkNc2A_2400_2834/0003s_00088_rgb.jpg", "mask_path": "assets/cat_12/v_-YMpwZkNc2A_2400_2834/0003s_00088_mask.png", "mask_status": "available" }, { "frame": "00120", "timestamp_sec": 84.07, "offset_sec": 4, "rgb_path": "assets/cat_12/v_-YMpwZkNc2A_2400_2834/0004s_00120_rgb.jpg", "mask_path": "assets/cat_12/v_-YMpwZkNc2A_2400_2834/0004s_00120_mask.png", "mask_status": "available" }, { "frame": "00148", "timestamp_sec": 85.07, "offset_sec": 5, "rgb_path": "assets/cat_12/v_-YMpwZkNc2A_2400_2834/0005s_00148_rgb.jpg", "mask_path": "assets/cat_12/v_-YMpwZkNc2A_2400_2834/0005s_00148_mask.png", "mask_status": "available" }, { "frame": "00180", "timestamp_sec": 86.07, "offset_sec": 6, "rgb_path": "assets/cat_12/v_-YMpwZkNc2A_2400_2834/0006s_00180_rgb.jpg", "mask_path": "assets/cat_12/v_-YMpwZkNc2A_2400_2834/0006s_00180_mask.png", "mask_status": "available" }, { "frame": "00208", "timestamp_sec": 87.07, "offset_sec": 7, "rgb_path": "assets/cat_12/v_-YMpwZkNc2A_2400_2834/0007s_00208_rgb.jpg", "mask_path": "assets/cat_12/v_-YMpwZkNc2A_2400_2834/0007s_00208_mask.png", "mask_status": "available" }, { "frame": "00240", "timestamp_sec": 88.07, "offset_sec": 8, "rgb_path": "assets/cat_12/v_-YMpwZkNc2A_2400_2834/0008s_00240_rgb.jpg", "mask_path": "assets/cat_12/v_-YMpwZkNc2A_2400_2834/0008s_00240_mask.png", "mask_status": "available" }, { "frame": "00268", "timestamp_sec": 89.07, "offset_sec": 9, "rgb_path": "assets/cat_12/v_-YMpwZkNc2A_2400_2834/0009s_00268_rgb.jpg", "mask_path": "assets/cat_12/v_-YMpwZkNc2A_2400_2834/0009s_00268_mask.png", "mask_status": "available" }, { "frame": "00300", "timestamp_sec": 90.07, "offset_sec": 10, "rgb_path": "assets/cat_12/v_-YMpwZkNc2A_2400_2834/0010s_00300_rgb.jpg", "mask_path": "assets/cat_12/v_-YMpwZkNc2A_2400_2834/0010s_00300_mask.png", "mask_status": "available" }, { "frame": "00328", "timestamp_sec": 91.07, "offset_sec": 11, "rgb_path": "assets/cat_12/v_-YMpwZkNc2A_2400_2834/0011s_00328_rgb.jpg", "mask_path": "assets/cat_12/v_-YMpwZkNc2A_2400_2834/0011s_00328_mask.png", "mask_status": "available" }, { "frame": "00360", "timestamp_sec": 92.07, "offset_sec": 12, "rgb_path": "assets/cat_12/v_-YMpwZkNc2A_2400_2834/0012s_00360_rgb.jpg", "mask_path": "assets/cat_12/v_-YMpwZkNc2A_2400_2834/0012s_00360_mask.png", "mask_status": "available" }, { "frame": "00388", "timestamp_sec": 93.07, "offset_sec": 13, "rgb_path": "assets/cat_12/v_-YMpwZkNc2A_2400_2834/0013s_00388_rgb.jpg", "mask_path": "assets/cat_12/v_-YMpwZkNc2A_2400_2834/0013s_00388_mask.png", "mask_status": "available" }, { "frame": "00420", "timestamp_sec": 94.07, "offset_sec": 14, "rgb_path": "assets/cat_12/v_-YMpwZkNc2A_2400_2834/0014s_00420_rgb.jpg", "mask_path": "assets/cat_12/v_-YMpwZkNc2A_2400_2834/0014s_00420_mask.png", "mask_status": "available" } ], "obj_id": [ 0 ], "anno_id": [ 83 ], "fps_inferred_from_clip_interval": 29.931 }, "usage": { "input_tokens": 49273, "input_tokens_details": { "cache_write_tokens": 49270, "cached_tokens": 0 }, "output_tokens": 3918, "output_tokens_details": { "reasoning_tokens": 2588 }, "total_tokens": 53191 }, "model": "gpt-5.6-sol", "effort": "medium", "custom_id": "cat12_cap06_v_-YMpwZkNc2A_2400_2834", "annotation": { "caption": "The person continues shaving their face, focusing on the remaining patches of shaving cream on their cheeks and neck. After several strokes, they inspect the razor. They then continue to shave their neck.", "entity_groups": [ { "id": "person_1", "label": "person", "spans": [ { "text": "The person", "range": [ 0, 10 ] }, { "text": "their", "range": [ 29, 34 ] }, { "text": "their", "range": [ 95, 100 ] }, { "text": "they", "range": [ 141, 145 ] }, { "text": "They", "range": [ 165, 169 ] }, { "text": "their", "range": [ 193, 198 ] } ], "source_mask_link": "selected_person" }, { "id": "face_1", "label": "face", "spans": [ { "text": "their face", "range": [ 29, 39 ] } ], "source_mask_link": null }, { "id": "shaving_cream_1", "label": "shaving cream", "spans": [ { "text": "the remaining patches of shaving cream", "range": [ 53, 91 ] } ], "source_mask_link": null }, { "id": "cheeks_1", "label": "cheeks", "spans": [ { "text": "their cheeks", "range": [ 95, 107 ] } ], "source_mask_link": null }, { "id": "neck_1", "label": "neck", "spans": [ { "text": "neck", "range": [ 112, 116 ] }, { "text": "their neck", "range": [ 193, 203 ] } ], "source_mask_link": null }, { "id": "razor_1", "label": "razor", "spans": [ { "text": "the razor", "range": [ 154, 163 ] } ], "source_mask_link": null } ], "actions": [ { "id": "action_1", "mentions": [ { "text": "shaving", "ranges": [ [ 21, 28 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "type": "simple", "subject_refs": [ "person_1" ], "object_refs": [ "face_1" ], "confidence": "high", "temporal_status": "localized", "intervals": [ { "start_s": 0.0, "end_s": 1.0, "source_start_s": 80.07, "source_end_s": 81.07 }, { "start_s": 2.0, "end_s": 4.0, "source_start_s": 82.07, "source_end_s": 84.07 }, { "start_s": 5.0, "end_s": 6.0, "source_start_s": 85.07, "source_end_s": 86.07 } ], "evidence": "At 0, 2, 3, and 5 seconds, the mask-linked person visibly holds the razor against the cream-covered lower face or neck.", "mask_relation": "subject_matches_selected_person_mask", "source_subject_refs": "person_1" }, { "id": "action_2", "mentions": [ { "text": "inspect", "ranges": [ [ 146, 153 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "type": "simple", "subject_refs": [ "person_1" ], "object_refs": [ "razor_1" ], "confidence": "high", "temporal_status": "uncertain", "intervals": [], "evidence": "Around 6–7 seconds the person looks downward during a pause, but the razor is not clearly visible enough to temporally verify inspection.", "mask_relation": "subject_matches_selected_person_mask", "source_subject_refs": "person_1" }, { "id": "action_3", "mentions": [ { "text": "shave", "ranges": [ [ 187, 192 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "type": "simple", "subject_refs": [ "person_1" ], "object_refs": [ "neck_1" ], "confidence": "high", "temporal_status": "localized", "intervals": [ { "start_s": 8.0, "end_s": 12.0, "source_start_s": 88.07, "source_end_s": 92.07 }, { "start_s": 14.0, "end_s": 14.5, "source_start_s": 94.07, "source_end_s": 94.57 } ], "evidence": "At 8–11 seconds and again at 14 seconds, the mask-linked person visibly applies the razor to the side or front of the neck. The final supported interval reaches the clip boundary at 14.5 seconds.", "mask_relation": "subject_matches_selected_person_mask", "source_subject_refs": "person_1" } ], "notes": [ "The supplied mask consistently identifies the foreground person who performs the captioned actions.", "No separate masks were supplied for the face, cheeks, neck, shaving cream, or razor.", "The two shaving predicates are separate events because inspection intervenes before shaving resumes.", "The attention predicate \"focusing\" is omitted because it describes mental attention rather than an independently visible physical action.", "Source-video interval: [80.070000, 94.570000); clip-relative duration: 14.500000 seconds." ] }, "model_output_raw": { "caption": "The person continues shaving their face, focusing on the remaining patches of shaving cream on their cheeks and neck. After several strokes, they inspect the razor. They then continue to shave their neck.", "entity_groups": [ { "id": "person_1", "label": "person", "spans": [ { "text": "The person", "range": [ 0, 10 ] }, { "text": "their", "range": [ 29, 34 ] }, { "text": "their", "range": [ 95, 100 ] }, { "text": "they", "range": [ 141, 145 ] }, { "text": "They", "range": [ 165, 169 ] }, { "text": "their", "range": [ 193, 198 ] } ], "source_mask": "selected_person_mask" }, { "id": "face_1", "label": "face", "spans": [ { "text": "their face", "range": [ 29, 39 ] } ], "source_mask": null }, { "id": "shaving_cream_1", "label": "shaving cream", "spans": [ { "text": "the remaining patches of shaving cream", "range": [ 53, 91 ] } ], "source_mask": null }, { "id": "cheeks_1", "label": "cheeks", "spans": [ { "text": "their cheeks", "range": [ 95, 107 ] } ], "source_mask": null }, { "id": "neck_1", "label": "neck", "spans": [ { "text": "neck", "range": [ 112, 116 ] }, { "text": "their neck", "range": [ 193, 203 ] } ], "source_mask": null }, { "id": "razor_1", "label": "razor", "spans": [ { "text": "the razor", "range": [ 154, 163 ] } ], "source_mask": null } ], "actions": [ { "id": "action_1", "mentions": [ { "text": "shaving", "occurrence": 1, "ranges": [ [ 21, 28 ] ], "particle": { "text": null, "occurrence": 1 } } ], "type": "simple", "subject": "person_1", "objects": [ "face_1" ], "confidence": "high", "temporal_status": "localized", "intervals": [ { "start_s": 0.0, "end_s": 1.0 }, { "start_s": 2.0, "end_s": 4.0 }, { "start_s": 5.0, "end_s": 6.0 } ], "evidence": "At 0, 2, 3, and 5 seconds, the mask-linked person visibly holds the razor against the cream-covered lower face or neck.", "mask_relation": "subject_matches_selected_person_mask" }, { "id": "action_2", "mentions": [ { "text": "inspect", "occurrence": 1, "ranges": [ [ 146, 153 ] ], "particle": { "text": null, "occurrence": 1 } } ], "type": "simple", "subject": "person_1", "objects": [ "razor_1" ], "confidence": "high", "temporal_status": "uncertain", "intervals": [], "evidence": "Around 6–7 seconds the person looks downward during a pause, but the razor is not clearly visible enough to temporally verify inspection.", "mask_relation": "subject_matches_selected_person_mask" }, { "id": "action_3", "mentions": [ { "text": "shave", "occurrence": 1, "ranges": [ [ 187, 192 ] ], "particle": { "text": null, "occurrence": 1 } } ], "type": "simple", "subject": "person_1", "objects": [ "neck_1" ], "confidence": "high", "temporal_status": "localized", "intervals": [ { "start_s": 8.0, "end_s": 12.0 }, { "start_s": 14.0, "end_s": 14.5 } ], "evidence": "At 8–11 seconds and again at 14 seconds, the mask-linked person visibly applies the razor to the side or front of the neck. The final supported interval reaches the clip boundary at 14.5 seconds.", "mask_relation": "subject_matches_selected_person_mask" } ], "notes": [ "The supplied mask consistently identifies the foreground person who performs the captioned actions.", "No separate masks were supplied for the face, cheeks, neck, shaving cream, or razor.", "The two shaving predicates are separate events because inspection intervenes before shaving resumes.", "The attention predicate \"focusing\" is omitted because it describes mental attention rather than an independently visible physical action.", "Source-video interval: [80.070000, 94.570000); clip-relative duration: 14.500000 seconds." ] }, "normalization_repairs": 0 }, { "source_video_id": "v_-YMpwZkNc2A", "caption_count_category": 12, "clip": { "caption_index": 7, "caption": "The person finishes shaving and leans forward to rinse their face with water. They wipe away the remaining shaving cream from their face and neck. They check their reflection before wiping their face again.", "source_interval_sec": [ 94.57, 107.57 ], "loom_clip_id": "v_-YMpwZkNc2A_2834_3224", "clip_frame_interval_4fps": [ 2834, 3224 ], "sample_count_1fps": 13, "sampled_frames": [ { "frame": "00000", "timestamp_sec": 94.57, "offset_sec": 0, "rgb_path": "assets/cat_12/v_-YMpwZkNc2A_2834_3224/0000s_00000_rgb.jpg", "mask_path": "assets/cat_12/v_-YMpwZkNc2A_2834_3224/0000s_00000_mask.png", "mask_status": "available" }, { "frame": "00028", "timestamp_sec": 95.57, "offset_sec": 1, "rgb_path": "assets/cat_12/v_-YMpwZkNc2A_2834_3224/0001s_00028_rgb.jpg", "mask_path": "assets/cat_12/v_-YMpwZkNc2A_2834_3224/0001s_00028_mask.png", "mask_status": "available" }, { "frame": "00060", "timestamp_sec": 96.57, "offset_sec": 2, "rgb_path": "assets/cat_12/v_-YMpwZkNc2A_2834_3224/0002s_00060_rgb.jpg", "mask_path": "assets/cat_12/v_-YMpwZkNc2A_2834_3224/0002s_00060_mask.png", "mask_status": "available" }, { "frame": "00088", "timestamp_sec": 97.57, "offset_sec": 3, "rgb_path": "assets/cat_12/v_-YMpwZkNc2A_2834_3224/0003s_00088_rgb.jpg", "mask_path": "assets/cat_12/v_-YMpwZkNc2A_2834_3224/0003s_00088_mask.png", "mask_status": "available" }, { "frame": "00120", "timestamp_sec": 98.57, "offset_sec": 4, "rgb_path": "assets/cat_12/v_-YMpwZkNc2A_2834_3224/0004s_00120_rgb.jpg", "mask_path": "assets/cat_12/v_-YMpwZkNc2A_2834_3224/0004s_00120_mask.png", "mask_status": "available" }, { "frame": "00148", "timestamp_sec": 99.57, "offset_sec": 5, "rgb_path": "assets/cat_12/v_-YMpwZkNc2A_2834_3224/0005s_00148_rgb.jpg", "mask_path": "assets/cat_12/v_-YMpwZkNc2A_2834_3224/0005s_00148_mask.png", "mask_status": "available" }, { "frame": "00180", "timestamp_sec": 100.57, "offset_sec": 6, "rgb_path": "assets/cat_12/v_-YMpwZkNc2A_2834_3224/0006s_00180_rgb.jpg", "mask_path": "assets/cat_12/v_-YMpwZkNc2A_2834_3224/0006s_00180_mask.png", "mask_status": "available" }, { "frame": "00208", "timestamp_sec": 101.57, "offset_sec": 7, "rgb_path": "assets/cat_12/v_-YMpwZkNc2A_2834_3224/0007s_00208_rgb.jpg", "mask_path": "assets/cat_12/v_-YMpwZkNc2A_2834_3224/0007s_00208_mask.png", "mask_status": "available" }, { "frame": "00240", "timestamp_sec": 102.57, "offset_sec": 8, "rgb_path": "assets/cat_12/v_-YMpwZkNc2A_2834_3224/0008s_00240_rgb.jpg", "mask_path": "assets/cat_12/v_-YMpwZkNc2A_2834_3224/0008s_00240_mask.png", "mask_status": "available" }, { "frame": "00268", "timestamp_sec": 103.57, "offset_sec": 9, "rgb_path": "assets/cat_12/v_-YMpwZkNc2A_2834_3224/0009s_00268_rgb.jpg", "mask_path": "assets/cat_12/v_-YMpwZkNc2A_2834_3224/0009s_00268_mask.png", "mask_status": "available" }, { "frame": "00300", "timestamp_sec": 104.57, "offset_sec": 10, "rgb_path": "assets/cat_12/v_-YMpwZkNc2A_2834_3224/0010s_00300_rgb.jpg", "mask_path": "assets/cat_12/v_-YMpwZkNc2A_2834_3224/0010s_00300_mask.png", "mask_status": "available" }, { "frame": "00328", "timestamp_sec": 105.57, "offset_sec": 11, "rgb_path": "assets/cat_12/v_-YMpwZkNc2A_2834_3224/0011s_00328_rgb.jpg", "mask_path": "assets/cat_12/v_-YMpwZkNc2A_2834_3224/0011s_00328_mask.png", "mask_status": "available" }, { "frame": "00360", "timestamp_sec": 106.57, "offset_sec": 12, "rgb_path": "assets/cat_12/v_-YMpwZkNc2A_2834_3224/0012s_00360_rgb.jpg", "mask_path": "assets/cat_12/v_-YMpwZkNc2A_2834_3224/0012s_00360_mask.png", "mask_status": "available" } ], "obj_id": [ 0 ], "anno_id": [ 84 ], "fps_inferred_from_clip_interval": 30.0 }, "usage": { "input_tokens": 44747, "input_tokens_details": { "cache_write_tokens": 44744, "cached_tokens": 0 }, "output_tokens": 5478, "output_tokens_details": { "reasoning_tokens": 3411 }, "total_tokens": 50225 }, "model": "gpt-5.6-sol", "effort": "medium", "custom_id": "cat12_cap07_v_-YMpwZkNc2A_2834_3224", "annotation": { "caption": "The person finishes shaving and leans forward to rinse their face with water. They wipe away the remaining shaving cream from their face and neck. They check their reflection before wiping their face again.", "entity_groups": [ { "id": "entity_person", "label": "person", "spans": [ { "text": "The person", "range": [ 0, 10 ] }, { "text": "their", "range": [ 55, 60 ] }, { "text": "They", "range": [ 78, 82 ] }, { "text": "their", "range": [ 126, 131 ] }, { "text": "They", "range": [ 147, 151 ] }, { "text": "their", "range": [ 158, 163 ] }, { "text": "their", "range": [ 189, 194 ] } ], "source_mask_link": "selected_person" }, { "id": "entity_face", "label": "face", "spans": [ { "text": "their face", "range": [ 55, 65 ] }, { "text": "their face", "range": [ 126, 136 ] }, { "text": "their face", "range": [ 189, 199 ] } ], "source_mask_link": null }, { "id": "entity_water", "label": "water", "spans": [ { "text": "water", "range": [ 71, 76 ] } ], "source_mask_link": null }, { "id": "entity_shaving_cream", "label": "shaving cream", "spans": [ { "text": "the remaining shaving cream", "range": [ 93, 120 ] } ], "source_mask_link": null }, { "id": "entity_neck", "label": "neck", "spans": [ { "text": "neck", "range": [ 141, 145 ] } ], "source_mask_link": null }, { "id": "entity_reflection", "label": "reflection", "spans": [ { "text": "their reflection", "range": [ 158, 174 ] } ], "source_mask_link": null } ], "actions": [ { "id": "action_shaving", "mentions": [ { "text": "shaving", "ranges": [ [ 20, 27 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "type": "simple", "subject_refs": [ "entity_person" ], "object_refs": [ "entity_face" ], "confidence": "high", "temporal_status": "localized", "intervals": [ { "start_s": 0.0, "end_s": 2.0, "source_start_s": 94.57, "source_end_s": 96.57 } ], "evidence": "At 0.000 and 1.000 seconds, the masked person visibly moves a shaving tool against the neck and jaw.", "mask_relation": "subject matches the supplied selected-person mask", "source_subject_refs": [ "entity_person" ] }, { "id": "action_leaning", "mentions": [ { "text": "leans", "ranges": [ [ 32, 37 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "type": "simple", "subject_refs": [ "entity_person" ], "object_refs": [], "confidence": "high", "temporal_status": "localized", "intervals": [ { "start_s": 2.0, "end_s": 4.0, "source_start_s": 96.57, "source_end_s": 98.57 } ], "evidence": "At 2.000 and 3.000 seconds, the masked person visibly inclines his head and upper body toward the basin area.", "mask_relation": "subject matches the supplied selected-person mask", "source_subject_refs": [ "entity_person" ] }, { "id": "action_rinsing", "mentions": [ { "text": "rinse", "ranges": [ [ 49, 54 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "type": "simple", "subject_refs": [ "entity_person" ], "object_refs": [ "entity_face" ], "confidence": "medium", "temporal_status": "localized", "intervals": [ { "start_s": 4.0, "end_s": 5.0, "source_start_s": 98.57, "source_end_s": 99.57 } ], "evidence": "At 4.000 seconds, the masked person rubs his face with his hand after leaning toward the basin; the water itself is not clearly visible.", "mask_relation": "subject matches the supplied selected-person mask", "source_subject_refs": [ "entity_person" ] }, { "id": "action_wiping_cream", "mentions": [ { "text": "wipe away", "ranges": [ [ 83, 92 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "type": "phrasal", "subject_refs": [ "entity_person" ], "object_refs": [ "entity_shaving_cream" ], "confidence": "high", "temporal_status": "localized", "intervals": [ { "start_s": 6.0, "end_s": 7.0, "source_start_s": 100.57, "source_end_s": 101.57 } ], "evidence": "At 6.000 seconds, the masked person visibly wipes the cream-covered cheek with his hand.", "mask_relation": "subject matches the supplied selected-person mask", "source_subject_refs": [ "entity_person" ] }, { "id": "action_checking_reflection", "mentions": [ { "text": "check", "ranges": [ [ 152, 157 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "type": "simple", "subject_refs": [ "entity_person" ], "object_refs": [ "entity_reflection" ], "confidence": "medium", "temporal_status": "localized", "intervals": [ { "start_s": 7.0, "end_s": 12.0, "source_start_s": 101.57, "source_end_s": 106.57 } ], "evidence": "From 7.000 through 11.000 seconds, the masked person pauses wiping and repeatedly looks downward toward an off-frame reflective surface; the reflection is not visible.", "mask_relation": "subject matches the supplied selected-person mask", "source_subject_refs": [ "entity_person" ] }, { "id": "action_wiping_face_again", "mentions": [ { "text": "wiping", "ranges": [ [ 182, 188 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "type": "simple", "subject_refs": [ "entity_person" ], "object_refs": [ "entity_face" ], "confidence": "high", "temporal_status": "localized", "intervals": [ { "start_s": 12.0, "end_s": 13.0, "source_start_s": 106.57, "source_end_s": 107.57 } ], "evidence": "At 12.000 seconds, the masked person visibly wipes across his face; support reaches the clip's right boundary.", "mask_relation": "subject matches the supplied selected-person mask", "source_subject_refs": [ "entity_person" ] } ], "notes": [ "The supplied masklet consistently identifies the captioned person in all sampled frames.", "No separate masks were supplied for the face, neck, water, shaving cream, or reflection, so those entity groups have null source-mask links.", "The reflection and water are not directly visible in the sampled frames; their associated action localization is therefore less certain.", "The picture-in-picture content is not mentioned by the caption and was not treated as a caption entity." ] }, "model_output_raw": { "caption": "The person finishes shaving and leans forward to rinse their face with water. They wipe away the remaining shaving cream from their face and neck. They check their reflection before wiping their face again.", "entity_groups": [ { "id": "entity_person", "label": "person", "spans": [ { "text": "The person", "range": [ 0, 10 ] }, { "text": "their", "range": [ 55, 60 ] }, { "text": "They", "range": [ 78, 82 ] }, { "text": "their", "range": [ 126, 131 ] }, { "text": "They", "range": [ 147, 151 ] }, { "text": "their", "range": [ 158, 163 ] }, { "text": "their", "range": [ 189, 194 ] } ], "source_mask": "selected_person_masklet" }, { "id": "entity_face", "label": "face", "spans": [ { "text": "their face", "range": [ 55, 65 ] }, { "text": "their face", "range": [ 126, 136 ] }, { "text": "their face", "range": [ 189, 199 ] } ], "source_mask": null }, { "id": "entity_water", "label": "water", "spans": [ { "text": "water", "range": [ 71, 76 ] } ], "source_mask": null }, { "id": "entity_shaving_cream", "label": "shaving cream", "spans": [ { "text": "the remaining shaving cream", "range": [ 93, 120 ] } ], "source_mask": null }, { "id": "entity_neck", "label": "neck", "spans": [ { "text": "neck", "range": [ 141, 145 ] } ], "source_mask": null }, { "id": "entity_reflection", "label": "reflection", "spans": [ { "text": "their reflection", "range": [ 158, 174 ] } ], "source_mask": null } ], "actions": [ { "id": "action_shaving", "mentions": [ { "text": "shaving", "ranges": [ [ 20, 27 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "type": "simple", "subject_entity_group_ids": [ "entity_person" ], "object_entity_group_ids": [ "entity_face" ], "confidence": "high", "mask_relation": "subject matches the supplied selected-person mask", "temporal_status": "localized", "intervals": [ { "start_s": 0.0, "end_s": 2.0, "source_start_s": 94.57, "source_end_s": 96.57 } ], "evidence": "At 0.000 and 1.000 seconds, the masked person visibly moves a shaving tool against the neck and jaw." }, { "id": "action_leaning", "mentions": [ { "text": "leans", "ranges": [ [ 32, 37 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "type": "simple", "subject_entity_group_ids": [ "entity_person" ], "object_entity_group_ids": [], "confidence": "high", "mask_relation": "subject matches the supplied selected-person mask", "temporal_status": "localized", "intervals": [ { "start_s": 2.0, "end_s": 4.0, "source_start_s": 96.57, "source_end_s": 98.57 } ], "evidence": "At 2.000 and 3.000 seconds, the masked person visibly inclines his head and upper body toward the basin area." }, { "id": "action_rinsing", "mentions": [ { "text": "rinse", "ranges": [ [ 49, 54 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "type": "simple", "subject_entity_group_ids": [ "entity_person" ], "object_entity_group_ids": [ "entity_face" ], "confidence": "medium", "mask_relation": "subject matches the supplied selected-person mask", "temporal_status": "localized", "intervals": [ { "start_s": 4.0, "end_s": 5.0, "source_start_s": 98.57, "source_end_s": 99.57 } ], "evidence": "At 4.000 seconds, the masked person rubs his face with his hand after leaning toward the basin; the water itself is not clearly visible." }, { "id": "action_wiping_cream", "mentions": [ { "text": "wipe away", "ranges": [ [ 83, 92 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "type": "phrasal", "subject_entity_group_ids": [ "entity_person" ], "object_entity_group_ids": [ "entity_shaving_cream" ], "confidence": "high", "mask_relation": "subject matches the supplied selected-person mask", "temporal_status": "localized", "intervals": [ { "start_s": 6.0, "end_s": 7.0, "source_start_s": 100.57, "source_end_s": 101.57 } ], "evidence": "At 6.000 seconds, the masked person visibly wipes the cream-covered cheek with his hand." }, { "id": "action_checking_reflection", "mentions": [ { "text": "check", "ranges": [ [ 152, 157 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "type": "simple", "subject_entity_group_ids": [ "entity_person" ], "object_entity_group_ids": [ "entity_reflection" ], "confidence": "medium", "mask_relation": "subject matches the supplied selected-person mask", "temporal_status": "localized", "intervals": [ { "start_s": 7.0, "end_s": 12.0, "source_start_s": 101.57, "source_end_s": 106.57 } ], "evidence": "From 7.000 through 11.000 seconds, the masked person pauses wiping and repeatedly looks downward toward an off-frame reflective surface; the reflection is not visible." }, { "id": "action_wiping_face_again", "mentions": [ { "text": "wiping", "ranges": [ [ 182, 188 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "type": "simple", "subject_entity_group_ids": [ "entity_person" ], "object_entity_group_ids": [ "entity_face" ], "confidence": "high", "mask_relation": "subject matches the supplied selected-person mask", "temporal_status": "localized", "intervals": [ { "start_s": 12.0, "end_s": 13.0, "source_start_s": 106.57, "source_end_s": 107.57 } ], "evidence": "At 12.000 seconds, the masked person visibly wipes across his face; support reaches the clip's right boundary." } ], "notes": [ "The supplied masklet consistently identifies the captioned person in all sampled frames.", "No separate masks were supplied for the face, neck, water, shaving cream, or reflection, so those entity groups have null source-mask links.", "The reflection and water are not directly visible in the sampled frames; their associated action localization is therefore less certain.", "The picture-in-picture content is not mentioned by the caption and was not treated as a caption entity." ] }, "normalization_repairs": 0 }, { "source_video_id": "v_-YMpwZkNc2A", "caption_count_category": 12, "clip": { "caption_index": 8, "caption": "The person, now mostly clean-shaven with a remaining goatee, uses the razor to carefully trim their mustache and shape their goatee. They then lean forward and move out of view.", "source_interval_sec": [ 107.57, 119.39 ], "loom_clip_id": "v_-YMpwZkNc2A_3224_3578", "clip_frame_interval_4fps": [ 3224, 3578 ], "sample_count_1fps": 12, "sampled_frames": [ { "frame": "00000", "timestamp_sec": 107.57, "offset_sec": 0, "rgb_path": "assets/cat_12/v_-YMpwZkNc2A_3224_3578/0000s_00000_rgb.jpg", "mask_path": "assets/cat_12/v_-YMpwZkNc2A_3224_3578/0000s_00000_mask.png", "mask_status": "available" }, { "frame": "00028", "timestamp_sec": 108.57, "offset_sec": 1, "rgb_path": "assets/cat_12/v_-YMpwZkNc2A_3224_3578/0001s_00028_rgb.jpg", "mask_path": "assets/cat_12/v_-YMpwZkNc2A_3224_3578/0001s_00028_mask.png", "mask_status": "available" }, { "frame": "00060", "timestamp_sec": 109.57, "offset_sec": 2, "rgb_path": "assets/cat_12/v_-YMpwZkNc2A_3224_3578/0002s_00060_rgb.jpg", "mask_path": "assets/cat_12/v_-YMpwZkNc2A_3224_3578/0002s_00060_mask.png", "mask_status": "available" }, { "frame": "00088", "timestamp_sec": 110.57, "offset_sec": 3, "rgb_path": "assets/cat_12/v_-YMpwZkNc2A_3224_3578/0003s_00088_rgb.jpg", "mask_path": "assets/cat_12/v_-YMpwZkNc2A_3224_3578/0003s_00088_mask.png", "mask_status": "available" }, { "frame": "00120", "timestamp_sec": 111.57, "offset_sec": 4, "rgb_path": "assets/cat_12/v_-YMpwZkNc2A_3224_3578/0004s_00120_rgb.jpg", "mask_path": "assets/cat_12/v_-YMpwZkNc2A_3224_3578/0004s_00120_mask.png", "mask_status": "available" }, { "frame": "00148", "timestamp_sec": 112.57, "offset_sec": 5, "rgb_path": "assets/cat_12/v_-YMpwZkNc2A_3224_3578/0005s_00148_rgb.jpg", "mask_path": "assets/cat_12/v_-YMpwZkNc2A_3224_3578/0005s_00148_mask.png", "mask_status": "available" }, { "frame": "00180", "timestamp_sec": 113.57, "offset_sec": 6, "rgb_path": "assets/cat_12/v_-YMpwZkNc2A_3224_3578/0006s_00180_rgb.jpg", "mask_path": "assets/cat_12/v_-YMpwZkNc2A_3224_3578/0006s_00180_mask.png", "mask_status": "available" }, { "frame": "00208", "timestamp_sec": 114.57, "offset_sec": 7, "rgb_path": "assets/cat_12/v_-YMpwZkNc2A_3224_3578/0007s_00208_rgb.jpg", "mask_path": "assets/cat_12/v_-YMpwZkNc2A_3224_3578/0007s_00208_mask.png", "mask_status": "available" }, { "frame": "00240", "timestamp_sec": 115.57, "offset_sec": 8, "rgb_path": "assets/cat_12/v_-YMpwZkNc2A_3224_3578/0008s_00240_rgb.jpg", "mask_path": "assets/cat_12/v_-YMpwZkNc2A_3224_3578/0008s_00240_mask.png", "mask_status": "available" }, { "frame": "00268", "timestamp_sec": 116.57, "offset_sec": 9, "rgb_path": "assets/cat_12/v_-YMpwZkNc2A_3224_3578/0009s_00268_rgb.jpg", "mask_path": "assets/cat_12/v_-YMpwZkNc2A_3224_3578/0009s_00268_mask.png", "mask_status": "available" }, { "frame": "00300", "timestamp_sec": 117.57, "offset_sec": 10, "rgb_path": "assets/cat_12/v_-YMpwZkNc2A_3224_3578/0010s_00300_rgb.jpg", "mask_path": "assets/cat_12/v_-YMpwZkNc2A_3224_3578/0010s_00300_mask.png", "mask_status": "available" }, { "frame": "00328", "timestamp_sec": 118.57, "offset_sec": 11, "rgb_path": "assets/cat_12/v_-YMpwZkNc2A_3224_3578/0011s_00328_rgb.jpg", "mask_path": "assets/cat_12/v_-YMpwZkNc2A_3224_3578/0011s_00328_mask.png", "mask_status": "available" } ], "obj_id": [ 0 ], "anno_id": [ 85 ], "fps_inferred_from_clip_interval": 29.9492 }, "usage": { "input_tokens": 42485, "input_tokens_details": { "cache_write_tokens": 42482, "cached_tokens": 0 }, "output_tokens": 4127, "output_tokens_details": { "reasoning_tokens": 2551 }, "total_tokens": 46612 }, "model": "gpt-5.6-sol", "effort": "medium", "custom_id": "cat12_cap08_v_-YMpwZkNc2A_3224_3578", "annotation": { "caption": "The person, now mostly clean-shaven with a remaining goatee, uses the razor to carefully trim their mustache and shape their goatee. They then lean forward and move out of view.", "entity_groups": [ { "id": "entity_1", "label": "person", "spans": [ { "text": "The person", "range": [ 0, 10 ] }, { "text": "their", "range": [ 94, 99 ] }, { "text": "their", "range": [ 119, 124 ] }, { "text": "They", "range": [ 133, 137 ] } ], "source_mask_link": "selected_person" }, { "id": "entity_2", "label": "goatee", "spans": [ { "text": "a remaining goatee", "range": [ 41, 59 ] }, { "text": "their goatee", "range": [ 119, 131 ] } ], "source_mask_link": null }, { "id": "entity_3", "label": "razor", "spans": [ { "text": "the razor", "range": [ 66, 75 ] } ], "source_mask_link": null }, { "id": "entity_4", "label": "mustache", "spans": [ { "text": "their mustache", "range": [ 94, 108 ] } ], "source_mask_link": null } ], "actions": [ { "id": "action_1", "mentions": [ { "text": "uses", "ranges": [ [ 61, 65 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "type": "simple", "subject_refs": [ "entity_1" ], "object_refs": [ "entity_3" ], "confidence": "high", "temporal_status": "localized", "intervals": [ { "start_s": 0.0, "end_s": 1.0, "source_start_s": 107.57, "source_end_s": 108.57 }, { "start_s": 2.0, "end_s": 5.0, "source_start_s": 109.57, "source_end_s": 112.57 }, { "start_s": 6.0, "end_s": 9.0, "source_start_s": 113.57, "source_end_s": 116.57 } ], "evidence": "The mask-linked person visibly holds and applies a razor near the upper lip, jaw, and neck at the supported samples.", "mask_relation": "subject matches the supplied selected-person mask", "source_subject_refs": "entity_1" }, { "id": "action_2", "mentions": [ { "text": "trim", "ranges": [ [ 89, 93 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "type": "simple", "subject_refs": [ "entity_1" ], "object_refs": [ "entity_4" ], "confidence": "high", "temporal_status": "localized", "intervals": [ { "start_s": 0.0, "end_s": 1.0, "source_start_s": 107.57, "source_end_s": 108.57 }, { "start_s": 6.0, "end_s": 9.0, "source_start_s": 113.57, "source_end_s": 116.57 } ], "evidence": "The selected person applies the razor directly around the upper lip at 0 and 6–8 seconds.", "mask_relation": "subject matches the supplied selected-person mask", "source_subject_refs": "entity_1" }, { "id": "action_3", "mentions": [ { "text": "shape", "ranges": [ [ 113, 118 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "type": "simple", "subject_refs": [ "entity_1" ], "object_refs": [ "entity_2" ], "confidence": "high", "temporal_status": "localized", "intervals": [ { "start_s": 2.0, "end_s": 5.0, "source_start_s": 109.57, "source_end_s": 112.57 } ], "evidence": "At 2–4 seconds, the mask-linked person runs the razor along the hair around the chin and neck.", "mask_relation": "subject matches the supplied selected-person mask", "source_subject_refs": "entity_1" }, { "id": "action_4", "mentions": [ { "text": "lean", "ranges": [ [ 143, 147 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "type": "posture", "subject_refs": [ "entity_1" ], "object_refs": [], "confidence": "high", "temporal_status": "localized", "intervals": [ { "start_s": 10.0, "end_s": 11.0, "source_start_s": 117.57, "source_end_s": 118.57 } ], "evidence": "The selected person bends the head and upper body forward in the 10-second sample.", "mask_relation": "subject matches the supplied selected-person mask", "source_subject_refs": "entity_1" }, { "id": "action_5", "mentions": [ { "text": "move", "ranges": [ [ 160, 164 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "type": "simple", "subject_refs": [ "entity_1" ], "object_refs": [], "confidence": "high", "temporal_status": "localized", "intervals": [ { "start_s": 11.0, "end_s": 11.820000000000007, "source_start_s": 118.57, "source_end_s": 119.39 } ], "evidence": "By the 11-second sample, only a small portion of the selected person remains at the right edge, supporting movement out of view through the clip boundary.", "mask_relation": "subject matches the supplied selected-person mask", "source_subject_refs": "entity_1" } ], "notes": [ "The supplied masklet consistently identifies the captioned person.", "No separate source masks were supplied for the razor, mustache, or goatee, so those entity groups have null source-mask links.", "Action intervals are sparse clip-relative support intervals derived from the sampled clean RGB frames. The final movement interval ends exactly at the 11.82-second clip boundary." ] }, "model_output_raw": { "caption": "The person, now mostly clean-shaven with a remaining goatee, uses the razor to carefully trim their mustache and shape their goatee. They then lean forward and move out of view.", "entity_groups": [ { "id": "entity_1", "label": "person", "spans": [ { "text": "The person", "range": [ 0, 10 ] }, { "text": "their", "range": [ 94, 99 ] }, { "text": "their", "range": [ 119, 124 ] }, { "text": "They", "range": [ 133, 137 ] } ], "source_mask": "selected_person_masklet" }, { "id": "entity_2", "label": "goatee", "spans": [ { "text": "a remaining goatee", "range": [ 41, 59 ] }, { "text": "their goatee", "range": [ 119, 131 ] } ], "source_mask": null }, { "id": "entity_3", "label": "razor", "spans": [ { "text": "the razor", "range": [ 66, 75 ] } ], "source_mask": null }, { "id": "entity_4", "label": "mustache", "spans": [ { "text": "their mustache", "range": [ 94, 108 ] } ], "source_mask": null } ], "actions": [ { "id": "action_1", "mentions": [ { "text": "uses", "ranges": [ [ 61, 65 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "type": "simple", "subject_entity_group": "entity_1", "object_entity_groups": [ "entity_3" ], "instrument_entity_groups": [], "confidence": "high", "mask_relation": "subject matches the supplied selected-person mask", "temporal_status": "localized", "intervals": [ { "start_s": 0.0, "end_s": 1.0 }, { "start_s": 2.0, "end_s": 5.0 }, { "start_s": 6.0, "end_s": 9.0 } ], "evidence": "The mask-linked person visibly holds and applies a razor near the upper lip, jaw, and neck at the supported samples." }, { "id": "action_2", "mentions": [ { "text": "trim", "ranges": [ [ 89, 93 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "type": "simple", "subject_entity_group": "entity_1", "object_entity_groups": [ "entity_4" ], "instrument_entity_groups": [ "entity_3" ], "confidence": "high", "mask_relation": "subject matches the supplied selected-person mask", "temporal_status": "localized", "intervals": [ { "start_s": 0.0, "end_s": 1.0 }, { "start_s": 6.0, "end_s": 9.0 } ], "evidence": "The selected person applies the razor directly around the upper lip at 0 and 6–8 seconds." }, { "id": "action_3", "mentions": [ { "text": "shape", "ranges": [ [ 113, 118 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "type": "simple", "subject_entity_group": "entity_1", "object_entity_groups": [ "entity_2" ], "instrument_entity_groups": [ "entity_3" ], "confidence": "high", "mask_relation": "subject matches the supplied selected-person mask", "temporal_status": "localized", "intervals": [ { "start_s": 2.0, "end_s": 5.0 } ], "evidence": "At 2–4 seconds, the mask-linked person runs the razor along the hair around the chin and neck." }, { "id": "action_4", "mentions": [ { "text": "lean", "ranges": [ [ 143, 147 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "type": "posture", "subject_entity_group": "entity_1", "object_entity_groups": [], "instrument_entity_groups": [], "confidence": "high", "mask_relation": "subject matches the supplied selected-person mask", "temporal_status": "localized", "intervals": [ { "start_s": 10.0, "end_s": 11.0 } ], "evidence": "The selected person bends the head and upper body forward in the 10-second sample." }, { "id": "action_5", "mentions": [ { "text": "move", "ranges": [ [ 160, 164 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "type": "simple", "subject_entity_group": "entity_1", "object_entity_groups": [], "instrument_entity_groups": [], "confidence": "high", "mask_relation": "subject matches the supplied selected-person mask", "temporal_status": "localized", "intervals": [ { "start_s": 11.0, "end_s": 11.82 } ], "evidence": "By the 11-second sample, only a small portion of the selected person remains at the right edge, supporting movement out of view through the clip boundary." } ], "notes": [ "The supplied masklet consistently identifies the captioned person.", "No separate source masks were supplied for the razor, mustache, or goatee, so those entity groups have null source-mask links.", "Action intervals are sparse clip-relative support intervals derived from the sampled clean RGB frames. The final movement interval ends exactly at the 11.82-second clip boundary." ] }, "normalization_repairs": 0 }, { "source_video_id": "v_-YMpwZkNc2A", "caption_count_category": 12, "clip": { "caption_index": 9, "caption": "A man in a blue t-shirt and a goatee enters the frame from the right. He lifts a brown towel with both hands and begins to dry his face. He rubs his entire face with the towel before moving it down to dry his neck.", "source_interval_sec": [ 121.25, 129.25 ], "loom_clip_id": "v_-YMpwZkNc2A_3634_3874", "clip_frame_interval_4fps": [ 3634, 3874 ], "sample_count_1fps": 8, "sampled_frames": [ { "frame": "00000", "timestamp_sec": 121.25, "offset_sec": 0, "rgb_path": "assets/cat_12/v_-YMpwZkNc2A_3634_3874/0000s_00000_rgb.jpg", "mask_path": "assets/cat_12/v_-YMpwZkNc2A_3634_3874/0000s_00000_mask.png", "mask_status": "available" }, { "frame": "00028", "timestamp_sec": 122.25, "offset_sec": 1, "rgb_path": "assets/cat_12/v_-YMpwZkNc2A_3634_3874/0001s_00028_rgb.jpg", "mask_path": "assets/cat_12/v_-YMpwZkNc2A_3634_3874/0001s_00028_mask.png", "mask_status": "available" }, { "frame": "00060", "timestamp_sec": 123.25, "offset_sec": 2, "rgb_path": "assets/cat_12/v_-YMpwZkNc2A_3634_3874/0002s_00060_rgb.jpg", "mask_path": "assets/cat_12/v_-YMpwZkNc2A_3634_3874/0002s_00060_mask.png", "mask_status": "available" }, { "frame": "00088", "timestamp_sec": 124.25, "offset_sec": 3, "rgb_path": "assets/cat_12/v_-YMpwZkNc2A_3634_3874/0003s_00088_rgb.jpg", "mask_path": "assets/cat_12/v_-YMpwZkNc2A_3634_3874/0003s_00088_mask.png", "mask_status": "available" }, { "frame": "00120", "timestamp_sec": 125.25, "offset_sec": 4, "rgb_path": "assets/cat_12/v_-YMpwZkNc2A_3634_3874/0004s_00120_rgb.jpg", "mask_path": "assets/cat_12/v_-YMpwZkNc2A_3634_3874/0004s_00120_mask.png", "mask_status": "available" }, { "frame": "00148", "timestamp_sec": 126.25, "offset_sec": 5, "rgb_path": "assets/cat_12/v_-YMpwZkNc2A_3634_3874/0005s_00148_rgb.jpg", "mask_path": "assets/cat_12/v_-YMpwZkNc2A_3634_3874/0005s_00148_mask.png", "mask_status": "available" }, { "frame": "00180", "timestamp_sec": 127.25, "offset_sec": 6, "rgb_path": "assets/cat_12/v_-YMpwZkNc2A_3634_3874/0006s_00180_rgb.jpg", "mask_path": "assets/cat_12/v_-YMpwZkNc2A_3634_3874/0006s_00180_mask.png", "mask_status": "available" }, { "frame": "00208", "timestamp_sec": 128.25, "offset_sec": 7, "rgb_path": "assets/cat_12/v_-YMpwZkNc2A_3634_3874/0007s_00208_rgb.jpg", "mask_path": "assets/cat_12/v_-YMpwZkNc2A_3634_3874/0007s_00208_mask.png", "mask_status": "available" } ], "obj_id": [ 0 ], "anno_id": [ 86 ], "fps_inferred_from_clip_interval": 30.0 }, "usage": { "input_tokens": 33445, "input_tokens_details": { "cache_write_tokens": 33442, "cached_tokens": 0 }, "output_tokens": 4790, "output_tokens_details": { "reasoning_tokens": 2748 }, "total_tokens": 38235 }, "model": "gpt-5.6-sol", "effort": "medium", "custom_id": "cat12_cap09_v_-YMpwZkNc2A_3634_3874", "annotation": { "caption": "A man in a blue t-shirt and a goatee enters the frame from the right. He lifts a brown towel with both hands and begins to dry his face. He rubs his entire face with the towel before moving it down to dry his neck.", "entity_groups": [ { "id": "entity_man", "label": "man", "spans": [ { "text": "A man", "range": [ 0, 5 ] }, { "text": "He", "range": [ 70, 72 ] }, { "text": "his", "range": [ 127, 130 ] }, { "text": "He", "range": [ 137, 139 ] }, { "text": "his", "range": [ 145, 148 ] }, { "text": "his", "range": [ 205, 208 ] } ], "source_mask_link": "selected_person" }, { "id": "entity_tshirt", "label": "t-shirt", "spans": [ { "text": "a blue t-shirt", "range": [ 9, 23 ] } ], "source_mask_link": null }, { "id": "entity_goatee", "label": "goatee", "spans": [ { "text": "a goatee", "range": [ 28, 36 ] } ], "source_mask_link": null }, { "id": "entity_towel", "label": "towel", "spans": [ { "text": "a brown towel", "range": [ 79, 92 ] }, { "text": "the towel", "range": [ 166, 175 ] }, { "text": "it", "range": [ 190, 192 ] } ], "source_mask_link": null }, { "id": "entity_hands", "label": "hands", "spans": [ { "text": "both hands", "range": [ 98, 108 ] } ], "source_mask_link": null }, { "id": "entity_face", "label": "face", "spans": [ { "text": "his face", "range": [ 127, 135 ] }, { "text": "his entire face", "range": [ 145, 160 ] } ], "source_mask_link": null }, { "id": "entity_neck", "label": "neck", "spans": [ { "text": "his neck", "range": [ 205, 213 ] } ], "source_mask_link": null } ], "actions": [ { "id": "action_enter", "mentions": [ { "text": "enters", "ranges": [ [ 37, 43 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "type": "simple", "subject_refs": [ "entity_man" ], "object_refs": [], "confidence": "medium", "temporal_status": "localized", "intervals": [ { "start_s": 0.0, "end_s": 2.0, "source_start_s": 121.25, "source_end_s": 123.25 } ], "evidence": "The room is empty at 0.000 s; by 1.000 s the mask-linked man has entered from the right side.", "mask_relation": "subject_matches_selected_person_mask", "source_subject_refs": "entity_man" }, { "id": "action_lift_towel", "mentions": [ { "text": "lifts", "ranges": [ [ 73, 78 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "type": "simple", "subject_refs": [ "entity_man" ], "object_refs": [ "entity_towel" ], "confidence": "medium", "temporal_status": "localized", "intervals": [ { "start_s": 1.0, "end_s": 3.0, "source_start_s": 122.25, "source_end_s": 124.25 } ], "evidence": "The man has no raised towel at 1.000 s and holds the brown towel up against his face at 2.000 s.", "mask_relation": "subject_matches_selected_person_mask", "source_subject_refs": "entity_man" }, { "id": "action_dry_face", "mentions": [ { "text": "dry", "ranges": [ [ 123, 126 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "type": "simple", "subject_refs": [ "entity_man" ], "object_refs": [ "entity_face" ], "confidence": "high", "temporal_status": "localized", "intervals": [ { "start_s": 2.0, "end_s": 4.0, "source_start_s": 123.25, "source_end_s": 125.25 } ], "evidence": "At 2.000 and 3.000 s, the mask-linked man presses and moves the towel over his face.", "mask_relation": "subject_matches_selected_person_mask", "source_subject_refs": "entity_man" }, { "id": "action_rub_face", "mentions": [ { "text": "rubs", "ranges": [ [ 140, 144 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "type": "simple", "subject_refs": [ "entity_man" ], "object_refs": [ "entity_face" ], "confidence": "high", "temporal_status": "localized", "intervals": [ { "start_s": 2.0, "end_s": 4.0, "source_start_s": 123.25, "source_end_s": 125.25 } ], "evidence": "The towel is held against different areas of the man's face at 2.000 and 3.000 s, supporting rubbing motion.", "mask_relation": "subject_matches_selected_person_mask", "source_subject_refs": "entity_man" }, { "id": "action_move_towel_down", "mentions": [ { "text": "moving", "ranges": [ [ 183, 189 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "type": "simple", "subject_refs": [ "entity_man" ], "object_refs": [ "entity_towel" ], "confidence": "high", "temporal_status": "localized", "intervals": [ { "start_s": 3.0, "end_s": 5.0, "source_start_s": 124.25, "source_end_s": 126.25 } ], "evidence": "The towel is at the face at 3.000 s and has moved down to the neck by 4.000 s.", "mask_relation": "subject_matches_selected_person_mask", "source_subject_refs": "entity_man" }, { "id": "action_dry_neck", "mentions": [ { "text": "dry", "ranges": [ [ 201, 204 ] ], "occurrence": 2, "particle": { "text": null, "occurrence": 1 } } ], "type": "simple", "subject_refs": [ "entity_man" ], "object_refs": [ "entity_neck" ], "confidence": "high", "temporal_status": "localized", "intervals": [ { "start_s": 4.0, "end_s": 6.0, "source_start_s": 125.25, "source_end_s": 127.25 } ], "evidence": "At 4.000 and 5.000 s, the mask-linked man rubs the brown towel along the front and side of his neck.", "mask_relation": "subject_matches_selected_person_mask", "source_subject_refs": "entity_man" } ], "notes": [ "The supplied masklet identifies the single man from 1.000 through 7.000 s; it is empty at 0.000 s because he has not yet entered.", "No separate source masks were supplied for the t-shirt, goatee, towel, hands, face, or neck, so those entity groups have null source-mask links.", "The control predicate \"begins\" is excluded; its embedded physical action \"dry\" is annotated.", "The directional modifier \"down\" is excluded from the action span for \"moving\"." ] }, "model_output_raw": { "caption": "A man in a blue t-shirt and a goatee enters the frame from the right. He lifts a brown towel with both hands and begins to dry his face. He rubs his entire face with the towel before moving it down to dry his neck.", "entity_groups": [ { "id": "entity_man", "label": "man", "spans": [ { "text": "A man", "range": [ 0, 5 ] }, { "text": "He", "range": [ 70, 72 ] }, { "text": "his", "range": [ 127, 130 ] }, { "text": "He", "range": [ 137, 139 ] }, { "text": "his", "range": [ 145, 148 ] }, { "text": "his", "range": [ 205, 208 ] } ], "source_mask": "selected_person_masklet" }, { "id": "entity_tshirt", "label": "t-shirt", "spans": [ { "text": "a blue t-shirt", "range": [ 9, 23 ] } ], "source_mask": null }, { "id": "entity_goatee", "label": "goatee", "spans": [ { "text": "a goatee", "range": [ 28, 36 ] } ], "source_mask": null }, { "id": "entity_towel", "label": "towel", "spans": [ { "text": "a brown towel", "range": [ 79, 92 ] }, { "text": "the towel", "range": [ 166, 175 ] }, { "text": "it", "range": [ 190, 192 ] } ], "source_mask": null }, { "id": "entity_hands", "label": "hands", "spans": [ { "text": "both hands", "range": [ 98, 108 ] } ], "source_mask": null }, { "id": "entity_face", "label": "face", "spans": [ { "text": "his face", "range": [ 127, 135 ] }, { "text": "his entire face", "range": [ 145, 160 ] } ], "source_mask": null }, { "id": "entity_neck", "label": "neck", "spans": [ { "text": "his neck", "range": [ 205, 213 ] } ], "source_mask": null } ], "actions": [ { "id": "action_enter", "mentions": [ { "text": "enters", "ranges": [ [ 37, 43 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "type": "simple", "subject": "entity_man", "objects": [], "confidence": "medium", "mask_relation": "subject_matches_selected_person_mask", "temporal_status": "localized", "intervals": [ { "start_s": 0.0, "end_s": 2.0, "source_start_s": 121.25, "source_end_s": 123.25 } ], "evidence": "The room is empty at 0.000 s; by 1.000 s the mask-linked man has entered from the right side." }, { "id": "action_lift_towel", "mentions": [ { "text": "lifts", "ranges": [ [ 73, 78 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "type": "simple", "subject": "entity_man", "objects": [ "entity_towel" ], "confidence": "medium", "mask_relation": "subject_matches_selected_person_mask", "temporal_status": "localized", "intervals": [ { "start_s": 1.0, "end_s": 3.0, "source_start_s": 122.25, "source_end_s": 124.25 } ], "evidence": "The man has no raised towel at 1.000 s and holds the brown towel up against his face at 2.000 s." }, { "id": "action_dry_face", "mentions": [ { "text": "dry", "ranges": [ [ 123, 126 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "type": "simple", "subject": "entity_man", "objects": [ "entity_face" ], "confidence": "high", "mask_relation": "subject_matches_selected_person_mask", "temporal_status": "localized", "intervals": [ { "start_s": 2.0, "end_s": 4.0, "source_start_s": 123.25, "source_end_s": 125.25 } ], "evidence": "At 2.000 and 3.000 s, the mask-linked man presses and moves the towel over his face." }, { "id": "action_rub_face", "mentions": [ { "text": "rubs", "ranges": [ [ 140, 144 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "type": "simple", "subject": "entity_man", "objects": [ "entity_face" ], "confidence": "high", "mask_relation": "subject_matches_selected_person_mask", "temporal_status": "localized", "intervals": [ { "start_s": 2.0, "end_s": 4.0, "source_start_s": 123.25, "source_end_s": 125.25 } ], "evidence": "The towel is held against different areas of the man's face at 2.000 and 3.000 s, supporting rubbing motion." }, { "id": "action_move_towel_down", "mentions": [ { "text": "moving", "ranges": [ [ 183, 189 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "type": "simple", "subject": "entity_man", "objects": [ "entity_towel" ], "confidence": "high", "mask_relation": "subject_matches_selected_person_mask", "temporal_status": "localized", "intervals": [ { "start_s": 3.0, "end_s": 5.0, "source_start_s": 124.25, "source_end_s": 126.25 } ], "evidence": "The towel is at the face at 3.000 s and has moved down to the neck by 4.000 s." }, { "id": "action_dry_neck", "mentions": [ { "text": "dry", "ranges": [ [ 201, 204 ] ], "occurrence": 2, "particle": { "text": null, "occurrence": 1 } } ], "type": "simple", "subject": "entity_man", "objects": [ "entity_neck" ], "confidence": "high", "mask_relation": "subject_matches_selected_person_mask", "temporal_status": "localized", "intervals": [ { "start_s": 4.0, "end_s": 6.0, "source_start_s": 125.25, "source_end_s": 127.25 } ], "evidence": "At 4.000 and 5.000 s, the mask-linked man rubs the brown towel along the front and side of his neck." } ], "notes": [ "The supplied masklet identifies the single man from 1.000 through 7.000 s; it is empty at 0.000 s because he has not yet entered.", "No separate source masks were supplied for the t-shirt, goatee, towel, hands, face, or neck, so those entity groups have null source-mask links.", "The control predicate \"begins\" is excluded; its embedded physical action \"dry\" is annotated.", "The directional modifier \"down\" is excluded from the action span for \"moving\"." ] }, "normalization_repairs": 0 }, { "source_video_id": "v_-YMpwZkNc2A", "caption_count_category": 12, "clip": { "caption_index": 10, "caption": "The man lowers the towel out of view and runs his right hand through his hair. He then stands looking forward, turns his head slightly, and talks. He briefly touches his chin with his thumb while continuing to speak.", "source_interval_sec": [ 129.25, 139.75 ], "loom_clip_id": "v_-YMpwZkNc2A_3874_4188", "clip_frame_interval_4fps": [ 3874, 4188 ], "sample_count_1fps": 11, "sampled_frames": [ { "frame": "00000", "timestamp_sec": 129.25, "offset_sec": 0, "rgb_path": "assets/cat_12/v_-YMpwZkNc2A_3874_4188/0000s_00000_rgb.jpg", "mask_path": "assets/cat_12/v_-YMpwZkNc2A_3874_4188/0000s_00000_mask.png", "mask_status": "available" }, { "frame": "00028", "timestamp_sec": 130.25, "offset_sec": 1, "rgb_path": "assets/cat_12/v_-YMpwZkNc2A_3874_4188/0001s_00028_rgb.jpg", "mask_path": "assets/cat_12/v_-YMpwZkNc2A_3874_4188/0001s_00028_mask.png", "mask_status": "available" }, { "frame": "00060", "timestamp_sec": 131.25, "offset_sec": 2, "rgb_path": "assets/cat_12/v_-YMpwZkNc2A_3874_4188/0002s_00060_rgb.jpg", "mask_path": "assets/cat_12/v_-YMpwZkNc2A_3874_4188/0002s_00060_mask.png", "mask_status": "available" }, { "frame": "00088", "timestamp_sec": 132.25, "offset_sec": 3, "rgb_path": "assets/cat_12/v_-YMpwZkNc2A_3874_4188/0003s_00088_rgb.jpg", "mask_path": "assets/cat_12/v_-YMpwZkNc2A_3874_4188/0003s_00088_mask.png", "mask_status": "available" }, { "frame": "00120", "timestamp_sec": 133.25, "offset_sec": 4, "rgb_path": "assets/cat_12/v_-YMpwZkNc2A_3874_4188/0004s_00120_rgb.jpg", "mask_path": "assets/cat_12/v_-YMpwZkNc2A_3874_4188/0004s_00120_mask.png", "mask_status": "available" }, { "frame": "00148", "timestamp_sec": 134.25, "offset_sec": 5, "rgb_path": "assets/cat_12/v_-YMpwZkNc2A_3874_4188/0005s_00148_rgb.jpg", "mask_path": "assets/cat_12/v_-YMpwZkNc2A_3874_4188/0005s_00148_mask.png", "mask_status": "available" }, { "frame": "00180", "timestamp_sec": 135.25, "offset_sec": 6, "rgb_path": "assets/cat_12/v_-YMpwZkNc2A_3874_4188/0006s_00180_rgb.jpg", "mask_path": "assets/cat_12/v_-YMpwZkNc2A_3874_4188/0006s_00180_mask.png", "mask_status": "available" }, { "frame": "00208", "timestamp_sec": 136.25, "offset_sec": 7, "rgb_path": "assets/cat_12/v_-YMpwZkNc2A_3874_4188/0007s_00208_rgb.jpg", "mask_path": "assets/cat_12/v_-YMpwZkNc2A_3874_4188/0007s_00208_mask.png", "mask_status": "available" }, { "frame": "00240", "timestamp_sec": 137.25, "offset_sec": 8, "rgb_path": "assets/cat_12/v_-YMpwZkNc2A_3874_4188/0008s_00240_rgb.jpg", "mask_path": "assets/cat_12/v_-YMpwZkNc2A_3874_4188/0008s_00240_mask.png", "mask_status": "available" }, { "frame": "00268", "timestamp_sec": 138.25, "offset_sec": 9, "rgb_path": "assets/cat_12/v_-YMpwZkNc2A_3874_4188/0009s_00268_rgb.jpg", "mask_path": "assets/cat_12/v_-YMpwZkNc2A_3874_4188/0009s_00268_mask.png", "mask_status": "available" }, { "frame": "00300", "timestamp_sec": 139.25, "offset_sec": 10, "rgb_path": "assets/cat_12/v_-YMpwZkNc2A_3874_4188/0010s_00300_rgb.jpg", "mask_path": "assets/cat_12/v_-YMpwZkNc2A_3874_4188/0010s_00300_mask.png", "mask_status": "available" } ], "obj_id": [ 0 ], "anno_id": [ 87 ], "fps_inferred_from_clip_interval": 29.9048 }, "usage": { "input_tokens": 40227, "input_tokens_details": { "cache_write_tokens": 40224, "cached_tokens": 0 }, "output_tokens": 4518, "output_tokens_details": { "reasoning_tokens": 2387 }, "total_tokens": 44745 }, "model": "gpt-5.6-sol", "effort": "medium", "custom_id": "cat12_cap10_v_-YMpwZkNc2A_3874_4188", "annotation": { "caption": "The man lowers the towel out of view and runs his right hand through his hair. He then stands looking forward, turns his head slightly, and talks. He briefly touches his chin with his thumb while continuing to speak.", "entity_groups": [ { "id": "man_1", "label": "man", "spans": [ { "text": "The man", "range": [ 0, 7 ] }, { "text": "his", "range": [ 46, 49 ] }, { "text": "his", "range": [ 69, 72 ] }, { "text": "He", "range": [ 79, 81 ] }, { "text": "his", "range": [ 117, 120 ] }, { "text": "He", "range": [ 147, 149 ] }, { "text": "his", "range": [ 166, 169 ] }, { "text": "his", "range": [ 180, 183 ] } ], "source_mask_link": "selected_person" }, { "id": "towel_1", "label": "towel", "spans": [ { "text": "the towel", "range": [ 15, 24 ] } ], "source_mask_link": null }, { "id": "right_hand_1", "label": "right hand", "spans": [ { "text": "his right hand", "range": [ 46, 60 ] } ], "source_mask_link": null }, { "id": "hair_1", "label": "hair", "spans": [ { "text": "his hair", "range": [ 69, 77 ] } ], "source_mask_link": null }, { "id": "head_1", "label": "head", "spans": [ { "text": "his head", "range": [ 117, 125 ] } ], "source_mask_link": null }, { "id": "chin_1", "label": "chin", "spans": [ { "text": "his chin", "range": [ 166, 174 ] } ], "source_mask_link": null }, { "id": "thumb_1", "label": "thumb", "spans": [ { "text": "his thumb", "range": [ 180, 189 ] } ], "source_mask_link": null } ], "actions": [ { "id": "action_1", "mentions": [ { "text": "lowers", "ranges": [ [ 8, 14 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "type": "simple", "subject_refs": [ "man_1" ], "object_refs": [ "towel_1" ], "confidence": "high", "temporal_status": "unsupported_by_sampled_frames", "intervals": [], "evidence": "The towel is already out of view at the first sampled frame, so no sampled frame visibly supports the lowering process.", "mask_relation": "subject_matches_selected_person_mask", "source_subject_refs": "man_1" }, { "id": "action_2", "mentions": [ { "text": "runs", "ranges": [ [ 41, 45 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "type": "prepositional", "subject_refs": [ "man_1" ], "object_refs": [ "right_hand_1", "hair_1" ], "confidence": "high", "temporal_status": "supported", "intervals": [ { "start_s": 1.0, "end_s": 2.0, "source_start_s": 130.25, "source_end_s": 131.25 } ], "evidence": "At clip-relative 1.000 s (source 130.250 s), the masked man has his raised hand against and moving through his hair.", "mask_relation": "subject_matches_selected_person_mask", "source_subject_refs": "man_1" }, { "id": "action_3", "mentions": [ { "text": "stands", "ranges": [ [ 87, 93 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "type": "posture", "subject_refs": [ "man_1" ], "object_refs": [], "confidence": "high", "temporal_status": "uncertain", "intervals": [], "evidence": "The sampled frames crop out the man's lower body, so standing cannot be distinguished confidently from sitting.", "mask_relation": "subject_matches_selected_person_mask", "source_subject_refs": "man_1" }, { "id": "action_4", "mentions": [ { "text": "looking", "ranges": [ [ 94, 101 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "type": "simple", "subject_refs": [ "man_1" ], "object_refs": [], "confidence": "high", "temporal_status": "supported", "intervals": [ { "start_s": 2.0, "end_s": 4.0, "source_start_s": 131.25, "source_end_s": 133.25 } ], "evidence": "At clip-relative 2.000 and 3.000 s (source 131.250 and 132.250 s), the masked man is oriented and looking generally forward.", "mask_relation": "subject_matches_selected_person_mask", "source_subject_refs": "man_1" }, { "id": "action_5", "mentions": [ { "text": "turns", "ranges": [ [ 111, 116 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "type": "simple", "subject_refs": [ "man_1" ], "object_refs": [ "head_1" ], "confidence": "high", "temporal_status": "supported", "intervals": [ { "start_s": 3.0, "end_s": 5.0, "source_start_s": 132.25, "source_end_s": 134.25 } ], "evidence": "The masked man's head changes from a frontal orientation at 3.000 s to a pronounced side orientation at 4.000 s.", "mask_relation": "subject_matches_selected_person_mask", "source_subject_refs": "man_1" }, { "id": "action_6", "mentions": [ { "text": "talks", "ranges": [ [ 140, 145 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } }, { "text": "speak", "ranges": [ [ 210, 215 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "type": "simple", "subject_refs": [ "man_1" ], "object_refs": [], "confidence": "high", "temporal_status": "supported", "intervals": [ { "start_s": 6.0, "end_s": 10.5, "source_start_s": 135.25, "source_end_s": 139.75 } ], "evidence": "From clip-relative 6.000 s through the final sample at 10.000 s, the masked man's changing open-mouth configurations visibly support continuous speech; support reaches the clip boundary at 10.500 s.", "mask_relation": "subject_matches_selected_person_mask", "source_subject_refs": "man_1" }, { "id": "action_7", "mentions": [ { "text": "touches", "ranges": [ [ 158, 165 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "type": "simple", "subject_refs": [ "man_1" ], "object_refs": [ "chin_1" ], "confidence": "high", "temporal_status": "supported", "intervals": [ { "start_s": 4.0, "end_s": 5.0, "source_start_s": 133.25, "source_end_s": 134.25 } ], "evidence": "At clip-relative 4.000 s (source 133.250 s), the masked man visibly places his raised thumb against the underside of his chin.", "mask_relation": "subject_matches_selected_person_mask", "source_subject_refs": "man_1" } ], "notes": [ "The supplied masklet consistently identifies the single man, so all captioned actions have a mask-linked subject.", "Only the man has a supplied source mask. The towel and separately mentioned body parts are annotated as text entities but have null source-mask links.", "The original source interval is [129.250000, 139.750000), corresponding to the clip-relative interval [0.000000, 10.500000).", "The towel-lowering event is not temporally localized because the towel is absent from all sampled frames. Standing is also left unlocalized because the lower body is outside the frame." ] }, "model_output_raw": { "caption": "The man lowers the towel out of view and runs his right hand through his hair. He then stands looking forward, turns his head slightly, and talks. He briefly touches his chin with his thumb while continuing to speak.", "entity_groups": [ { "id": "man_1", "label": "man", "spans": [ { "text": "The man", "range": [ 0, 7 ] }, { "text": "his", "range": [ 46, 49 ] }, { "text": "his", "range": [ 69, 72 ] }, { "text": "He", "range": [ 79, 81 ] }, { "text": "his", "range": [ 117, 120 ] }, { "text": "He", "range": [ 147, 149 ] }, { "text": "his", "range": [ 166, 169 ] }, { "text": "his", "range": [ 180, 183 ] } ], "source_mask": "selected_person_masklet" }, { "id": "towel_1", "label": "towel", "spans": [ { "text": "the towel", "range": [ 15, 24 ] } ], "source_mask": null }, { "id": "right_hand_1", "label": "right hand", "spans": [ { "text": "his right hand", "range": [ 46, 60 ] } ], "source_mask": null }, { "id": "hair_1", "label": "hair", "spans": [ { "text": "his hair", "range": [ 69, 77 ] } ], "source_mask": null }, { "id": "head_1", "label": "head", "spans": [ { "text": "his head", "range": [ 117, 125 ] } ], "source_mask": null }, { "id": "chin_1", "label": "chin", "spans": [ { "text": "his chin", "range": [ 166, 174 ] } ], "source_mask": null }, { "id": "thumb_1", "label": "thumb", "spans": [ { "text": "his thumb", "range": [ 180, 189 ] } ], "source_mask": null } ], "actions": [ { "id": "action_1", "mentions": [ { "text": "lowers", "ranges": [ [ 8, 14 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "type": "simple", "subject_ref": "man_1", "object_refs": [ "towel_1" ], "confidence": "high", "mask_relation": "subject_matches_selected_person_mask", "temporal_status": "unsupported_by_sampled_frames", "intervals": [], "evidence": "The towel is already out of view at the first sampled frame, so no sampled frame visibly supports the lowering process." }, { "id": "action_2", "mentions": [ { "text": "runs", "ranges": [ [ 41, 45 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "type": "prepositional", "subject_ref": "man_1", "object_refs": [ "right_hand_1", "hair_1" ], "confidence": "high", "mask_relation": "subject_matches_selected_person_mask", "temporal_status": "supported", "intervals": [ { "start_s": 1.0, "end_s": 2.0 } ], "evidence": "At clip-relative 1.000 s (source 130.250 s), the masked man has his raised hand against and moving through his hair." }, { "id": "action_3", "mentions": [ { "text": "stands", "ranges": [ [ 87, 93 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "type": "posture", "subject_ref": "man_1", "object_refs": [], "confidence": "high", "mask_relation": "subject_matches_selected_person_mask", "temporal_status": "uncertain", "intervals": [], "evidence": "The sampled frames crop out the man's lower body, so standing cannot be distinguished confidently from sitting." }, { "id": "action_4", "mentions": [ { "text": "looking", "ranges": [ [ 94, 101 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "type": "simple", "subject_ref": "man_1", "object_refs": [], "confidence": "high", "mask_relation": "subject_matches_selected_person_mask", "temporal_status": "supported", "intervals": [ { "start_s": 2.0, "end_s": 4.0 } ], "evidence": "At clip-relative 2.000 and 3.000 s (source 131.250 and 132.250 s), the masked man is oriented and looking generally forward." }, { "id": "action_5", "mentions": [ { "text": "turns", "ranges": [ [ 111, 116 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "type": "simple", "subject_ref": "man_1", "object_refs": [ "head_1" ], "confidence": "high", "mask_relation": "subject_matches_selected_person_mask", "temporal_status": "supported", "intervals": [ { "start_s": 3.0, "end_s": 5.0 } ], "evidence": "The masked man's head changes from a frontal orientation at 3.000 s to a pronounced side orientation at 4.000 s." }, { "id": "action_6", "mentions": [ { "text": "talks", "ranges": [ [ 140, 145 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } }, { "text": "speak", "ranges": [ [ 210, 215 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "type": "simple", "subject_ref": "man_1", "object_refs": [], "confidence": "high", "mask_relation": "subject_matches_selected_person_mask", "temporal_status": "supported", "intervals": [ { "start_s": 6.0, "end_s": 10.5 } ], "evidence": "From clip-relative 6.000 s through the final sample at 10.000 s, the masked man's changing open-mouth configurations visibly support continuous speech; support reaches the clip boundary at 10.500 s." }, { "id": "action_7", "mentions": [ { "text": "touches", "ranges": [ [ 158, 165 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "type": "simple", "subject_ref": "man_1", "object_refs": [ "chin_1" ], "instrument_refs": [ "thumb_1" ], "confidence": "high", "mask_relation": "subject_matches_selected_person_mask", "temporal_status": "supported", "intervals": [ { "start_s": 4.0, "end_s": 5.0 } ], "evidence": "At clip-relative 4.000 s (source 133.250 s), the masked man visibly places his raised thumb against the underside of his chin." } ], "notes": [ "The supplied masklet consistently identifies the single man, so all captioned actions have a mask-linked subject.", "Only the man has a supplied source mask. The towel and separately mentioned body parts are annotated as text entities but have null source-mask links.", "The original source interval is [129.250000, 139.750000), corresponding to the clip-relative interval [0.000000, 10.500000).", "The towel-lowering event is not temporally localized because the towel is absent from all sampled frames. Standing is also left unlocalized because the lower body is outside the frame." ] }, "normalization_repairs": 0 }, { "source_video_id": "v_-YMpwZkNc2A", "caption_count_category": 12, "clip": { "caption_index": 11, "caption": "The man picks up a black razor in his right hand and holds it up while talking. He gestures with the razor before bringing it to his chin and nose. He moves the razor in a simulated shaving motion across his face.", "source_interval_sec": [ 139.75, 148.75 ], "loom_clip_id": "v_-YMpwZkNc2A_4188_4458", "clip_frame_interval_4fps": [ 4188, 4458 ], "sample_count_1fps": 9, "sampled_frames": [ { "frame": "00000", "timestamp_sec": 139.75, "offset_sec": 0, "rgb_path": "assets/cat_12/v_-YMpwZkNc2A_4188_4458/0000s_00000_rgb.jpg", "mask_path": "assets/cat_12/v_-YMpwZkNc2A_4188_4458/0000s_00000_mask.png", "mask_status": "available" }, { "frame": "00028", "timestamp_sec": 140.75, "offset_sec": 1, "rgb_path": "assets/cat_12/v_-YMpwZkNc2A_4188_4458/0001s_00028_rgb.jpg", "mask_path": "assets/cat_12/v_-YMpwZkNc2A_4188_4458/0001s_00028_mask.png", "mask_status": "available" }, { "frame": "00060", "timestamp_sec": 141.75, "offset_sec": 2, "rgb_path": "assets/cat_12/v_-YMpwZkNc2A_4188_4458/0002s_00060_rgb.jpg", "mask_path": "assets/cat_12/v_-YMpwZkNc2A_4188_4458/0002s_00060_mask.png", "mask_status": "available" }, { "frame": "00088", "timestamp_sec": 142.75, "offset_sec": 3, "rgb_path": "assets/cat_12/v_-YMpwZkNc2A_4188_4458/0003s_00088_rgb.jpg", "mask_path": "assets/cat_12/v_-YMpwZkNc2A_4188_4458/0003s_00088_mask.png", "mask_status": "available" }, { "frame": "00120", "timestamp_sec": 143.75, "offset_sec": 4, "rgb_path": "assets/cat_12/v_-YMpwZkNc2A_4188_4458/0004s_00120_rgb.jpg", "mask_path": "assets/cat_12/v_-YMpwZkNc2A_4188_4458/0004s_00120_mask.png", "mask_status": "available" }, { "frame": "00148", "timestamp_sec": 144.75, "offset_sec": 5, "rgb_path": "assets/cat_12/v_-YMpwZkNc2A_4188_4458/0005s_00148_rgb.jpg", "mask_path": "assets/cat_12/v_-YMpwZkNc2A_4188_4458/0005s_00148_mask.png", "mask_status": "available" }, { "frame": "00180", "timestamp_sec": 145.75, "offset_sec": 6, "rgb_path": "assets/cat_12/v_-YMpwZkNc2A_4188_4458/0006s_00180_rgb.jpg", "mask_path": "assets/cat_12/v_-YMpwZkNc2A_4188_4458/0006s_00180_mask.png", "mask_status": "available" }, { "frame": "00208", "timestamp_sec": 146.75, "offset_sec": 7, "rgb_path": "assets/cat_12/v_-YMpwZkNc2A_4188_4458/0007s_00208_rgb.jpg", "mask_path": "assets/cat_12/v_-YMpwZkNc2A_4188_4458/0007s_00208_mask.png", "mask_status": "available" }, { "frame": "00240", "timestamp_sec": 147.75, "offset_sec": 8, "rgb_path": "assets/cat_12/v_-YMpwZkNc2A_4188_4458/0008s_00240_rgb.jpg", "mask_path": "assets/cat_12/v_-YMpwZkNc2A_4188_4458/0008s_00240_mask.png", "mask_status": "available" } ], "obj_id": [ 0 ], "anno_id": [ 88 ], "fps_inferred_from_clip_interval": 30.0 }, "usage": { "input_tokens": 35702, "input_tokens_details": { "cache_write_tokens": 35699, "cached_tokens": 0 }, "output_tokens": 5037, "output_tokens_details": { "reasoning_tokens": 2836 }, "total_tokens": 40739 }, "model": "gpt-5.6-sol", "effort": "medium", "custom_id": "cat12_cap11_v_-YMpwZkNc2A_4188_4458", "annotation": { "caption": "The man picks up a black razor in his right hand and holds it up while talking. He gestures with the razor before bringing it to his chin and nose. He moves the razor in a simulated shaving motion across his face.", "entity_groups": [ { "id": "man", "label": "man", "spans": [ { "text": "The man", "range": [ 0, 7 ] }, { "text": "his", "range": [ 34, 37 ] }, { "text": "He", "range": [ 80, 82 ] }, { "text": "his", "range": [ 129, 132 ] }, { "text": "He", "range": [ 148, 150 ] }, { "text": "his", "range": [ 204, 207 ] } ], "source_mask_link": "selected_person" }, { "id": "razor", "label": "razor", "spans": [ { "text": "a black razor", "range": [ 17, 30 ] }, { "text": "it", "range": [ 59, 61 ] }, { "text": "the razor", "range": [ 97, 106 ] }, { "text": "it", "range": [ 123, 125 ] }, { "text": "the razor", "range": [ 157, 166 ] } ], "source_mask_link": null }, { "id": "right_hand", "label": "right hand", "spans": [ { "text": "his right hand", "range": [ 34, 48 ] } ], "source_mask_link": null }, { "id": "chin", "label": "chin", "spans": [ { "text": "his chin", "range": [ 129, 137 ] } ], "source_mask_link": null }, { "id": "nose", "label": "nose", "spans": [ { "text": "nose", "range": [ 142, 146 ] } ], "source_mask_link": null }, { "id": "face", "label": "face", "spans": [ { "text": "his face", "range": [ 204, 212 ] } ], "source_mask_link": null } ], "actions": [ { "id": "action_1", "mentions": [ { "text": "picks up", "ranges": [ [ 8, 16 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "type": "phrasal", "subject_refs": [ "man" ], "object_refs": [ "razor" ], "confidence": "medium", "temporal_status": "localized", "intervals": [ { "start_s": 0.0, "end_s": 2.0, "source_start_s": 139.75, "source_end_s": 141.75 } ], "evidence": "The razor changes from being low and mostly out of view at 0.0 seconds to being raised in the masked man's hand at 1.0 seconds.", "mask_relation": "subject_matches_selected_person_mask", "source_subject_refs": "man" }, { "id": "action_2", "mentions": [ { "text": "holds", "ranges": [ [ 53, 58 ] ], "occurrence": 1, "particle": { "text": "up", "occurrence": 2 } } ], "type": "phrasal", "subject_refs": [ "man" ], "object_refs": [ "razor" ], "confidence": "high", "temporal_status": "localized", "intervals": [ { "start_s": 1.0, "end_s": 9.0, "source_start_s": 140.75, "source_end_s": 148.75 } ], "evidence": "From 1.0 seconds through the final sampled frame, the masked man visibly keeps the razor raised in his hand.", "mask_relation": "subject_matches_selected_person_mask", "source_subject_refs": "man" }, { "id": "action_3", "mentions": [ { "text": "talking", "ranges": [ [ 71, 78 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "type": "simple", "subject_refs": [ "man" ], "object_refs": [], "confidence": "medium", "temporal_status": "localized", "intervals": [ { "start_s": 0.0, "end_s": 4.0, "source_start_s": 139.75, "source_end_s": 143.75 }, { "start_s": 5.0, "end_s": 8.0, "source_start_s": 144.75, "source_end_s": 147.75 } ], "evidence": "The masked man's mouth and facial expression vary across the supported frames in a manner consistent with speaking.", "mask_relation": "subject_matches_selected_person_mask", "source_subject_refs": "man" }, { "id": "action_4", "mentions": [ { "text": "gestures", "ranges": [ [ 83, 91 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "type": "prepositional", "subject_refs": [ "man" ], "object_refs": [ "razor" ], "confidence": "high", "temporal_status": "localized", "intervals": [ { "start_s": 1.0, "end_s": 2.0, "source_start_s": 140.75, "source_end_s": 141.75 }, { "start_s": 5.0, "end_s": 8.0, "source_start_s": 144.75, "source_end_s": 147.75 } ], "evidence": "The masked man displays and repositions the razor away from his face, with clear hand-and-razor pose changes at 1.0 and 5.0–7.0 seconds.", "mask_relation": "subject_matches_selected_person_mask", "source_subject_refs": "man" }, { "id": "action_5", "mentions": [ { "text": "bringing", "ranges": [ [ 114, 122 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "type": "simple", "subject_refs": [ "man" ], "object_refs": [ "razor", "chin", "nose" ], "confidence": "high", "temporal_status": "localized", "intervals": [ { "start_s": 1.0, "end_s": 3.0, "source_start_s": 140.75, "source_end_s": 142.75 }, { "start_s": 7.0, "end_s": 9.0, "source_start_s": 146.75, "source_end_s": 148.75 } ], "evidence": "The razor moves from a displayed position toward the masked man's mouth, nose, chin, and neck regions across these frame transitions.", "mask_relation": "subject_matches_selected_person_mask", "source_subject_refs": "man" }, { "id": "action_6", "mentions": [ { "text": "moves", "ranges": [ [ 151, 156 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "type": "simple", "subject_refs": [ "man" ], "object_refs": [ "razor", "face" ], "confidence": "high", "temporal_status": "localized", "intervals": [ { "start_s": 2.0, "end_s": 5.0, "source_start_s": 141.75, "source_end_s": 144.75 }, { "start_s": 7.0, "end_s": 9.0, "source_start_s": 146.75, "source_end_s": 148.75 } ], "evidence": "The masked man visibly moves the razor around his upper lip, chin, and neck in shaving-like passes.", "mask_relation": "subject_matches_selected_person_mask", "source_subject_refs": "man" } ], "notes": [ "The supplied mask consistently identifies the sole man, so every captioned action has a mask-linked subject.", "No separate masks were supplied for the razor, right hand, chin, nose, or face; their source-mask links remain null.", "The pickup interval has medium confidence because the sparse samples show the razor transitioning from a low position to a raised position but do not show the exact grasp instant.", "The phrase 'a simulated shaving motion' describes the manner of the razor-moving event and is not extracted as a separate light-verb action." ] }, "model_output_raw": { "caption": "The man picks up a black razor in his right hand and holds it up while talking. He gestures with the razor before bringing it to his chin and nose. He moves the razor in a simulated shaving motion across his face.", "entity_groups": [ { "id": "man", "label": "man", "spans": [ { "text": "The man", "range": [ 0, 7 ] }, { "text": "his", "range": [ 34, 37 ] }, { "text": "He", "range": [ 80, 82 ] }, { "text": "his", "range": [ 129, 132 ] }, { "text": "He", "range": [ 148, 150 ] }, { "text": "his", "range": [ 204, 207 ] } ], "source_mask": "selected_person" }, { "id": "razor", "label": "razor", "spans": [ { "text": "a black razor", "range": [ 17, 30 ] }, { "text": "it", "range": [ 59, 61 ] }, { "text": "the razor", "range": [ 97, 106 ] }, { "text": "it", "range": [ 123, 125 ] }, { "text": "the razor", "range": [ 157, 166 ] } ], "source_mask": null }, { "id": "right_hand", "label": "right hand", "spans": [ { "text": "his right hand", "range": [ 34, 48 ] } ], "source_mask": null }, { "id": "chin", "label": "chin", "spans": [ { "text": "his chin", "range": [ 129, 137 ] } ], "source_mask": null }, { "id": "nose", "label": "nose", "spans": [ { "text": "nose", "range": [ 142, 146 ] } ], "source_mask": null }, { "id": "face", "label": "face", "spans": [ { "text": "his face", "range": [ 204, 212 ] } ], "source_mask": null } ], "actions": [ { "id": "action_1", "mentions": [ { "text": "picks up", "ranges": [ [ 8, 16 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "type": "phrasal", "subject_ref": "man", "object_refs": [ "razor" ], "confidence": "medium", "temporal_status": "localized", "intervals": [ { "start_s": 0.0, "end_s": 2.0, "source_start_s": 139.75, "source_end_s": 141.75 } ], "evidence": "The razor changes from being low and mostly out of view at 0.0 seconds to being raised in the masked man's hand at 1.0 seconds.", "mask_relation": "subject_matches_selected_person_mask" }, { "id": "action_2", "mentions": [ { "text": "holds", "ranges": [ [ 53, 58 ] ], "occurrence": 1, "particle": { "text": "up", "occurrence": 2 } } ], "type": "phrasal", "subject_ref": "man", "object_refs": [ "razor" ], "confidence": "high", "temporal_status": "localized", "intervals": [ { "start_s": 1.0, "end_s": 9.0, "source_start_s": 140.75, "source_end_s": 148.75 } ], "evidence": "From 1.0 seconds through the final sampled frame, the masked man visibly keeps the razor raised in his hand.", "mask_relation": "subject_matches_selected_person_mask" }, { "id": "action_3", "mentions": [ { "text": "talking", "ranges": [ [ 71, 78 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "type": "simple", "subject_ref": "man", "object_refs": [], "confidence": "medium", "temporal_status": "localized", "intervals": [ { "start_s": 0.0, "end_s": 4.0, "source_start_s": 139.75, "source_end_s": 143.75 }, { "start_s": 5.0, "end_s": 8.0, "source_start_s": 144.75, "source_end_s": 147.75 } ], "evidence": "The masked man's mouth and facial expression vary across the supported frames in a manner consistent with speaking.", "mask_relation": "subject_matches_selected_person_mask" }, { "id": "action_4", "mentions": [ { "text": "gestures", "ranges": [ [ 83, 91 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "type": "prepositional", "subject_ref": "man", "object_refs": [ "razor" ], "confidence": "high", "temporal_status": "localized", "intervals": [ { "start_s": 1.0, "end_s": 2.0, "source_start_s": 140.75, "source_end_s": 141.75 }, { "start_s": 5.0, "end_s": 8.0, "source_start_s": 144.75, "source_end_s": 147.75 } ], "evidence": "The masked man displays and repositions the razor away from his face, with clear hand-and-razor pose changes at 1.0 and 5.0–7.0 seconds.", "mask_relation": "subject_matches_selected_person_mask" }, { "id": "action_5", "mentions": [ { "text": "bringing", "ranges": [ [ 114, 122 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "type": "simple", "subject_ref": "man", "object_refs": [ "razor", "chin", "nose" ], "confidence": "high", "temporal_status": "localized", "intervals": [ { "start_s": 1.0, "end_s": 3.0, "source_start_s": 140.75, "source_end_s": 142.75 }, { "start_s": 7.0, "end_s": 9.0, "source_start_s": 146.75, "source_end_s": 148.75 } ], "evidence": "The razor moves from a displayed position toward the masked man's mouth, nose, chin, and neck regions across these frame transitions.", "mask_relation": "subject_matches_selected_person_mask" }, { "id": "action_6", "mentions": [ { "text": "moves", "ranges": [ [ 151, 156 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "type": "simple", "subject_ref": "man", "object_refs": [ "razor", "face" ], "confidence": "high", "temporal_status": "localized", "intervals": [ { "start_s": 2.0, "end_s": 5.0, "source_start_s": 141.75, "source_end_s": 144.75 }, { "start_s": 7.0, "end_s": 9.0, "source_start_s": 146.75, "source_end_s": 148.75 } ], "evidence": "The masked man visibly moves the razor around his upper lip, chin, and neck in shaving-like passes.", "mask_relation": "subject_matches_selected_person_mask" } ], "notes": [ "The supplied mask consistently identifies the sole man, so every captioned action has a mask-linked subject.", "No separate masks were supplied for the razor, right hand, chin, nose, or face; their source-mask links remain null.", "The pickup interval has medium confidence because the sparse samples show the razor transitioning from a low position to a raised position but do not show the exact grasp instant.", "The phrase 'a simulated shaving motion' describes the manner of the razor-moving event and is not extracted as a separate light-verb action." ] }, "normalization_repairs": 0 }, { "source_video_id": "v_-YMpwZkNc2A", "caption_count_category": 12, "clip": { "caption_index": 12, "caption": "He lowers the razor, looks down briefly, and then looks up to continue speaking. After a few moments, he turns his head to the right and moves completely out of the frame.", "source_interval_sec": [ 148.75, 157.39 ], "loom_clip_id": "v_-YMpwZkNc2A_4458_4717", "clip_frame_interval_4fps": [ 4458, 4717 ], "sample_count_1fps": 9, "sampled_frames": [ { "frame": "00000", "timestamp_sec": 148.75, "offset_sec": 0, "rgb_path": "assets/cat_12/v_-YMpwZkNc2A_4458_4717/0000s_00000_rgb.jpg", "mask_path": "assets/cat_12/v_-YMpwZkNc2A_4458_4717/0000s_00000_mask.png", "mask_status": "available" }, { "frame": "00028", "timestamp_sec": 149.75, "offset_sec": 1, "rgb_path": "assets/cat_12/v_-YMpwZkNc2A_4458_4717/0001s_00028_rgb.jpg", "mask_path": "assets/cat_12/v_-YMpwZkNc2A_4458_4717/0001s_00028_mask.png", "mask_status": "available" }, { "frame": "00060", "timestamp_sec": 150.75, "offset_sec": 2, "rgb_path": "assets/cat_12/v_-YMpwZkNc2A_4458_4717/0002s_00060_rgb.jpg", "mask_path": "assets/cat_12/v_-YMpwZkNc2A_4458_4717/0002s_00060_mask.png", "mask_status": "available" }, { "frame": "00088", "timestamp_sec": 151.75, "offset_sec": 3, "rgb_path": "assets/cat_12/v_-YMpwZkNc2A_4458_4717/0003s_00088_rgb.jpg", "mask_path": "assets/cat_12/v_-YMpwZkNc2A_4458_4717/0003s_00088_mask.png", "mask_status": "available" }, { "frame": "00120", "timestamp_sec": 152.75, "offset_sec": 4, "rgb_path": "assets/cat_12/v_-YMpwZkNc2A_4458_4717/0004s_00120_rgb.jpg", "mask_path": "assets/cat_12/v_-YMpwZkNc2A_4458_4717/0004s_00120_mask.png", "mask_status": "available" }, { "frame": "00148", "timestamp_sec": 153.75, "offset_sec": 5, "rgb_path": "assets/cat_12/v_-YMpwZkNc2A_4458_4717/0005s_00148_rgb.jpg", "mask_path": "assets/cat_12/v_-YMpwZkNc2A_4458_4717/0005s_00148_mask.png", "mask_status": "available" }, { "frame": "00180", "timestamp_sec": 154.75, "offset_sec": 6, "rgb_path": "assets/cat_12/v_-YMpwZkNc2A_4458_4717/0006s_00180_rgb.jpg", "mask_path": "assets/cat_12/v_-YMpwZkNc2A_4458_4717/0006s_00180_mask.png", "mask_status": "available" }, { "frame": "00208", "timestamp_sec": 155.75, "offset_sec": 7, "rgb_path": "assets/cat_12/v_-YMpwZkNc2A_4458_4717/0007s_00208_rgb.jpg", "mask_path": "assets/cat_12/v_-YMpwZkNc2A_4458_4717/0007s_00208_mask.png", "mask_status": "available" }, { "frame": "00240", "timestamp_sec": 156.75, "offset_sec": 8, "rgb_path": "assets/cat_12/v_-YMpwZkNc2A_4458_4717/0008s_00240_rgb.jpg", "mask_path": "assets/cat_12/v_-YMpwZkNc2A_4458_4717/0008s_00240_mask.png", "mask_status": "available" } ], "obj_id": [ 0 ], "anno_id": [ 89 ], "fps_inferred_from_clip_interval": 29.9769 }, "usage": { "input_tokens": 35694, "input_tokens_details": { "cache_write_tokens": 35691, "cached_tokens": 0 }, "output_tokens": 4013, "output_tokens_details": { "reasoning_tokens": 2326 }, "total_tokens": 39707 }, "model": "gpt-5.6-sol", "effort": "medium", "custom_id": "cat12_cap12_v_-YMpwZkNc2A_4458_4717", "annotation": { "caption": "He lowers the razor, looks down briefly, and then looks up to continue speaking. After a few moments, he turns his head to the right and moves completely out of the frame.", "entity_groups": [ { "id": "entity_person_1", "label": "person", "spans": [ { "text": "He", "range": [ 0, 2 ] }, { "text": "he", "range": [ 102, 104 ] }, { "text": "his", "range": [ 111, 114 ] } ], "source_mask_link": "selected_person" }, { "id": "entity_razor_1", "label": "razor", "spans": [ { "text": "the razor", "range": [ 10, 19 ] } ], "source_mask_link": null }, { "id": "entity_head_1", "label": "head", "spans": [ { "text": "his head", "range": [ 111, 119 ] } ], "source_mask_link": null } ], "actions": [ { "id": "action_1", "mentions": [ { "text": "lowers", "ranges": [ [ 3, 9 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "type": "simple", "subject_refs": [ "entity_person_1" ], "object_refs": [ "entity_razor_1" ], "confidence": "high", "temporal_status": "supported_by_sampled_frames", "intervals": [ { "start_s": 0.0, "end_s": 2.0, "source_start_s": 148.75, "source_end_s": 150.75 } ], "evidence": "At 0.000 s the masked man holds the razor near his face; at 1.000 s it is visibly lower, and by 2.000 s it has left the visible area.", "mask_relation": "subject matches the supplied selected-person mask", "source_subject_refs": "entity_person_1" }, { "id": "action_2", "mentions": [ { "text": "looks", "ranges": [ [ 21, 26 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "type": "simple", "subject_refs": [ "entity_person_1" ], "object_refs": [], "confidence": "high", "temporal_status": "supported_by_sampled_frames", "intervals": [ { "start_s": 1.0, "end_s": 3.0, "source_start_s": 149.75, "source_end_s": 151.75 } ], "evidence": "The masked man's gaze and head are directed downward in the 1.000 s and 2.000 s frames.", "mask_relation": "subject matches the supplied selected-person mask", "source_subject_refs": "entity_person_1" }, { "id": "action_3", "mentions": [ { "text": "looks", "ranges": [ [ 50, 55 ] ], "occurrence": 2, "particle": { "text": null, "occurrence": 1 } } ], "type": "simple", "subject_refs": [ "entity_person_1" ], "object_refs": [], "confidence": "high", "temporal_status": "supported_by_sampled_frames", "intervals": [ { "start_s": 2.0, "end_s": 4.0, "source_start_s": 150.75, "source_end_s": 152.75 } ], "evidence": "The masked man changes from looking downward at 2.000 s to looking upward toward the camera at 3.000 s.", "mask_relation": "subject matches the supplied selected-person mask", "source_subject_refs": "entity_person_1" }, { "id": "action_4", "mentions": [ { "text": "speaking", "ranges": [ [ 71, 79 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "type": "simple", "subject_refs": [ "entity_person_1" ], "object_refs": [], "confidence": "medium", "temporal_status": "supported_by_sampled_frames", "intervals": [ { "start_s": 3.0, "end_s": 7.0, "source_start_s": 151.75, "source_end_s": 155.75 } ], "evidence": "From 3.000 s through 6.000 s, the masked man faces generally toward the camera with visible mouth and facial changes consistent with speaking.", "mask_relation": "subject matches the supplied selected-person mask", "source_subject_refs": "entity_person_1" }, { "id": "action_5", "mentions": [ { "text": "turns", "ranges": [ [ 105, 110 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "type": "simple", "subject_refs": [ "entity_person_1" ], "object_refs": [ "entity_head_1" ], "confidence": "medium", "temporal_status": "partially_observed_at_clip_boundary", "intervals": [ { "start_s": 8.0, "end_s": 8.639999999999986, "source_start_s": 156.75, "source_end_s": 157.39 } ], "evidence": "At 8.000 s the masked man's head is motion-blurred and oriented toward the right side of the scene, supporting an ongoing head turn.", "mask_relation": "subject matches the supplied selected-person mask", "source_subject_refs": "entity_person_1" }, { "id": "action_6", "mentions": [ { "text": "moves", "ranges": [ [ 137, 142 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "type": "simple", "subject_refs": [ "entity_person_1" ], "object_refs": [], "confidence": "medium", "temporal_status": "partially_observed_at_clip_boundary", "intervals": [ { "start_s": 8.0, "end_s": 8.639999999999986, "source_start_s": 156.75, "source_end_s": 157.39 } ], "evidence": "The 8.000 s frame shows strong motion blur and rightward displacement of the masked man. His complete exit is not visible in the sampled frames, but motion continues into the clip boundary.", "mask_relation": "subject matches the supplied selected-person mask", "source_subject_refs": "entity_person_1" } ], "notes": [ "The supplied masklet consistently identifies the sole captioned person.", "No separate source masks were supplied for the razor or the man's head.", "The temporal phrase \"a few moments,\" the spatial expression \"the right,\" and the image-medium reference \"the frame\" are not entity groups.", "Directional words \"down,\" \"up,\" and \"out\" and the modifiers \"briefly\" and \"completely\" are excluded from the action spans.", "\"continue\" is an aspectual/control predicate, so only \"speaking\" is selected for that action." ] }, "model_output_raw": { "caption": "He lowers the razor, looks down briefly, and then looks up to continue speaking. After a few moments, he turns his head to the right and moves completely out of the frame.", "entity_groups": [ { "id": "entity_person_1", "entity_type": "person", "spans": [ { "text": "He", "range": [ 0, 2 ] }, { "text": "he", "range": [ 102, 104 ] }, { "text": "his", "range": [ 111, 114 ] } ], "source_mask": "selected_person_masklet" }, { "id": "entity_razor_1", "entity_type": "razor", "spans": [ { "text": "the razor", "range": [ 10, 19 ] } ], "source_mask": null }, { "id": "entity_head_1", "entity_type": "head", "spans": [ { "text": "his head", "range": [ 111, 119 ] } ], "source_mask": null } ], "actions": [ { "id": "action_1", "mentions": [ { "text": "lowers", "ranges": [ [ 3, 9 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "type": "simple", "subject": "entity_person_1", "object": "entity_razor_1", "confidence": "high", "mask_relation": "subject matches the supplied selected-person mask", "temporal_status": "supported_by_sampled_frames", "intervals": [ { "clip_relative_s": [ 0.0, 2.0 ], "source_video_s": [ 148.75, 150.75 ] } ], "evidence": "At 0.000 s the masked man holds the razor near his face; at 1.000 s it is visibly lower, and by 2.000 s it has left the visible area." }, { "id": "action_2", "mentions": [ { "text": "looks", "ranges": [ [ 21, 26 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "type": "simple", "subject": "entity_person_1", "object": null, "confidence": "high", "mask_relation": "subject matches the supplied selected-person mask", "temporal_status": "supported_by_sampled_frames", "intervals": [ { "clip_relative_s": [ 1.0, 3.0 ], "source_video_s": [ 149.75, 151.75 ] } ], "evidence": "The masked man's gaze and head are directed downward in the 1.000 s and 2.000 s frames." }, { "id": "action_3", "mentions": [ { "text": "looks", "ranges": [ [ 50, 55 ] ], "occurrence": 2, "particle": { "text": null, "occurrence": 1 } } ], "type": "simple", "subject": "entity_person_1", "object": null, "confidence": "high", "mask_relation": "subject matches the supplied selected-person mask", "temporal_status": "supported_by_sampled_frames", "intervals": [ { "clip_relative_s": [ 2.0, 4.0 ], "source_video_s": [ 150.75, 152.75 ] } ], "evidence": "The masked man changes from looking downward at 2.000 s to looking upward toward the camera at 3.000 s." }, { "id": "action_4", "mentions": [ { "text": "speaking", "ranges": [ [ 71, 79 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "type": "simple", "subject": "entity_person_1", "object": null, "confidence": "medium", "mask_relation": "subject matches the supplied selected-person mask", "temporal_status": "supported_by_sampled_frames", "intervals": [ { "clip_relative_s": [ 3.0, 7.0 ], "source_video_s": [ 151.75, 155.75 ] } ], "evidence": "From 3.000 s through 6.000 s, the masked man faces generally toward the camera with visible mouth and facial changes consistent with speaking." }, { "id": "action_5", "mentions": [ { "text": "turns", "ranges": [ [ 105, 110 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "type": "simple", "subject": "entity_person_1", "object": "entity_head_1", "confidence": "medium", "mask_relation": "subject matches the supplied selected-person mask", "temporal_status": "partially_observed_at_clip_boundary", "intervals": [ { "clip_relative_s": [ 8.0, 8.64 ], "source_video_s": [ 156.75, 157.39 ] } ], "evidence": "At 8.000 s the masked man's head is motion-blurred and oriented toward the right side of the scene, supporting an ongoing head turn." }, { "id": "action_6", "mentions": [ { "text": "moves", "ranges": [ [ 137, 142 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "type": "simple", "subject": "entity_person_1", "object": null, "confidence": "medium", "mask_relation": "subject matches the supplied selected-person mask", "temporal_status": "partially_observed_at_clip_boundary", "intervals": [ { "clip_relative_s": [ 8.0, 8.64 ], "source_video_s": [ 156.75, 157.39 ] } ], "evidence": "The 8.000 s frame shows strong motion blur and rightward displacement of the masked man. His complete exit is not visible in the sampled frames, but motion continues into the clip boundary." } ], "notes": [ "The supplied masklet consistently identifies the sole captioned person.", "No separate source masks were supplied for the razor or the man's head.", "The temporal phrase \"a few moments,\" the spatial expression \"the right,\" and the image-medium reference \"the frame\" are not entity groups.", "Directional words \"down,\" \"up,\" and \"out\" and the modifiers \"briefly\" and \"completely\" are excluded from the action spans.", "\"continue\" is an aspectual/control predicate, so only \"speaking\" is selected for that action." ] }, "normalization_repairs": 0 }, { "source_video_id": "v_-Sp-gOEGy2o", "caption_count_category": 13, "clip": { "caption_index": 1, "caption": "A man with glasses and a mustache, wearing a green shirt and jeans, kneels while looking forward and speaking. He rests his left elbow on his knee and keeps his right hand on his other leg.", "source_interval_sec": [ 6.41, 12.91 ], "loom_clip_id": "v_-Sp-gOEGy2o_192_387", "clip_frame_interval_4fps": [ 192, 387 ], "sample_count_1fps": 7, "sampled_frames": [ { "frame": "00000", "timestamp_sec": 6.41, "offset_sec": 0, "rgb_path": "assets/cat_13/v_-Sp-gOEGy2o_192_387/0000s_00000_rgb.jpg", "mask_path": "assets/cat_13/v_-Sp-gOEGy2o_192_387/0000s_00000_mask.png", "mask_status": "available" }, { "frame": "00028", "timestamp_sec": 7.41, "offset_sec": 1, "rgb_path": "assets/cat_13/v_-Sp-gOEGy2o_192_387/0001s_00028_rgb.jpg", "mask_path": "assets/cat_13/v_-Sp-gOEGy2o_192_387/0001s_00028_mask.png", "mask_status": "available" }, { "frame": "00060", "timestamp_sec": 8.41, "offset_sec": 2, "rgb_path": "assets/cat_13/v_-Sp-gOEGy2o_192_387/0002s_00060_rgb.jpg", "mask_path": "assets/cat_13/v_-Sp-gOEGy2o_192_387/0002s_00060_mask.png", "mask_status": "available" }, { "frame": "00088", "timestamp_sec": 9.41, "offset_sec": 3, "rgb_path": "assets/cat_13/v_-Sp-gOEGy2o_192_387/0003s_00088_rgb.jpg", "mask_path": "assets/cat_13/v_-Sp-gOEGy2o_192_387/0003s_00088_mask.png", "mask_status": "available" }, { "frame": "00120", "timestamp_sec": 10.41, "offset_sec": 4, "rgb_path": "assets/cat_13/v_-Sp-gOEGy2o_192_387/0004s_00120_rgb.jpg", "mask_path": "assets/cat_13/v_-Sp-gOEGy2o_192_387/0004s_00120_mask.png", "mask_status": "available" }, { "frame": "00148", "timestamp_sec": 11.41, "offset_sec": 5, "rgb_path": "assets/cat_13/v_-Sp-gOEGy2o_192_387/0005s_00148_rgb.jpg", "mask_path": "assets/cat_13/v_-Sp-gOEGy2o_192_387/0005s_00148_mask.png", "mask_status": "available" }, { "frame": "00180", "timestamp_sec": 12.41, "offset_sec": 6, "rgb_path": "assets/cat_13/v_-Sp-gOEGy2o_192_387/0006s_00180_rgb.jpg", "mask_path": "assets/cat_13/v_-Sp-gOEGy2o_192_387/0006s_00180_mask.png", "mask_status": "available" } ], "obj_id": [ 0 ], "anno_id": [ 70 ], "fps_inferred_from_clip_interval": 30.0 }, "usage": { "input_tokens": 31173, "input_tokens_details": { "cache_write_tokens": 31170, "cached_tokens": 0 }, "output_tokens": 5222, "output_tokens_details": { "reasoning_tokens": 3095 }, "total_tokens": 36395 }, "model": "gpt-5.6-sol", "effort": "medium", "custom_id": "cat13_cap01_v_-Sp-gOEGy2o_192_387", "annotation": { "caption": "A man with glasses and a mustache, wearing a green shirt and jeans, kneels while looking forward and speaking. He rests his left elbow on his knee and keeps his right hand on his other leg.", "entity_groups": [ { "id": "person_1", "label": "man", "spans": [ { "text": "A man", "range": [ 0, 5 ] }, { "text": "He", "range": [ 111, 113 ] }, { "text": "his", "range": [ 120, 123 ] }, { "text": "his", "range": [ 138, 141 ] }, { "text": "his", "range": [ 157, 160 ] }, { "text": "his", "range": [ 175, 178 ] } ], "source_mask_link": "selected_person" }, { "id": "glasses_1", "label": "glasses", "spans": [ { "text": "glasses", "range": [ 11, 18 ] } ], "source_mask_link": null }, { "id": "mustache_1", "label": "mustache", "spans": [ { "text": "a mustache", "range": [ 23, 33 ] } ], "source_mask_link": null }, { "id": "shirt_1", "label": "shirt", "spans": [ { "text": "a green shirt", "range": [ 43, 56 ] } ], "source_mask_link": null }, { "id": "jeans_1", "label": "jeans", "spans": [ { "text": "jeans", "range": [ 61, 66 ] } ], "source_mask_link": null }, { "id": "elbow_1", "label": "left elbow", "spans": [ { "text": "his left elbow", "range": [ 120, 134 ] } ], "source_mask_link": null }, { "id": "knee_1", "label": "knee", "spans": [ { "text": "his knee", "range": [ 138, 146 ] } ], "source_mask_link": null }, { "id": "hand_1", "label": "right hand", "spans": [ { "text": "his right hand", "range": [ 157, 171 ] } ], "source_mask_link": null }, { "id": "leg_1", "label": "other leg", "spans": [ { "text": "his other leg", "range": [ 175, 188 ] } ], "source_mask_link": null } ], "actions": [ { "id": "action_1", "mentions": [ { "text": "wearing", "ranges": [ [ 35, 42 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "type": "simple", "subject_refs": [ "person_1" ], "object_refs": [ "shirt_1", "jeans_1" ], "confidence": "high", "temporal_status": "localized", "intervals": [ { "start_s": 1.0, "end_s": 6.5, "source_start_s": 7.41, "source_end_s": 12.91 } ], "evidence": "The masked man visibly wears a green shirt and jeans in all non-black sampled frames from 1.0 seconds through the clip boundary.", "mask_relation": "subject matches the supplied selected-person mask", "source_subject_refs": "person_1" }, { "id": "action_2", "mentions": [ { "text": "kneels", "ranges": [ [ 68, 74 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "type": "posture", "subject_refs": [ "person_1" ], "object_refs": [], "confidence": "high", "temporal_status": "localized", "intervals": [ { "start_s": 1.0, "end_s": 6.5, "source_start_s": 7.41, "source_end_s": 12.91 } ], "evidence": "The masked man remains low to the ground with one knee raised throughout the visible sampled sequence.", "mask_relation": "subject matches the supplied selected-person mask", "source_subject_refs": "person_1" }, { "id": "action_3", "mentions": [ { "text": "looking", "ranges": [ [ 81, 88 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "type": "simple", "subject_refs": [ "person_1" ], "object_refs": [], "confidence": "medium", "temporal_status": "localized", "intervals": [ { "start_s": 3.0, "end_s": 6.5, "source_start_s": 9.41, "source_end_s": 12.91 } ], "evidence": "From the 3.0-second sample onward, the masked man's head and gaze are directed generally forward; earlier samples show him looking downward.", "mask_relation": "subject matches the supplied selected-person mask", "source_subject_refs": "person_1" }, { "id": "action_4", "mentions": [ { "text": "speaking", "ranges": [ [ 101, 109 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "type": "simple", "subject_refs": [ "person_1" ], "object_refs": [], "confidence": "high", "temporal_status": "localized", "intervals": [ { "start_s": 1.0, "end_s": 5.0, "source_start_s": 7.41, "source_end_s": 11.41 }, { "start_s": 6.0, "end_s": 6.5, "source_start_s": 12.41, "source_end_s": 12.91 } ], "evidence": "The masked man's mouth is visibly open or changing as if speaking in the 1.0–4.0-second samples and again at 6.0 seconds; the 5.0-second frame is less supportive.", "mask_relation": "subject matches the supplied selected-person mask", "source_subject_refs": "person_1" }, { "id": "action_5", "mentions": [ { "text": "rests", "ranges": [ [ 114, 119 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "type": "simple", "subject_refs": [ "person_1" ], "object_refs": [ "elbow_1", "knee_1" ], "confidence": "medium", "temporal_status": "localized", "intervals": [ { "start_s": 1.0, "end_s": 6.5, "source_start_s": 7.41, "source_end_s": 12.91 } ], "evidence": "An elbow and forearm of the masked man remain supported on his raised knee, although anatomical left/right laterality is difficult to verify from the frontal view.", "mask_relation": "subject matches the supplied selected-person mask", "source_subject_refs": "person_1" }, { "id": "action_6", "mentions": [ { "text": "keeps", "ranges": [ [ 151, 156 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "type": "simple", "subject_refs": [ "person_1" ], "object_refs": [ "hand_1", "leg_1" ], "confidence": "low", "temporal_status": "localized", "intervals": [ { "start_s": 1.0, "end_s": 2.0, "source_start_s": 7.41, "source_end_s": 8.41 } ], "evidence": "A hand appears near or on the other leg at 1.0 seconds, but the relevant arm extends outward in subsequent frames and later hand placement is partly obscured by the lower-third graphic.", "mask_relation": "subject matches the supplied selected-person mask", "source_subject_refs": "person_1" } ], "notes": [ "The sample at clip-relative 0.0 seconds is black and its paired mask is empty, so it provides no visual action support.", "The supplied mask identifies only the man and is nonempty from 1.0 through 6.0 seconds.", "No separate source masks were supplied for the glasses, mustache, clothing, elbow, knee, hand, or leg.", "Left/right body-part identity is uncertain because of the frontal viewpoint, and the lower-third graphic obscures parts of the man's legs and hands in later frames." ] }, "model_output_raw": { "caption": "A man with glasses and a mustache, wearing a green shirt and jeans, kneels while looking forward and speaking. He rests his left elbow on his knee and keeps his right hand on his other leg.", "entity_groups": [ { "id": "person_1", "label": "man", "spans": [ { "text": "A man", "range": [ 0, 5 ] }, { "text": "He", "range": [ 111, 113 ] }, { "text": "his", "range": [ 120, 123 ] }, { "text": "his", "range": [ 138, 141 ] }, { "text": "his", "range": [ 157, 160 ] }, { "text": "his", "range": [ 175, 178 ] } ], "source_mask": "selected_person_mask" }, { "id": "glasses_1", "label": "glasses", "spans": [ { "text": "glasses", "range": [ 11, 18 ] } ], "source_mask": null }, { "id": "mustache_1", "label": "mustache", "spans": [ { "text": "a mustache", "range": [ 23, 33 ] } ], "source_mask": null }, { "id": "shirt_1", "label": "shirt", "spans": [ { "text": "a green shirt", "range": [ 43, 56 ] } ], "source_mask": null }, { "id": "jeans_1", "label": "jeans", "spans": [ { "text": "jeans", "range": [ 61, 66 ] } ], "source_mask": null }, { "id": "elbow_1", "label": "left elbow", "spans": [ { "text": "his left elbow", "range": [ 120, 134 ] } ], "source_mask": null }, { "id": "knee_1", "label": "knee", "spans": [ { "text": "his knee", "range": [ 138, 146 ] } ], "source_mask": null }, { "id": "hand_1", "label": "right hand", "spans": [ { "text": "his right hand", "range": [ 157, 171 ] } ], "source_mask": null }, { "id": "leg_1", "label": "other leg", "spans": [ { "text": "his other leg", "range": [ 175, 188 ] } ], "source_mask": null } ], "actions": [ { "id": "action_1", "mentions": [ { "text": "wearing", "ranges": [ [ 35, 42 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "type": "simple", "subject_ref": "person_1", "object_refs": [ "shirt_1", "jeans_1" ], "confidence": "high", "mask_relation": "subject matches the supplied selected-person mask", "temporal_status": "localized", "intervals": [ { "clip_relative_s": [ 1.0, 6.5 ], "source_video_s": [ 7.41, 12.91 ] } ], "evidence": "The masked man visibly wears a green shirt and jeans in all non-black sampled frames from 1.0 seconds through the clip boundary." }, { "id": "action_2", "mentions": [ { "text": "kneels", "ranges": [ [ 68, 74 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "type": "posture", "subject_ref": "person_1", "object_refs": [], "confidence": "high", "mask_relation": "subject matches the supplied selected-person mask", "temporal_status": "localized", "intervals": [ { "clip_relative_s": [ 1.0, 6.5 ], "source_video_s": [ 7.41, 12.91 ] } ], "evidence": "The masked man remains low to the ground with one knee raised throughout the visible sampled sequence." }, { "id": "action_3", "mentions": [ { "text": "looking", "ranges": [ [ 81, 88 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "type": "simple", "subject_ref": "person_1", "object_refs": [], "confidence": "medium", "mask_relation": "subject matches the supplied selected-person mask", "temporal_status": "localized", "intervals": [ { "clip_relative_s": [ 3.0, 6.5 ], "source_video_s": [ 9.41, 12.91 ] } ], "evidence": "From the 3.0-second sample onward, the masked man's head and gaze are directed generally forward; earlier samples show him looking downward." }, { "id": "action_4", "mentions": [ { "text": "speaking", "ranges": [ [ 101, 109 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "type": "simple", "subject_ref": "person_1", "object_refs": [], "confidence": "high", "mask_relation": "subject matches the supplied selected-person mask", "temporal_status": "localized", "intervals": [ { "clip_relative_s": [ 1.0, 5.0 ], "source_video_s": [ 7.41, 11.41 ] }, { "clip_relative_s": [ 6.0, 6.5 ], "source_video_s": [ 12.41, 12.91 ] } ], "evidence": "The masked man's mouth is visibly open or changing as if speaking in the 1.0–4.0-second samples and again at 6.0 seconds; the 5.0-second frame is less supportive." }, { "id": "action_5", "mentions": [ { "text": "rests", "ranges": [ [ 114, 119 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "type": "simple", "subject_ref": "person_1", "object_refs": [ "elbow_1", "knee_1" ], "confidence": "medium", "mask_relation": "subject matches the supplied selected-person mask", "temporal_status": "localized", "intervals": [ { "clip_relative_s": [ 1.0, 6.5 ], "source_video_s": [ 7.41, 12.91 ] } ], "evidence": "An elbow and forearm of the masked man remain supported on his raised knee, although anatomical left/right laterality is difficult to verify from the frontal view." }, { "id": "action_6", "mentions": [ { "text": "keeps", "ranges": [ [ 151, 156 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "type": "simple", "subject_ref": "person_1", "object_refs": [ "hand_1", "leg_1" ], "confidence": "low", "mask_relation": "subject matches the supplied selected-person mask", "temporal_status": "localized", "intervals": [ { "clip_relative_s": [ 1.0, 2.0 ], "source_video_s": [ 7.41, 8.41 ] } ], "evidence": "A hand appears near or on the other leg at 1.0 seconds, but the relevant arm extends outward in subsequent frames and later hand placement is partly obscured by the lower-third graphic." } ], "notes": [ "The sample at clip-relative 0.0 seconds is black and its paired mask is empty, so it provides no visual action support.", "The supplied mask identifies only the man and is nonempty from 1.0 through 6.0 seconds.", "No separate source masks were supplied for the glasses, mustache, clothing, elbow, knee, hand, or leg.", "Left/right body-part identity is uncertain because of the frontal viewpoint, and the lower-third graphic obscures parts of the man's legs and hands in later frames." ] }, "normalization_repairs": 0 }, { "source_video_id": "v_-Sp-gOEGy2o", "caption_count_category": 13, "clip": { "caption_index": 2, "caption": "The man leans forward, looking down to his right. He reaches out with his right hand, picks up a white box of matches from a stone ledge, and prepares to strike a match.", "source_interval_sec": [ 12.91, 20.41 ], "loom_clip_id": "v_-Sp-gOEGy2o_387_612", "clip_frame_interval_4fps": [ 387, 612 ], "sample_count_1fps": 8, "sampled_frames": [ { "frame": "00000", "timestamp_sec": 12.91, "offset_sec": 0, "rgb_path": "assets/cat_13/v_-Sp-gOEGy2o_387_612/0000s_00000_rgb.jpg", "mask_path": "assets/cat_13/v_-Sp-gOEGy2o_387_612/0000s_00000_mask.png", "mask_status": "available" }, { "frame": "00028", "timestamp_sec": 13.91, "offset_sec": 1, "rgb_path": "assets/cat_13/v_-Sp-gOEGy2o_387_612/0001s_00028_rgb.jpg", "mask_path": "assets/cat_13/v_-Sp-gOEGy2o_387_612/0001s_00028_mask.png", "mask_status": "available" }, { "frame": "00060", "timestamp_sec": 14.91, "offset_sec": 2, "rgb_path": "assets/cat_13/v_-Sp-gOEGy2o_387_612/0002s_00060_rgb.jpg", "mask_path": "assets/cat_13/v_-Sp-gOEGy2o_387_612/0002s_00060_mask.png", "mask_status": "available" }, { "frame": "00088", "timestamp_sec": 15.91, "offset_sec": 3, "rgb_path": "assets/cat_13/v_-Sp-gOEGy2o_387_612/0003s_00088_rgb.jpg", "mask_path": "assets/cat_13/v_-Sp-gOEGy2o_387_612/0003s_00088_mask.png", "mask_status": "available" }, { "frame": "00120", "timestamp_sec": 16.91, "offset_sec": 4, "rgb_path": "assets/cat_13/v_-Sp-gOEGy2o_387_612/0004s_00120_rgb.jpg", "mask_path": "assets/cat_13/v_-Sp-gOEGy2o_387_612/0004s_00120_mask.png", "mask_status": "available" }, { "frame": "00148", "timestamp_sec": 17.91, "offset_sec": 5, "rgb_path": "assets/cat_13/v_-Sp-gOEGy2o_387_612/0005s_00148_rgb.jpg", "mask_path": "assets/cat_13/v_-Sp-gOEGy2o_387_612/0005s_00148_mask.png", "mask_status": "available" }, { "frame": "00180", "timestamp_sec": 18.91, "offset_sec": 6, "rgb_path": "assets/cat_13/v_-Sp-gOEGy2o_387_612/0006s_00180_rgb.jpg", "mask_path": "assets/cat_13/v_-Sp-gOEGy2o_387_612/0006s_00180_mask.png", "mask_status": "available" }, { "frame": "00208", "timestamp_sec": 19.91, "offset_sec": 7, "rgb_path": "assets/cat_13/v_-Sp-gOEGy2o_387_612/0007s_00208_rgb.jpg", "mask_path": "assets/cat_13/v_-Sp-gOEGy2o_387_612/0007s_00208_mask.png", "mask_status": "available" } ], "obj_id": [ 0 ], "anno_id": [ 74 ], "fps_inferred_from_clip_interval": 30.0 }, "usage": { "input_tokens": 33434, "input_tokens_details": { "cache_write_tokens": 33431, "cached_tokens": 0 }, "output_tokens": 4139, "output_tokens_details": { "reasoning_tokens": 2588 }, "total_tokens": 37573 }, "model": "gpt-5.6-sol", "effort": "medium", "custom_id": "cat13_cap02_v_-Sp-gOEGy2o_387_612", "annotation": { "caption": "The man leans forward, looking down to his right. He reaches out with his right hand, picks up a white box of matches from a stone ledge, and prepares to strike a match.", "entity_groups": [ { "id": "entity_1", "label": "man", "spans": [ { "text": "The man", "range": [ 0, 7 ] }, { "text": "his", "range": [ 39, 42 ] }, { "text": "He", "range": [ 50, 52 ] }, { "text": "his", "range": [ 70, 73 ] } ], "source_mask_link": "selected_person" }, { "id": "entity_2", "label": "right hand", "spans": [ { "text": "his right hand", "range": [ 70, 84 ] } ], "source_mask_link": null }, { "id": "entity_3", "label": "box", "spans": [ { "text": "a white box of matches", "range": [ 95, 117 ] } ], "source_mask_link": null }, { "id": "entity_4", "label": "matches", "spans": [ { "text": "a white box of matches", "range": [ 95, 117 ] } ], "source_mask_link": null }, { "id": "entity_5", "label": "stone ledge", "spans": [ { "text": "a stone ledge", "range": [ 123, 136 ] } ], "source_mask_link": null }, { "id": "entity_6", "label": "match", "spans": [ { "text": "a match", "range": [ 161, 168 ] } ], "source_mask_link": null } ], "actions": [ { "id": "action_1", "mentions": [ { "text": "leans", "ranges": [ [ 8, 13 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "type": "posture", "subject_refs": [ "entity_1" ], "object_refs": [], "confidence": "high", "temporal_status": "supported", "intervals": [ { "start_s": 1.0, "end_s": 7.5, "source_start_s": 13.91, "source_end_s": 20.41 } ], "evidence": "From the 1.000-second frame through the final sample, the masked man holds his torso forward over the ledge.", "mask_relation": "subject_matches_supplied_selected_person_mask", "source_subject_refs": "entity_1" }, { "id": "action_2", "mentions": [ { "text": "looking", "ranges": [ [ 23, 30 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "type": "prepositional", "subject_refs": [ "entity_1" ], "object_refs": [], "confidence": "high", "temporal_status": "supported", "intervals": [ { "start_s": 1.0, "end_s": 7.5, "source_start_s": 13.91, "source_end_s": 20.41 } ], "evidence": "At 1.000–7.000 seconds, the masked man's head and gaze are directed downward toward the work area on his right.", "mask_relation": "subject_matches_supplied_selected_person_mask", "source_subject_refs": "entity_1" }, { "id": "action_3", "mentions": [ { "text": "reaches out", "ranges": [ [ 53, 64 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "type": "phrasal", "subject_refs": [ "entity_1" ], "object_refs": [], "confidence": "high", "temporal_status": "supported", "intervals": [ { "start_s": 1.0, "end_s": 3.0, "source_start_s": 13.91, "source_end_s": 15.91 } ], "evidence": "The clean frames at 1.000 and 2.000 seconds show the masked man extending his right arm downward toward the ledge.", "mask_relation": "subject_matches_supplied_selected_person_mask", "source_subject_refs": "entity_1" }, { "id": "action_4", "mentions": [ { "text": "picks up", "ranges": [ [ 86, 94 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "type": "phrasal", "subject_refs": [ "entity_1" ], "object_refs": [ "entity_3", "entity_4" ], "confidence": "high", "temporal_status": "supported", "intervals": [ { "start_s": 6.0, "end_s": 7.5, "source_start_s": 18.91, "source_end_s": 20.41 } ], "evidence": "At 6.000 seconds the masked man grips the white match box, and by 7.000 seconds he has lifted and tilted it from its resting position.", "mask_relation": "subject_matches_supplied_selected_person_mask", "source_subject_refs": "entity_1" }, { "id": "action_5", "mentions": [ { "text": "strike", "ranges": [ [ 154, 160 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "type": "simple", "subject_refs": [ "entity_1" ], "object_refs": [ "entity_6" ], "confidence": "high", "temporal_status": "uncertain", "intervals": [], "evidence": "The final clean frame shows the masked man manipulating the match box, but no sampled frame clearly shows a match being struck.", "mask_relation": "subject_matches_supplied_selected_person_mask", "source_subject_refs": "entity_1" } ], "notes": [ "The supplied masklet consistently identifies the man, so only actions with the man as subject are included.", "No separate source masks were supplied for the hand, box, matches, ledge, or individual match.", "The phrase \"prepares\" is excluded as a preparation predicate; its embedded physical action \"strike\" is retained but has no visually supported sampled-frame interval.", "The source-video interval is [12.910000, 20.410000), corresponding to the clip-relative interval [0.000000, 7.500000)." ] }, "model_output_raw": { "caption": "The man leans forward, looking down to his right. He reaches out with his right hand, picks up a white box of matches from a stone ledge, and prepares to strike a match.", "entity_groups": [ { "id": "entity_1", "label": "man", "spans": [ { "text": "The man", "range": [ 0, 7 ] }, { "text": "his", "range": [ 39, 42 ] }, { "text": "He", "range": [ 50, 52 ] }, { "text": "his", "range": [ 70, 73 ] } ], "source_mask": "selected_person_masklet" }, { "id": "entity_2", "label": "right hand", "spans": [ { "text": "his right hand", "range": [ 70, 84 ] } ], "source_mask": null }, { "id": "entity_3", "label": "box", "spans": [ { "text": "a white box of matches", "range": [ 95, 117 ] } ], "source_mask": null }, { "id": "entity_4", "label": "matches", "spans": [ { "text": "a white box of matches", "range": [ 95, 117 ] } ], "source_mask": null }, { "id": "entity_5", "label": "stone ledge", "spans": [ { "text": "a stone ledge", "range": [ 123, 136 ] } ], "source_mask": null }, { "id": "entity_6", "label": "match", "spans": [ { "text": "a match", "range": [ 161, 168 ] } ], "source_mask": null } ], "actions": [ { "mentions": [ { "text": "leans", "ranges": [ [ 8, 13 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "type": "posture", "subject_entity_group": "entity_1", "object_entity_groups": [], "confidence": "high", "mask_relation": "subject_matches_supplied_selected_person_mask", "temporal_status": "supported", "intervals": [ { "start_s": 1.0, "end_s": 7.5 } ], "evidence": "From the 1.000-second frame through the final sample, the masked man holds his torso forward over the ledge." }, { "mentions": [ { "text": "looking", "ranges": [ [ 23, 30 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "type": "prepositional", "subject_entity_group": "entity_1", "object_entity_groups": [], "confidence": "high", "mask_relation": "subject_matches_supplied_selected_person_mask", "temporal_status": "supported", "intervals": [ { "start_s": 1.0, "end_s": 7.5 } ], "evidence": "At 1.000–7.000 seconds, the masked man's head and gaze are directed downward toward the work area on his right." }, { "mentions": [ { "text": "reaches out", "ranges": [ [ 53, 64 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "type": "phrasal", "subject_entity_group": "entity_1", "object_entity_groups": [], "confidence": "high", "mask_relation": "subject_matches_supplied_selected_person_mask", "temporal_status": "supported", "intervals": [ { "start_s": 1.0, "end_s": 3.0 } ], "evidence": "The clean frames at 1.000 and 2.000 seconds show the masked man extending his right arm downward toward the ledge." }, { "mentions": [ { "text": "picks up", "ranges": [ [ 86, 94 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "type": "phrasal", "subject_entity_group": "entity_1", "object_entity_groups": [ "entity_3", "entity_4" ], "confidence": "high", "mask_relation": "subject_matches_supplied_selected_person_mask", "temporal_status": "supported", "intervals": [ { "start_s": 6.0, "end_s": 7.5 } ], "evidence": "At 6.000 seconds the masked man grips the white match box, and by 7.000 seconds he has lifted and tilted it from its resting position." }, { "mentions": [ { "text": "strike", "ranges": [ [ 154, 160 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "type": "simple", "subject_entity_group": "entity_1", "object_entity_groups": [ "entity_6" ], "confidence": "high", "mask_relation": "subject_matches_supplied_selected_person_mask", "temporal_status": "uncertain", "intervals": [], "evidence": "The final clean frame shows the masked man manipulating the match box, but no sampled frame clearly shows a match being struck." } ], "notes": [ "The supplied masklet consistently identifies the man, so only actions with the man as subject are included.", "No separate source masks were supplied for the hand, box, matches, ledge, or individual match.", "The phrase \"prepares\" is excluded as a preparation predicate; its embedded physical action \"strike\" is retained but has no visually supported sampled-frame interval.", "The source-video interval is [12.910000, 20.410000), corresponding to the clip-relative interval [0.000000, 7.500000)." ] }, "normalization_repairs": 0 }, { "source_video_id": "v_-Sp-gOEGy2o", "caption_count_category": 13, "clip": { "caption_index": 3, "caption": "After striking a match with his right hand, he transfers the lit match to his left hand. He holds the burning match while looking down at a pile of kindling in a fire pit.", "source_interval_sec": [ 20.41, 27.41 ], "loom_clip_id": "v_-Sp-gOEGy2o_612_821", "clip_frame_interval_4fps": [ 612, 821 ], "sample_count_1fps": 7, "sampled_frames": [ { "frame": "00000", "timestamp_sec": 20.41, "offset_sec": 0, "rgb_path": "assets/cat_13/v_-Sp-gOEGy2o_612_821/0000s_00000_rgb.jpg", "mask_path": "assets/cat_13/v_-Sp-gOEGy2o_612_821/0000s_00000_mask.png", "mask_status": "available" }, { "frame": "00028", "timestamp_sec": 21.41, "offset_sec": 1, "rgb_path": "assets/cat_13/v_-Sp-gOEGy2o_612_821/0001s_00028_rgb.jpg", "mask_path": "assets/cat_13/v_-Sp-gOEGy2o_612_821/0001s_00028_mask.png", "mask_status": "available" }, { "frame": "00060", "timestamp_sec": 22.41, "offset_sec": 2, "rgb_path": "assets/cat_13/v_-Sp-gOEGy2o_612_821/0002s_00060_rgb.jpg", "mask_path": "assets/cat_13/v_-Sp-gOEGy2o_612_821/0002s_00060_mask.png", "mask_status": "available" }, { "frame": "00088", "timestamp_sec": 23.41, "offset_sec": 3, "rgb_path": "assets/cat_13/v_-Sp-gOEGy2o_612_821/0003s_00088_rgb.jpg", "mask_path": "assets/cat_13/v_-Sp-gOEGy2o_612_821/0003s_00088_mask.png", "mask_status": "available" }, { "frame": "00120", "timestamp_sec": 24.41, "offset_sec": 4, "rgb_path": "assets/cat_13/v_-Sp-gOEGy2o_612_821/0004s_00120_rgb.jpg", "mask_path": "assets/cat_13/v_-Sp-gOEGy2o_612_821/0004s_00120_mask.png", "mask_status": "available" }, { "frame": "00148", "timestamp_sec": 25.41, "offset_sec": 5, "rgb_path": "assets/cat_13/v_-Sp-gOEGy2o_612_821/0005s_00148_rgb.jpg", "mask_path": "assets/cat_13/v_-Sp-gOEGy2o_612_821/0005s_00148_mask.png", "mask_status": "available" }, { "frame": "00180", "timestamp_sec": 26.41, "offset_sec": 6, "rgb_path": "assets/cat_13/v_-Sp-gOEGy2o_612_821/0006s_00180_rgb.jpg", "mask_path": "assets/cat_13/v_-Sp-gOEGy2o_612_821/0006s_00180_mask.png", "mask_status": "available" } ], "obj_id": [ 0 ], "anno_id": [ 78 ], "fps_inferred_from_clip_interval": 29.8571 }, "usage": { "input_tokens": 31169, "input_tokens_details": { "cache_write_tokens": 31166, "cached_tokens": 0 }, "output_tokens": 4143, "output_tokens_details": { "reasoning_tokens": 2588 }, "total_tokens": 35312 }, "model": "gpt-5.6-sol", "effort": "medium", "custom_id": "cat13_cap03_v_-Sp-gOEGy2o_612_821", "annotation": { "caption": "After striking a match with his right hand, he transfers the lit match to his left hand. He holds the burning match while looking down at a pile of kindling in a fire pit.", "entity_groups": [ { "id": "entity_person_1", "label": "person", "spans": [ { "text": "his", "range": [ 28, 31 ] }, { "text": "he", "range": [ 44, 46 ] }, { "text": "his", "range": [ 74, 77 ] }, { "text": "He", "range": [ 89, 91 ] } ], "source_mask_link": "selected_person" }, { "id": "entity_match_1", "label": "match", "spans": [ { "text": "a match", "range": [ 15, 22 ] }, { "text": "the lit match", "range": [ 57, 70 ] }, { "text": "the burning match", "range": [ 98, 115 ] } ], "source_mask_link": null }, { "id": "entity_right_hand_1", "label": "right hand", "spans": [ { "text": "his right hand", "range": [ 28, 42 ] } ], "source_mask_link": null }, { "id": "entity_left_hand_1", "label": "left hand", "spans": [ { "text": "his left hand", "range": [ 74, 87 ] } ], "source_mask_link": null }, { "id": "entity_kindling_1", "label": "kindling", "spans": [ { "text": "a pile of kindling", "range": [ 138, 156 ] } ], "source_mask_link": null }, { "id": "entity_fire_pit_1", "label": "fire pit", "spans": [ { "text": "a fire pit", "range": [ 160, 170 ] } ], "source_mask_link": null } ], "actions": [ { "id": "action_strike_1", "mentions": [ { "text": "striking", "ranges": [ [ 6, 14 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "type": "simple", "subject_refs": [ "entity_person_1" ], "object_refs": [ "entity_match_1" ], "confidence": "medium", "temporal_status": "uncertain", "intervals": [], "evidence": "At 0.000 s the masked man manipulates a matchbox near his hands, but no sampled frame clearly resolves the match-striking stroke.", "mask_relation": "subject matches the supplied selected-person mask", "source_subject_refs": "entity_person_1" }, { "id": "action_transfer_1", "mentions": [ { "text": "transfers", "ranges": [ [ 47, 56 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "type": "simple", "subject_refs": [ "entity_person_1" ], "object_refs": [ "entity_match_1" ], "confidence": "medium", "temporal_status": "uncertain", "intervals": [], "evidence": "The sampled frames show the masked man handling a thin match and moving both hands, but they do not clearly show a transfer from one hand to the other.", "mask_relation": "subject matches the supplied selected-person mask", "source_subject_refs": "entity_person_1" }, { "id": "action_hold_1", "mentions": [ { "text": "holds", "ranges": [ [ 92, 97 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "type": "simple", "subject_refs": [ "entity_person_1" ], "object_refs": [ "entity_match_1" ], "confidence": "high", "temporal_status": "supported_sparse", "intervals": [ { "start_s": 1.0, "end_s": 5.0, "source_start_s": 21.41, "source_end_s": 25.41 }, { "start_s": 6.0, "end_s": 7.0, "source_start_s": 26.41, "source_end_s": 27.41 } ], "evidence": "At 1.000–4.000 s and again at 6.000 s, the masked man visibly holds a thin match-like object. A flame is not clearly resolved.", "mask_relation": "subject matches the supplied selected-person mask", "source_subject_refs": "entity_person_1" }, { "id": "action_look_1", "mentions": [ { "text": "looking", "ranges": [ [ 122, 129 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "type": "prepositional", "subject_refs": [ "entity_person_1" ], "object_refs": [ "entity_kindling_1" ], "confidence": "high", "temporal_status": "supported_sparse", "intervals": [ { "start_s": 0.0, "end_s": 6.0, "source_start_s": 20.41, "source_end_s": 26.41 } ], "evidence": "From 0.000 through 5.000 s, the masked man's head and gaze are directed downward toward the kindling and fire-pit area; his head is cropped at 6.000 s.", "mask_relation": "subject matches the supplied selected-person mask", "source_subject_refs": "entity_person_1" } ], "notes": [ "The supplied mask track identifies only the single man and includes his body and limbs; it does not provide separate instance masks for the match, hands, kindling, or fire pit.", "The match's participle “burning” denotes a process whose subject is the unmasked match, so it is omitted as a separate action under the mask-linked-subject rule.", "The modifier “lit” describes the match's resulting state and is not a separate action.", "The sampled frames do not clearly show a visible flame, a discrete striking stroke, or a hand-to-hand transfer.", "The source-video interval is [20.410000, 27.410000), corresponding to clip-relative [0.000000, 7.000000)." ] }, "model_output_raw": { "caption": "After striking a match with his right hand, he transfers the lit match to his left hand. He holds the burning match while looking down at a pile of kindling in a fire pit.", "entity_groups": [ { "id": "entity_person_1", "label": "person", "spans": [ { "text": "his", "range": [ 28, 31 ] }, { "text": "he", "range": [ 44, 46 ] }, { "text": "his", "range": [ 74, 77 ] }, { "text": "He", "range": [ 89, 91 ] } ], "source_mask": "selected_person" }, { "id": "entity_match_1", "label": "match", "spans": [ { "text": "a match", "range": [ 15, 22 ] }, { "text": "the lit match", "range": [ 57, 70 ] }, { "text": "the burning match", "range": [ 98, 115 ] } ], "source_mask": null }, { "id": "entity_right_hand_1", "label": "right hand", "spans": [ { "text": "his right hand", "range": [ 28, 42 ] } ], "source_mask": null }, { "id": "entity_left_hand_1", "label": "left hand", "spans": [ { "text": "his left hand", "range": [ 74, 87 ] } ], "source_mask": null }, { "id": "entity_kindling_1", "label": "kindling", "spans": [ { "text": "a pile of kindling", "range": [ 138, 156 ] } ], "source_mask": null }, { "id": "entity_fire_pit_1", "label": "fire pit", "spans": [ { "text": "a fire pit", "range": [ 160, 170 ] } ], "source_mask": null } ], "actions": [ { "id": "action_strike_1", "mentions": [ { "text": "striking", "ranges": [ [ 6, 14 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "type": "simple", "subject_ref": "entity_person_1", "object_refs": [ "entity_match_1" ], "instrument_refs": [ "entity_right_hand_1" ], "confidence": "medium", "mask_relation": "subject matches the supplied selected-person mask", "temporal_status": "uncertain", "intervals": [], "evidence": "At 0.000 s the masked man manipulates a matchbox near his hands, but no sampled frame clearly resolves the match-striking stroke." }, { "id": "action_transfer_1", "mentions": [ { "text": "transfers", "ranges": [ [ 47, 56 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "type": "simple", "subject_ref": "entity_person_1", "object_refs": [ "entity_match_1" ], "destination_refs": [ "entity_left_hand_1" ], "confidence": "medium", "mask_relation": "subject matches the supplied selected-person mask", "temporal_status": "uncertain", "intervals": [], "evidence": "The sampled frames show the masked man handling a thin match and moving both hands, but they do not clearly show a transfer from one hand to the other." }, { "id": "action_hold_1", "mentions": [ { "text": "holds", "ranges": [ [ 92, 97 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "type": "simple", "subject_ref": "entity_person_1", "object_refs": [ "entity_match_1" ], "confidence": "high", "mask_relation": "subject matches the supplied selected-person mask", "temporal_status": "supported_sparse", "intervals": [ { "clip_relative": [ 1.0, 5.0 ], "source_video": [ 21.41, 25.41 ] }, { "clip_relative": [ 6.0, 7.0 ], "source_video": [ 26.41, 27.41 ] } ], "evidence": "At 1.000–4.000 s and again at 6.000 s, the masked man visibly holds a thin match-like object. A flame is not clearly resolved." }, { "id": "action_look_1", "mentions": [ { "text": "looking", "ranges": [ [ 122, 129 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "type": "prepositional", "subject_ref": "entity_person_1", "object_refs": [ "entity_kindling_1" ], "confidence": "high", "mask_relation": "subject matches the supplied selected-person mask", "temporal_status": "supported_sparse", "intervals": [ { "clip_relative": [ 0.0, 6.0 ], "source_video": [ 20.41, 26.41 ] } ], "evidence": "From 0.000 through 5.000 s, the masked man's head and gaze are directed downward toward the kindling and fire-pit area; his head is cropped at 6.000 s." } ], "notes": [ "The supplied mask track identifies only the single man and includes his body and limbs; it does not provide separate instance masks for the match, hands, kindling, or fire pit.", "The match's participle “burning” denotes a process whose subject is the unmasked match, so it is omitted as a separate action under the mask-linked-subject rule.", "The modifier “lit” describes the match's resulting state and is not a separate action.", "The sampled frames do not clearly show a visible flame, a discrete striking stroke, or a hand-to-hand transfer.", "The source-video interval is [20.410000, 27.410000), corresponding to clip-relative [0.000000, 7.000000)." ] }, "normalization_repairs": 0 }, { "source_video_id": "v_-Sp-gOEGy2o", "caption_count_category": 13, "clip": { "caption_index": 4, "caption": "The man carefully places the burning match into the kindling, successfully igniting it. As a small flame appears, he uses his right hand to add a small piece of wood to the fire.", "source_interval_sec": [ 27.41, 36.91 ], "loom_clip_id": "v_-Sp-gOEGy2o_821_1106", "clip_frame_interval_4fps": [ 821, 1106 ], "sample_count_1fps": 10, "sampled_frames": [ { "frame": "00000", "timestamp_sec": 27.41, "offset_sec": 0, "rgb_path": "assets/cat_13/v_-Sp-gOEGy2o_821_1106/0000s_00000_rgb.jpg", "mask_path": "assets/cat_13/v_-Sp-gOEGy2o_821_1106/0000s_00000_mask.png", "mask_status": "available" }, { "frame": "00032", "timestamp_sec": 28.41, "offset_sec": 1, "rgb_path": "assets/cat_13/v_-Sp-gOEGy2o_821_1106/0001s_00032_rgb.jpg", "mask_path": "assets/cat_13/v_-Sp-gOEGy2o_821_1106/0001s_00032_mask.png", "mask_status": "available" }, { "frame": "00060", "timestamp_sec": 29.41, "offset_sec": 2, "rgb_path": "assets/cat_13/v_-Sp-gOEGy2o_821_1106/0002s_00060_rgb.jpg", "mask_path": "assets/cat_13/v_-Sp-gOEGy2o_821_1106/0002s_00060_mask.png", "mask_status": "available" }, { "frame": "00092", "timestamp_sec": 30.41, "offset_sec": 3, "rgb_path": "assets/cat_13/v_-Sp-gOEGy2o_821_1106/0003s_00092_rgb.jpg", "mask_path": "assets/cat_13/v_-Sp-gOEGy2o_821_1106/0003s_00092_mask.png", "mask_status": "available" }, { "frame": "00120", "timestamp_sec": 31.41, "offset_sec": 4, "rgb_path": "assets/cat_13/v_-Sp-gOEGy2o_821_1106/0004s_00120_rgb.jpg", "mask_path": "assets/cat_13/v_-Sp-gOEGy2o_821_1106/0004s_00120_mask.png", "mask_status": "available" }, { "frame": "00152", "timestamp_sec": 32.41, "offset_sec": 5, "rgb_path": "assets/cat_13/v_-Sp-gOEGy2o_821_1106/0005s_00152_rgb.jpg", "mask_path": "assets/cat_13/v_-Sp-gOEGy2o_821_1106/0005s_00152_mask.png", "mask_status": "available" }, { "frame": "00180", "timestamp_sec": 33.41, "offset_sec": 6, "rgb_path": "assets/cat_13/v_-Sp-gOEGy2o_821_1106/0006s_00180_rgb.jpg", "mask_path": "assets/cat_13/v_-Sp-gOEGy2o_821_1106/0006s_00180_mask.png", "mask_status": "available" }, { "frame": "00212", "timestamp_sec": 34.41, "offset_sec": 7, "rgb_path": "assets/cat_13/v_-Sp-gOEGy2o_821_1106/0007s_00212_rgb.jpg", "mask_path": "assets/cat_13/v_-Sp-gOEGy2o_821_1106/0007s_00212_mask.png", "mask_status": "available" }, { "frame": "00240", "timestamp_sec": 35.41, "offset_sec": 8, "rgb_path": "assets/cat_13/v_-Sp-gOEGy2o_821_1106/0008s_00240_rgb.jpg", "mask_path": "assets/cat_13/v_-Sp-gOEGy2o_821_1106/0008s_00240_mask.png", "mask_status": "available" }, { "frame": "00272", "timestamp_sec": 36.41, "offset_sec": 9, "rgb_path": "assets/cat_13/v_-Sp-gOEGy2o_821_1106/0009s_00272_rgb.jpg", "mask_path": "assets/cat_13/v_-Sp-gOEGy2o_821_1106/0009s_00272_mask.png", "mask_status": "available" } ], "obj_id": [ 0 ], "anno_id": [ 80 ], "fps_inferred_from_clip_interval": 30.0 }, "usage": { "input_tokens": 37959, "input_tokens_details": { "cache_write_tokens": 37956, "cached_tokens": 0 }, "output_tokens": 3805, "output_tokens_details": { "reasoning_tokens": 2225 }, "total_tokens": 41764 }, "model": "gpt-5.6-sol", "effort": "medium", "custom_id": "cat13_cap04_v_-Sp-gOEGy2o_821_1106", "annotation": { "caption": "The man carefully places the burning match into the kindling, successfully igniting it. As a small flame appears, he uses his right hand to add a small piece of wood to the fire.", "entity_groups": [ { "id": "entity_1", "label": "man", "spans": [ { "text": "The man", "range": [ 0, 7 ] }, { "text": "he", "range": [ 114, 116 ] }, { "text": "his", "range": [ 122, 125 ] } ], "source_mask_link": "selected_person" }, { "id": "entity_2", "label": "match", "spans": [ { "text": "the burning match", "range": [ 25, 42 ] } ], "source_mask_link": null }, { "id": "entity_3", "label": "kindling", "spans": [ { "text": "the kindling", "range": [ 48, 60 ] }, { "text": "it", "range": [ 84, 86 ] } ], "source_mask_link": null }, { "id": "entity_4", "label": "flame/fire", "spans": [ { "text": "a small flame", "range": [ 91, 104 ] }, { "text": "the fire", "range": [ 169, 177 ] } ], "source_mask_link": null }, { "id": "entity_5", "label": "right hand", "spans": [ { "text": "his right hand", "range": [ 122, 136 ] } ], "source_mask_link": null }, { "id": "entity_6", "label": "piece of wood", "spans": [ { "text": "a small piece of wood", "range": [ 144, 165 ] } ], "source_mask_link": null } ], "actions": [ { "id": "action_1", "mentions": [ { "text": "places", "ranges": [ [ 18, 24 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "type": "simple", "subject_refs": [ "entity_1" ], "object_refs": [], "confidence": "medium", "temporal_status": "localized_from_sparse_samples", "intervals": [ { "start_s": 0.0, "end_s": 8.0, "source_start_s": 27.41, "source_end_s": 35.41 } ], "evidence": "At clip-relative samples 0.0–7.0 s, the mask-linked man keeps his hands and the match close to the kindling while positioning it within the tinder structure.", "mask_relation": "subject_matches_supplied_selected-person_mask", "source_subject_refs": "entity_1" }, { "id": "action_2", "mentions": [ { "text": "igniting", "ranges": [ [ 75, 83 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "type": "simple", "subject_refs": [ "entity_1" ], "object_refs": [], "confidence": "high", "temporal_status": "localized_from_sparse_samples", "intervals": [ { "start_s": 8.0, "end_s": 9.0, "source_start_s": 35.41, "source_end_s": 36.41 } ], "evidence": "A visible orange flame emerges in the kindling at the 8.0 s sample while the mask-linked man's hand remains at the ignition point.", "mask_relation": "subject_matches_supplied_selected-person_mask", "source_subject_refs": "entity_1" }, { "id": "action_3", "mentions": [ { "text": "uses", "ranges": [ [ 117, 121 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "type": "simple", "subject_refs": [ "entity_1" ], "object_refs": [], "confidence": "high", "temporal_status": "localized_from_sparse_samples", "intervals": [ { "start_s": 9.0, "end_s": 9.499999999999996, "source_start_s": 36.41, "source_end_s": 36.91 } ], "evidence": "At 9.0 s, the selected man's right hand visibly reaches over the burning kindling while manipulating a piece of wood.", "mask_relation": "subject_matches_supplied_selected-person_mask", "source_subject_refs": "entity_1" }, { "id": "action_4", "mentions": [ { "text": "add", "ranges": [ [ 140, 143 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "type": "simple", "subject_refs": [ "entity_1" ], "object_refs": [], "confidence": "high", "temporal_status": "localized_from_sparse_samples", "intervals": [ { "start_s": 9.0, "end_s": 9.499999999999996, "source_start_s": 36.41, "source_end_s": 36.91 } ], "evidence": "The 9.0 s sample shows the mask-linked man lowering a small wood piece onto the burning kindling; support reaches the clip's 9.5 s right boundary.", "mask_relation": "subject_matches_supplied_selected-person_mask", "source_subject_refs": "entity_1" } ], "notes": [ "The supplied masklet consistently identifies the man, including his arms and hands, but it is a person mask rather than a separate hand-instance mask.", "No independent source masks are supplied for the match, kindling, flame/fire, hand, or wood piece.", "The process expressed by “burning” has the match as its subject, and “appears” has the flame as its subject. These actions are omitted because neither subject has a corresponding supplied mask.", "Temporal intervals are sparse sampled-frame support intervals within the source interval [27.410000, 36.910000)." ] }, "model_output_raw": { "caption": "The man carefully places the burning match into the kindling, successfully igniting it. As a small flame appears, he uses his right hand to add a small piece of wood to the fire.", "entity_groups": [ { "id": "entity_1", "label": "man", "spans": [ { "text": "The man", "range": [ 0, 7 ] }, { "text": "he", "range": [ 114, 116 ] }, { "text": "his", "range": [ 122, 125 ] } ], "source_mask": "selected_person_masklet" }, { "id": "entity_2", "label": "match", "spans": [ { "text": "the burning match", "range": [ 25, 42 ] } ], "source_mask": null }, { "id": "entity_3", "label": "kindling", "spans": [ { "text": "the kindling", "range": [ 48, 60 ] }, { "text": "it", "range": [ 84, 86 ] } ], "source_mask": null }, { "id": "entity_4", "label": "flame/fire", "spans": [ { "text": "a small flame", "range": [ 91, 104 ] }, { "text": "the fire", "range": [ 169, 177 ] } ], "source_mask": null }, { "id": "entity_5", "label": "right hand", "spans": [ { "text": "his right hand", "range": [ 122, 136 ] } ], "source_mask": null }, { "id": "entity_6", "label": "piece of wood", "spans": [ { "text": "a small piece of wood", "range": [ 144, 165 ] } ], "source_mask": null } ], "actions": [ { "id": "action_1", "mentions": [ { "text": "places", "occurrence": 1, "ranges": [ [ 18, 24 ] ], "particle": { "text": null, "occurrence": 1 } } ], "type": "simple", "subject_group": "entity_1", "object_groups": [ { "group_id": "entity_2", "role": "theme" }, { "group_id": "entity_3", "role": "destination" } ], "confidence": "medium", "temporal_status": "localized_from_sparse_samples", "intervals": [ { "start_s": 0.0, "end_s": 8.0 } ], "evidence": "At clip-relative samples 0.0–7.0 s, the mask-linked man keeps his hands and the match close to the kindling while positioning it within the tinder structure.", "mask_relation": "subject_matches_supplied_selected-person_mask" }, { "id": "action_2", "mentions": [ { "text": "igniting", "occurrence": 1, "ranges": [ [ 75, 83 ] ], "particle": { "text": null, "occurrence": 1 } } ], "type": "simple", "subject_group": "entity_1", "object_groups": [ { "group_id": "entity_3", "role": "patient" } ], "confidence": "high", "temporal_status": "localized_from_sparse_samples", "intervals": [ { "start_s": 8.0, "end_s": 9.0 } ], "evidence": "A visible orange flame emerges in the kindling at the 8.0 s sample while the mask-linked man's hand remains at the ignition point.", "mask_relation": "subject_matches_supplied_selected-person_mask" }, { "id": "action_3", "mentions": [ { "text": "uses", "occurrence": 1, "ranges": [ [ 117, 121 ] ], "particle": { "text": null, "occurrence": 1 } } ], "type": "simple", "subject_group": "entity_1", "object_groups": [ { "group_id": "entity_5", "role": "instrument" } ], "confidence": "high", "temporal_status": "localized_from_sparse_samples", "intervals": [ { "start_s": 9.0, "end_s": 9.5 } ], "evidence": "At 9.0 s, the selected man's right hand visibly reaches over the burning kindling while manipulating a piece of wood.", "mask_relation": "subject_matches_supplied_selected-person_mask" }, { "id": "action_4", "mentions": [ { "text": "add", "occurrence": 1, "ranges": [ [ 140, 143 ] ], "particle": { "text": null, "occurrence": 1 } } ], "type": "simple", "subject_group": "entity_1", "object_groups": [ { "group_id": "entity_6", "role": "theme" }, { "group_id": "entity_4", "role": "destination" } ], "confidence": "high", "temporal_status": "localized_from_sparse_samples", "intervals": [ { "start_s": 9.0, "end_s": 9.5 } ], "evidence": "The 9.0 s sample shows the mask-linked man lowering a small wood piece onto the burning kindling; support reaches the clip's 9.5 s right boundary.", "mask_relation": "subject_matches_supplied_selected-person_mask" } ], "notes": [ "The supplied masklet consistently identifies the man, including his arms and hands, but it is a person mask rather than a separate hand-instance mask.", "No independent source masks are supplied for the match, kindling, flame/fire, hand, or wood piece.", "The process expressed by “burning” has the match as its subject, and “appears” has the flame as its subject. These actions are omitted because neither subject has a corresponding supplied mask.", "Temporal intervals are sparse sampled-frame support intervals within the source interval [27.410000, 36.910000)." ] }, "normalization_repairs": 0 }, { "source_video_id": "v_-Sp-gOEGy2o", "caption_count_category": 13, "clip": { "caption_index": 5, "caption": "He picks up a black handheld blower with his left hand. He then uses his right hand to pick up another small piece of wood and place it onto the smoldering fire.", "source_interval_sec": [ 36.91, 46.41 ], "loom_clip_id": "v_-Sp-gOEGy2o_1106_1391", "clip_frame_interval_4fps": [ 1106, 1391 ], "sample_count_1fps": 10, "sampled_frames": [ { "frame": "00000", "timestamp_sec": 36.91, "offset_sec": 0, "rgb_path": "assets/cat_13/v_-Sp-gOEGy2o_1106_1391/0000s_00000_rgb.jpg", "mask_path": "assets/cat_13/v_-Sp-gOEGy2o_1106_1391/0000s_00000_mask.png", "mask_status": "available" }, { "frame": "00028", "timestamp_sec": 37.91, "offset_sec": 1, "rgb_path": "assets/cat_13/v_-Sp-gOEGy2o_1106_1391/0001s_00028_rgb.jpg", "mask_path": "assets/cat_13/v_-Sp-gOEGy2o_1106_1391/0001s_00028_mask.png", "mask_status": "available" }, { "frame": "00060", "timestamp_sec": 38.91, "offset_sec": 2, "rgb_path": "assets/cat_13/v_-Sp-gOEGy2o_1106_1391/0002s_00060_rgb.jpg", "mask_path": "assets/cat_13/v_-Sp-gOEGy2o_1106_1391/0002s_00060_mask.png", "mask_status": "available" }, { "frame": "00088", "timestamp_sec": 39.91, "offset_sec": 3, "rgb_path": "assets/cat_13/v_-Sp-gOEGy2o_1106_1391/0003s_00088_rgb.jpg", "mask_path": "assets/cat_13/v_-Sp-gOEGy2o_1106_1391/0003s_00088_mask.png", "mask_status": "available" }, { "frame": "00120", "timestamp_sec": 40.91, "offset_sec": 4, "rgb_path": "assets/cat_13/v_-Sp-gOEGy2o_1106_1391/0004s_00120_rgb.jpg", "mask_path": "assets/cat_13/v_-Sp-gOEGy2o_1106_1391/0004s_00120_mask.png", "mask_status": "available" }, { "frame": "00148", "timestamp_sec": 41.91, "offset_sec": 5, "rgb_path": "assets/cat_13/v_-Sp-gOEGy2o_1106_1391/0005s_00148_rgb.jpg", "mask_path": "assets/cat_13/v_-Sp-gOEGy2o_1106_1391/0005s_00148_mask.png", "mask_status": "available" }, { "frame": "00180", "timestamp_sec": 42.91, "offset_sec": 6, "rgb_path": "assets/cat_13/v_-Sp-gOEGy2o_1106_1391/0006s_00180_rgb.jpg", "mask_path": "assets/cat_13/v_-Sp-gOEGy2o_1106_1391/0006s_00180_mask.png", "mask_status": "available" }, { "frame": "00208", "timestamp_sec": 43.91, "offset_sec": 7, "rgb_path": "assets/cat_13/v_-Sp-gOEGy2o_1106_1391/0007s_00208_rgb.jpg", "mask_path": "assets/cat_13/v_-Sp-gOEGy2o_1106_1391/0007s_00208_mask.png", "mask_status": "available" }, { "frame": "00240", "timestamp_sec": 44.91, "offset_sec": 8, "rgb_path": "assets/cat_13/v_-Sp-gOEGy2o_1106_1391/0008s_00240_rgb.jpg", "mask_path": "assets/cat_13/v_-Sp-gOEGy2o_1106_1391/0008s_00240_mask.png", "mask_status": "available" }, { "frame": "00268", "timestamp_sec": 45.91, "offset_sec": 9, "rgb_path": "assets/cat_13/v_-Sp-gOEGy2o_1106_1391/0009s_00268_rgb.jpg", "mask_path": "assets/cat_13/v_-Sp-gOEGy2o_1106_1391/0009s_00268_mask.png", "mask_status": "available" } ], "obj_id": [ 0 ], "anno_id": [ 68 ], "fps_inferred_from_clip_interval": 30.0 }, "usage": { "input_tokens": 37956, "input_tokens_details": { "cache_write_tokens": 37953, "cached_tokens": 0 }, "output_tokens": 4610, "output_tokens_details": { "reasoning_tokens": 3010 }, "total_tokens": 42566 }, "model": "gpt-5.6-sol", "effort": "medium", "custom_id": "cat13_cap05_v_-Sp-gOEGy2o_1106_1391", "annotation": { "caption": "He picks up a black handheld blower with his left hand. He then uses his right hand to pick up another small piece of wood and place it onto the smoldering fire.", "entity_groups": [ { "id": "person_1", "label": "person", "spans": [ { "text": "He", "range": [ 0, 2 ] }, { "text": "his", "range": [ 41, 44 ] }, { "text": "He", "range": [ 56, 58 ] }, { "text": "his", "range": [ 69, 72 ] } ], "source_mask_link": "selected_person" }, { "id": "blower_1", "label": "blower", "spans": [ { "text": "a black handheld blower", "range": [ 12, 35 ] } ], "source_mask_link": null }, { "id": "left_hand_1", "label": "left hand", "spans": [ { "text": "his left hand", "range": [ 41, 54 ] } ], "source_mask_link": null }, { "id": "right_hand_1", "label": "right hand", "spans": [ { "text": "his right hand", "range": [ 69, 83 ] } ], "source_mask_link": null }, { "id": "wood_piece_1", "label": "piece of wood", "spans": [ { "text": "another small piece of wood", "range": [ 95, 122 ] }, { "text": "it", "range": [ 133, 135 ] } ], "source_mask_link": null }, { "id": "fire_1", "label": "fire", "spans": [ { "text": "the smoldering fire", "range": [ 141, 160 ] } ], "source_mask_link": null } ], "actions": [ { "id": "action_1", "mentions": [ { "text": "picks up", "ranges": [ [ 3, 11 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "type": "phrasal", "subject_refs": [ "person_1" ], "object_refs": [ "blower_1" ], "confidence": "medium", "temporal_status": "sparsely localized", "intervals": [ { "start_s": 0.0, "end_s": 2.0, "source_start_s": 36.91, "source_end_s": 38.91 } ], "evidence": "At clip times 0.000–1.000 s (source 36.910–37.910 s), the masked person brings the black blower into view and holds it in his left hand; the exact pickup transition is only sparsely sampled.", "mask_relation": "subject matches the supplied selected-person mask", "source_subject_refs": "person_1" }, { "id": "action_2", "mentions": [ { "text": "uses", "ranges": [ [ 64, 68 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "type": "simple", "subject_refs": [ "person_1" ], "object_refs": [ "right_hand_1" ], "confidence": "high", "temporal_status": "localized", "intervals": [ { "start_s": 7.0, "end_s": 9.5, "source_start_s": 43.91, "source_end_s": 46.41 } ], "evidence": "At clip times 7.000–9.000 s (source 43.910–45.910 s), the masked person visibly reaches with and manipulates the wood using his right hand.", "mask_relation": "subject matches the supplied selected-person mask", "source_subject_refs": "person_1" }, { "id": "action_3", "mentions": [ { "text": "pick up", "ranges": [ [ 87, 94 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "type": "phrasal", "subject_refs": [ "person_1" ], "object_refs": [ "wood_piece_1" ], "confidence": "high", "temporal_status": "localized", "intervals": [ { "start_s": 7.0, "end_s": 9.0, "source_start_s": 43.91, "source_end_s": 45.91 } ], "evidence": "Between clip times 7.000 and 8.000 s (source 43.910–44.910 s), the masked person reaches toward a small piece of wood and brings it toward the fire with his right hand.", "mask_relation": "subject matches the supplied selected-person mask", "source_subject_refs": "person_1" }, { "id": "action_4", "mentions": [ { "text": "place", "ranges": [ [ 127, 132 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "type": "simple", "subject_refs": [ "person_1" ], "object_refs": [ "wood_piece_1" ], "confidence": "high", "temporal_status": "localized", "intervals": [ { "start_s": 8.0, "end_s": 9.5, "source_start_s": 44.91, "source_end_s": 46.41 } ], "evidence": "At clip times 8.000–9.000 s (source 44.910–45.910 s), the masked person lowers and positions the wood onto the fire area. The final sampled frame supports continuation to the clip boundary at 9.500 s.", "mask_relation": "subject matches the supplied selected-person mask", "source_subject_refs": "person_1" } ], "notes": [ "The supplied mask consistently identifies the single person referred to by both occurrences of “He” and the possessives “his.”", "The blower, hands, wood piece, and fire have no dedicated supplied instance masks, so their source-mask links are null.", "The process “smoldering” is not included as an action because its subject is the unmasked fire; VideoLoom actions are restricted to subjects identified by the supplied mask.", "The first blower pickup is only sparsely supported because the sampled frames show the blower entering or already held in the left hand rather than a complete pickup transition." ] }, "model_output_raw": { "caption": "He picks up a black handheld blower with his left hand. He then uses his right hand to pick up another small piece of wood and place it onto the smoldering fire.", "entity_groups": [ { "id": "person_1", "label": "person", "spans": [ { "text": "He", "range": [ 0, 2 ] }, { "text": "his", "range": [ 41, 44 ] }, { "text": "He", "range": [ 56, 58 ] }, { "text": "his", "range": [ 69, 72 ] } ], "source_mask": "selected_person_masklet" }, { "id": "blower_1", "label": "blower", "spans": [ { "text": "a black handheld blower", "range": [ 12, 35 ] } ], "source_mask": null }, { "id": "left_hand_1", "label": "left hand", "spans": [ { "text": "his left hand", "range": [ 41, 54 ] } ], "source_mask": null }, { "id": "right_hand_1", "label": "right hand", "spans": [ { "text": "his right hand", "range": [ 69, 83 ] } ], "source_mask": null }, { "id": "wood_piece_1", "label": "piece of wood", "spans": [ { "text": "another small piece of wood", "range": [ 95, 122 ] }, { "text": "it", "range": [ 133, 135 ] } ], "source_mask": null }, { "id": "fire_1", "label": "fire", "spans": [ { "text": "the smoldering fire", "range": [ 141, 160 ] } ], "source_mask": null } ], "actions": [ { "id": "action_1", "mentions": [ { "text": "picks up", "ranges": [ [ 3, 11 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "type": "phrasal", "subject_ref": "person_1", "object_refs": [ "blower_1" ], "instrument_refs": [ "left_hand_1" ], "confidence": "medium", "mask_relation": "subject matches the supplied selected-person mask", "temporal_status": "sparsely localized", "intervals": [ { "start_s": 0.0, "end_s": 2.0 } ], "evidence": "At clip times 0.000–1.000 s (source 36.910–37.910 s), the masked person brings the black blower into view and holds it in his left hand; the exact pickup transition is only sparsely sampled." }, { "id": "action_2", "mentions": [ { "text": "uses", "ranges": [ [ 64, 68 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "type": "simple", "subject_ref": "person_1", "object_refs": [ "right_hand_1" ], "confidence": "high", "mask_relation": "subject matches the supplied selected-person mask", "temporal_status": "localized", "intervals": [ { "start_s": 7.0, "end_s": 9.5 } ], "evidence": "At clip times 7.000–9.000 s (source 43.910–45.910 s), the masked person visibly reaches with and manipulates the wood using his right hand." }, { "id": "action_3", "mentions": [ { "text": "pick up", "ranges": [ [ 87, 94 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "type": "phrasal", "subject_ref": "person_1", "object_refs": [ "wood_piece_1" ], "instrument_refs": [ "right_hand_1" ], "confidence": "high", "mask_relation": "subject matches the supplied selected-person mask", "temporal_status": "localized", "intervals": [ { "start_s": 7.0, "end_s": 9.0 } ], "evidence": "Between clip times 7.000 and 8.000 s (source 43.910–44.910 s), the masked person reaches toward a small piece of wood and brings it toward the fire with his right hand." }, { "id": "action_4", "mentions": [ { "text": "place", "ranges": [ [ 127, 132 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "type": "simple", "subject_ref": "person_1", "object_refs": [ "wood_piece_1" ], "destination_refs": [ "fire_1" ], "instrument_refs": [ "right_hand_1" ], "confidence": "high", "mask_relation": "subject matches the supplied selected-person mask", "temporal_status": "localized", "intervals": [ { "start_s": 8.0, "end_s": 9.5 } ], "evidence": "At clip times 8.000–9.000 s (source 44.910–45.910 s), the masked person lowers and positions the wood onto the fire area. The final sampled frame supports continuation to the clip boundary at 9.500 s." } ], "notes": [ "The supplied mask consistently identifies the single person referred to by both occurrences of “He” and the possessives “his.”", "The blower, hands, wood piece, and fire have no dedicated supplied instance masks, so their source-mask links are null.", "The process “smoldering” is not included as an action because its subject is the unmasked fire; VideoLoom actions are restricted to subjects identified by the supplied mask.", "The first blower pickup is only sparsely supported because the sampled frames show the blower entering or already held in the left hand rather than a complete pickup transition." ] }, "normalization_repairs": 0 }, { "source_video_id": "v_-Sp-gOEGy2o", "caption_count_category": 13, "clip": { "caption_index": 6, "caption": "He picks up a curved piece of bark with his right hand and places it on the smoking kindling. He continues to add more small pieces of wood as the flame begins to grow.", "source_interval_sec": [ 46.41, 64.91 ], "loom_clip_id": "v_-Sp-gOEGy2o_1391_1945", "clip_frame_interval_4fps": [ 1391, 1945 ], "sample_count_1fps": 19, "sampled_frames": [ { "frame": "00000", "timestamp_sec": 46.41, "offset_sec": 0, "rgb_path": "assets/cat_13/v_-Sp-gOEGy2o_1391_1945/0000s_00000_rgb.jpg", "mask_path": "assets/cat_13/v_-Sp-gOEGy2o_1391_1945/0000s_00000_mask.png", "mask_status": "available" }, { "frame": "00028", "timestamp_sec": 47.41, "offset_sec": 1, "rgb_path": "assets/cat_13/v_-Sp-gOEGy2o_1391_1945/0001s_00028_rgb.jpg", "mask_path": "assets/cat_13/v_-Sp-gOEGy2o_1391_1945/0001s_00028_mask.png", "mask_status": "available" }, { "frame": "00060", "timestamp_sec": 48.41, "offset_sec": 2, "rgb_path": "assets/cat_13/v_-Sp-gOEGy2o_1391_1945/0002s_00060_rgb.jpg", "mask_path": "assets/cat_13/v_-Sp-gOEGy2o_1391_1945/0002s_00060_mask.png", "mask_status": "available" }, { "frame": "00088", "timestamp_sec": 49.41, "offset_sec": 3, "rgb_path": "assets/cat_13/v_-Sp-gOEGy2o_1391_1945/0003s_00088_rgb.jpg", "mask_path": "assets/cat_13/v_-Sp-gOEGy2o_1391_1945/0003s_00088_mask.png", "mask_status": "available" }, { "frame": "00120", "timestamp_sec": 50.41, "offset_sec": 4, "rgb_path": "assets/cat_13/v_-Sp-gOEGy2o_1391_1945/0004s_00120_rgb.jpg", "mask_path": "assets/cat_13/v_-Sp-gOEGy2o_1391_1945/0004s_00120_mask.png", "mask_status": "available" }, { "frame": "00148", "timestamp_sec": 51.41, "offset_sec": 5, "rgb_path": "assets/cat_13/v_-Sp-gOEGy2o_1391_1945/0005s_00148_rgb.jpg", "mask_path": "assets/cat_13/v_-Sp-gOEGy2o_1391_1945/0005s_00148_mask.png", "mask_status": "available" }, { "frame": "00180", "timestamp_sec": 52.41, "offset_sec": 6, "rgb_path": "assets/cat_13/v_-Sp-gOEGy2o_1391_1945/0006s_00180_rgb.jpg", "mask_path": "assets/cat_13/v_-Sp-gOEGy2o_1391_1945/0006s_00180_mask.png", "mask_status": "available" }, { "frame": "00208", "timestamp_sec": 53.41, "offset_sec": 7, "rgb_path": "assets/cat_13/v_-Sp-gOEGy2o_1391_1945/0007s_00208_rgb.jpg", "mask_path": "assets/cat_13/v_-Sp-gOEGy2o_1391_1945/0007s_00208_mask.png", "mask_status": "available" }, { "frame": "00240", "timestamp_sec": 54.41, "offset_sec": 8, "rgb_path": "assets/cat_13/v_-Sp-gOEGy2o_1391_1945/0008s_00240_rgb.jpg", "mask_path": "assets/cat_13/v_-Sp-gOEGy2o_1391_1945/0008s_00240_mask.png", "mask_status": "available" }, { "frame": "00268", "timestamp_sec": 55.41, "offset_sec": 9, "rgb_path": "assets/cat_13/v_-Sp-gOEGy2o_1391_1945/0009s_00268_rgb.jpg", "mask_path": "assets/cat_13/v_-Sp-gOEGy2o_1391_1945/0009s_00268_mask.png", "mask_status": "available" }, { "frame": "00300", "timestamp_sec": 56.41, "offset_sec": 10, "rgb_path": "assets/cat_13/v_-Sp-gOEGy2o_1391_1945/0010s_00300_rgb.jpg", "mask_path": "assets/cat_13/v_-Sp-gOEGy2o_1391_1945/0010s_00300_mask.png", "mask_status": "available" }, { "frame": "00328", "timestamp_sec": 57.41, "offset_sec": 11, "rgb_path": "assets/cat_13/v_-Sp-gOEGy2o_1391_1945/0011s_00328_rgb.jpg", "mask_path": "assets/cat_13/v_-Sp-gOEGy2o_1391_1945/0011s_00328_mask.png", "mask_status": "available" }, { "frame": "00360", "timestamp_sec": 58.41, "offset_sec": 12, "rgb_path": "assets/cat_13/v_-Sp-gOEGy2o_1391_1945/0012s_00360_rgb.jpg", "mask_path": "assets/cat_13/v_-Sp-gOEGy2o_1391_1945/0012s_00360_mask.png", "mask_status": "available" }, { "frame": "00388", "timestamp_sec": 59.41, "offset_sec": 13, "rgb_path": "assets/cat_13/v_-Sp-gOEGy2o_1391_1945/0013s_00388_rgb.jpg", "mask_path": "assets/cat_13/v_-Sp-gOEGy2o_1391_1945/0013s_00388_mask.png", "mask_status": "available" }, { "frame": "00420", "timestamp_sec": 60.41, "offset_sec": 14, "rgb_path": "assets/cat_13/v_-Sp-gOEGy2o_1391_1945/0014s_00420_rgb.jpg", "mask_path": "assets/cat_13/v_-Sp-gOEGy2o_1391_1945/0014s_00420_mask.png", "mask_status": "available" }, { "frame": "00448", "timestamp_sec": 61.41, "offset_sec": 15, "rgb_path": "assets/cat_13/v_-Sp-gOEGy2o_1391_1945/0015s_00448_rgb.jpg", "mask_path": "assets/cat_13/v_-Sp-gOEGy2o_1391_1945/0015s_00448_mask.png", "mask_status": "available" }, { "frame": "00480", "timestamp_sec": 62.41, "offset_sec": 16, "rgb_path": "assets/cat_13/v_-Sp-gOEGy2o_1391_1945/0016s_00480_rgb.jpg", "mask_path": "assets/cat_13/v_-Sp-gOEGy2o_1391_1945/0016s_00480_mask.png", "mask_status": "available" }, { "frame": "00508", "timestamp_sec": 63.41, "offset_sec": 17, "rgb_path": "assets/cat_13/v_-Sp-gOEGy2o_1391_1945/0017s_00508_rgb.jpg", "mask_path": "assets/cat_13/v_-Sp-gOEGy2o_1391_1945/0017s_00508_mask.png", "mask_status": "available" }, { "frame": "00540", "timestamp_sec": 64.41, "offset_sec": 18, "rgb_path": "assets/cat_13/v_-Sp-gOEGy2o_1391_1945/0018s_00540_rgb.jpg", "mask_path": "assets/cat_13/v_-Sp-gOEGy2o_1391_1945/0018s_00540_mask.png", "mask_status": "available" } ], "obj_id": [ 0 ], "anno_id": [ 69 ], "fps_inferred_from_clip_interval": 29.9459 }, "usage": { "input_tokens": 58324, "input_tokens_details": { "cache_write_tokens": 58321, "cached_tokens": 0 }, "output_tokens": 3428, "output_tokens_details": { "reasoning_tokens": 2070 }, "total_tokens": 61752 }, "model": "gpt-5.6-sol", "effort": "medium", "custom_id": "cat13_cap06_v_-Sp-gOEGy2o_1391_1945", "annotation": { "caption": "He picks up a curved piece of bark with his right hand and places it on the smoking kindling. He continues to add more small pieces of wood as the flame begins to grow.", "entity_groups": [ { "id": "entity_1", "label": "person", "spans": [ { "text": "He", "range": [ 0, 2 ] }, { "text": "his", "range": [ 40, 43 ] }, { "text": "He", "range": [ 94, 96 ] } ], "source_mask_link": "selected_person" }, { "id": "entity_2", "label": "piece of bark", "spans": [ { "text": "a curved piece of bark", "range": [ 12, 34 ] }, { "text": "it", "range": [ 66, 68 ] } ], "source_mask_link": null }, { "id": "entity_3", "label": "right hand", "spans": [ { "text": "his right hand", "range": [ 40, 54 ] } ], "source_mask_link": null }, { "id": "entity_4", "label": "kindling", "spans": [ { "text": "the smoking kindling", "range": [ 72, 92 ] } ], "source_mask_link": null }, { "id": "entity_5", "label": "pieces of wood", "spans": [ { "text": "more small pieces of wood", "range": [ 114, 139 ] } ], "source_mask_link": null }, { "id": "entity_6", "label": "flame", "spans": [ { "text": "the flame", "range": [ 143, 152 ] } ], "source_mask_link": null } ], "actions": [ { "id": "action_1", "mentions": [ { "text": "picks up", "ranges": [ [ 3, 11 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "type": "phrasal", "subject_refs": [ "entity_1" ], "object_refs": [ "entity_2" ], "confidence": "high", "temporal_status": "localized", "intervals": [ { "start_s": 1.0, "end_s": 3.0, "source_start_s": 47.41, "source_end_s": 49.41 } ], "evidence": "At 1.0 seconds the masked man reaches toward the curved bark; at 2.0 seconds he has lifted it from the stone.", "mask_relation": "subject matches the supplied selected-person mask", "source_subject_refs": [ "entity_1" ] }, { "id": "action_2", "mentions": [ { "text": "places", "ranges": [ [ 59, 65 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "type": "simple", "subject_refs": [ "entity_1" ], "object_refs": [ "entity_2" ], "confidence": "high", "temporal_status": "localized", "intervals": [ { "start_s": 3.0, "end_s": 7.0, "source_start_s": 49.41, "source_end_s": 53.41 } ], "evidence": "From 3.0 through 6.0 seconds, the masked man lowers and positions the long curved bark over the kindling; by 7.0 seconds it has left his hand.", "mask_relation": "subject matches the supplied selected-person mask", "source_subject_refs": [ "entity_1" ] }, { "id": "action_3", "mentions": [ { "text": "add", "ranges": [ [ 110, 113 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "type": "simple", "subject_refs": [ "entity_1" ], "object_refs": [ "entity_5" ], "confidence": "high", "temporal_status": "localized", "intervals": [ { "start_s": 7.0, "end_s": 15.0, "source_start_s": 53.41, "source_end_s": 61.41 } ], "evidence": "Between 7.0 and 14.0 seconds, the masked man repeatedly handles thin wood pieces and moves them into the kindling pile; his hand withdraws by 15.0 seconds.", "mask_relation": "subject matches the supplied selected-person mask", "source_subject_refs": [ "entity_1" ] } ], "notes": [ "The supplied masklet consistently identifies the sole visible man and therefore links to entity_1.", "The hand is annotated as a separate body-part entity, but the supplied mask segments the person as a whole rather than providing an independent hand mask.", "No separate source masks are supplied for the bark, hand, kindling, wood pieces, or flame.", "The captioned process \"smoking\" has the unmasked kindling as its subject, so it is not included in actions.", "The captioned process \"grow\" has the unmasked flame as its subject, so it is not included in actions.", "\"continues\" and \"begins\" are excluded as aspectual predicates; the corresponding core actions are \"add\" and \"grow\"." ] }, "model_output_raw": { "caption": "He picks up a curved piece of bark with his right hand and places it on the smoking kindling. He continues to add more small pieces of wood as the flame begins to grow.", "entity_groups": [ { "id": "entity_1", "label": "person", "spans": [ { "text": "He", "range": [ 0, 2 ] }, { "text": "his", "range": [ 40, 43 ] }, { "text": "He", "range": [ 94, 96 ] } ], "source_mask_link": "selected_person_masklet" }, { "id": "entity_2", "label": "piece of bark", "spans": [ { "text": "a curved piece of bark", "range": [ 12, 34 ] }, { "text": "it", "range": [ 66, 68 ] } ], "source_mask_link": null }, { "id": "entity_3", "label": "right hand", "spans": [ { "text": "his right hand", "range": [ 40, 54 ] } ], "source_mask_link": null }, { "id": "entity_4", "label": "kindling", "spans": [ { "text": "the smoking kindling", "range": [ 72, 92 ] } ], "source_mask_link": null }, { "id": "entity_5", "label": "pieces of wood", "spans": [ { "text": "more small pieces of wood", "range": [ 114, 139 ] } ], "source_mask_link": null }, { "id": "entity_6", "label": "flame", "spans": [ { "text": "the flame", "range": [ 143, 152 ] } ], "source_mask_link": null } ], "actions": [ { "id": "action_1", "mentions": [ { "text": "picks up", "ranges": [ [ 3, 11 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "type": "phrasal", "subject_refs": [ "entity_1" ], "object_refs": [ "entity_2" ], "confidence": "high", "mask_relation": "subject matches the supplied selected-person mask", "temporal_status": "localized", "intervals": [ { "start_s": 1.0, "end_s": 3.0, "source_start_s": 47.41, "source_end_s": 49.41 } ], "evidence": "At 1.0 seconds the masked man reaches toward the curved bark; at 2.0 seconds he has lifted it from the stone." }, { "id": "action_2", "mentions": [ { "text": "places", "ranges": [ [ 59, 65 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "type": "simple", "subject_refs": [ "entity_1" ], "object_refs": [ "entity_2" ], "target_refs": [ "entity_4" ], "confidence": "high", "mask_relation": "subject matches the supplied selected-person mask", "temporal_status": "localized", "intervals": [ { "start_s": 3.0, "end_s": 7.0, "source_start_s": 49.41, "source_end_s": 53.41 } ], "evidence": "From 3.0 through 6.0 seconds, the masked man lowers and positions the long curved bark over the kindling; by 7.0 seconds it has left his hand." }, { "id": "action_3", "mentions": [ { "text": "add", "ranges": [ [ 110, 113 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "type": "simple", "subject_refs": [ "entity_1" ], "object_refs": [ "entity_5" ], "confidence": "high", "mask_relation": "subject matches the supplied selected-person mask", "temporal_status": "localized", "intervals": [ { "start_s": 7.0, "end_s": 15.0, "source_start_s": 53.41, "source_end_s": 61.41 } ], "evidence": "Between 7.0 and 14.0 seconds, the masked man repeatedly handles thin wood pieces and moves them into the kindling pile; his hand withdraws by 15.0 seconds." } ], "notes": [ "The supplied masklet consistently identifies the sole visible man and therefore links to entity_1.", "The hand is annotated as a separate body-part entity, but the supplied mask segments the person as a whole rather than providing an independent hand mask.", "No separate source masks are supplied for the bark, hand, kindling, wood pieces, or flame.", "The captioned process \"smoking\" has the unmasked kindling as its subject, so it is not included in actions.", "The captioned process \"grow\" has the unmasked flame as its subject, so it is not included in actions.", "\"continues\" and \"begins\" are excluded as aspectual predicates; the corresponding core actions are \"add\" and \"grow\"." ] }, "normalization_repairs": 0 }, { "source_video_id": "v_-Sp-gOEGy2o", "caption_count_category": 13, "clip": { "caption_index": 7, "caption": "While holding the blower in his left hand, the man uses his right hand to pick up a larger split log. He carefully positions and places the log on top of the burning kindling.", "source_interval_sec": [ 64.91, 85.41 ], "loom_clip_id": "v_-Sp-gOEGy2o_1945_2560", "clip_frame_interval_4fps": [ 1945, 2560 ], "sample_count_1fps": 21, "sampled_frames": [ { "frame": "00000", "timestamp_sec": 64.91, "offset_sec": 0, "rgb_path": "assets/cat_13/v_-Sp-gOEGy2o_1945_2560/0000s_00000_rgb.jpg", "mask_path": "assets/cat_13/v_-Sp-gOEGy2o_1945_2560/0000s_00000_mask.png", "mask_status": "available" }, { "frame": "00028", "timestamp_sec": 65.91, "offset_sec": 1, "rgb_path": "assets/cat_13/v_-Sp-gOEGy2o_1945_2560/0001s_00028_rgb.jpg", "mask_path": "assets/cat_13/v_-Sp-gOEGy2o_1945_2560/0001s_00028_mask.png", "mask_status": "available" }, { "frame": "00060", "timestamp_sec": 66.91, "offset_sec": 2, "rgb_path": "assets/cat_13/v_-Sp-gOEGy2o_1945_2560/0002s_00060_rgb.jpg", "mask_path": "assets/cat_13/v_-Sp-gOEGy2o_1945_2560/0002s_00060_mask.png", "mask_status": "available" }, { "frame": "00088", "timestamp_sec": 67.91, "offset_sec": 3, "rgb_path": "assets/cat_13/v_-Sp-gOEGy2o_1945_2560/0003s_00088_rgb.jpg", "mask_path": "assets/cat_13/v_-Sp-gOEGy2o_1945_2560/0003s_00088_mask.png", "mask_status": "available" }, { "frame": "00120", "timestamp_sec": 68.91, "offset_sec": 4, "rgb_path": "assets/cat_13/v_-Sp-gOEGy2o_1945_2560/0004s_00120_rgb.jpg", "mask_path": "assets/cat_13/v_-Sp-gOEGy2o_1945_2560/0004s_00120_mask.png", "mask_status": "available" }, { "frame": "00148", "timestamp_sec": 69.91, "offset_sec": 5, "rgb_path": "assets/cat_13/v_-Sp-gOEGy2o_1945_2560/0005s_00148_rgb.jpg", "mask_path": "assets/cat_13/v_-Sp-gOEGy2o_1945_2560/0005s_00148_mask.png", "mask_status": "available" }, { "frame": "00180", "timestamp_sec": 70.91, "offset_sec": 6, "rgb_path": "assets/cat_13/v_-Sp-gOEGy2o_1945_2560/0006s_00180_rgb.jpg", "mask_path": "assets/cat_13/v_-Sp-gOEGy2o_1945_2560/0006s_00180_mask.png", "mask_status": "available" }, { "frame": "00208", "timestamp_sec": 71.91, "offset_sec": 7, "rgb_path": "assets/cat_13/v_-Sp-gOEGy2o_1945_2560/0007s_00208_rgb.jpg", "mask_path": "assets/cat_13/v_-Sp-gOEGy2o_1945_2560/0007s_00208_mask.png", "mask_status": "available" }, { "frame": "00240", "timestamp_sec": 72.91, "offset_sec": 8, "rgb_path": "assets/cat_13/v_-Sp-gOEGy2o_1945_2560/0008s_00240_rgb.jpg", "mask_path": "assets/cat_13/v_-Sp-gOEGy2o_1945_2560/0008s_00240_mask.png", "mask_status": "available" }, { "frame": "00268", "timestamp_sec": 73.91, "offset_sec": 9, "rgb_path": "assets/cat_13/v_-Sp-gOEGy2o_1945_2560/0009s_00268_rgb.jpg", "mask_path": "assets/cat_13/v_-Sp-gOEGy2o_1945_2560/0009s_00268_mask.png", "mask_status": "available" }, { "frame": "00300", "timestamp_sec": 74.91, "offset_sec": 10, "rgb_path": "assets/cat_13/v_-Sp-gOEGy2o_1945_2560/0010s_00300_rgb.jpg", "mask_path": "assets/cat_13/v_-Sp-gOEGy2o_1945_2560/0010s_00300_mask.png", "mask_status": "available" }, { "frame": "00328", "timestamp_sec": 75.91, "offset_sec": 11, "rgb_path": "assets/cat_13/v_-Sp-gOEGy2o_1945_2560/0011s_00328_rgb.jpg", "mask_path": "assets/cat_13/v_-Sp-gOEGy2o_1945_2560/0011s_00328_mask.png", "mask_status": "available" }, { "frame": "00360", "timestamp_sec": 76.91, "offset_sec": 12, "rgb_path": "assets/cat_13/v_-Sp-gOEGy2o_1945_2560/0012s_00360_rgb.jpg", "mask_path": "assets/cat_13/v_-Sp-gOEGy2o_1945_2560/0012s_00360_mask.png", "mask_status": "available" }, { "frame": "00388", "timestamp_sec": 77.91, "offset_sec": 13, "rgb_path": "assets/cat_13/v_-Sp-gOEGy2o_1945_2560/0013s_00388_rgb.jpg", "mask_path": "assets/cat_13/v_-Sp-gOEGy2o_1945_2560/0013s_00388_mask.png", "mask_status": "available" }, { "frame": "00420", "timestamp_sec": 78.91, "offset_sec": 14, "rgb_path": "assets/cat_13/v_-Sp-gOEGy2o_1945_2560/0014s_00420_rgb.jpg", "mask_path": "assets/cat_13/v_-Sp-gOEGy2o_1945_2560/0014s_00420_mask.png", "mask_status": "available" }, { "frame": "00448", "timestamp_sec": 79.91, "offset_sec": 15, "rgb_path": "assets/cat_13/v_-Sp-gOEGy2o_1945_2560/0015s_00448_rgb.jpg", "mask_path": "assets/cat_13/v_-Sp-gOEGy2o_1945_2560/0015s_00448_mask.png", "mask_status": "available" }, { "frame": "00480", "timestamp_sec": 80.91, "offset_sec": 16, "rgb_path": "assets/cat_13/v_-Sp-gOEGy2o_1945_2560/0016s_00480_rgb.jpg", "mask_path": "assets/cat_13/v_-Sp-gOEGy2o_1945_2560/0016s_00480_mask.png", "mask_status": "available" }, { "frame": "00508", "timestamp_sec": 81.91, "offset_sec": 17, "rgb_path": "assets/cat_13/v_-Sp-gOEGy2o_1945_2560/0017s_00508_rgb.jpg", "mask_path": "assets/cat_13/v_-Sp-gOEGy2o_1945_2560/0017s_00508_mask.png", "mask_status": "available" }, { "frame": "00540", "timestamp_sec": 82.91, "offset_sec": 18, "rgb_path": "assets/cat_13/v_-Sp-gOEGy2o_1945_2560/0018s_00540_rgb.jpg", "mask_path": "assets/cat_13/v_-Sp-gOEGy2o_1945_2560/0018s_00540_mask.png", "mask_status": "available" }, { "frame": "00568", "timestamp_sec": 83.91, "offset_sec": 19, "rgb_path": "assets/cat_13/v_-Sp-gOEGy2o_1945_2560/0019s_00568_rgb.jpg", "mask_path": "assets/cat_13/v_-Sp-gOEGy2o_1945_2560/0019s_00568_mask.png", "mask_status": "available" }, { "frame": "00600", "timestamp_sec": 84.91, "offset_sec": 20, "rgb_path": "assets/cat_13/v_-Sp-gOEGy2o_1945_2560/0020s_00600_rgb.jpg", "mask_path": "assets/cat_13/v_-Sp-gOEGy2o_1945_2560/0020s_00600_mask.png", "mask_status": "available" } ], "obj_id": [ 0 ], "anno_id": [ 71 ], "fps_inferred_from_clip_interval": 30.0 }, "usage": { "input_tokens": 62851, "input_tokens_details": { "cache_write_tokens": 62848, "cached_tokens": 0 }, "output_tokens": 3544, "output_tokens_details": { "reasoning_tokens": 1853 }, "total_tokens": 66395 }, "model": "gpt-5.6-sol", "effort": "medium", "custom_id": "cat13_cap07_v_-Sp-gOEGy2o_1945_2560", "annotation": { "caption": "While holding the blower in his left hand, the man uses his right hand to pick up a larger split log. He carefully positions and places the log on top of the burning kindling.", "entity_groups": [ { "id": "person_1", "label": "man", "spans": [ { "text": "his", "range": [ 28, 31 ] }, { "text": "the man", "range": [ 43, 50 ] }, { "text": "his", "range": [ 56, 59 ] }, { "text": "He", "range": [ 102, 104 ] } ], "source_mask_link": "selected_person" }, { "id": "blower_1", "label": "blower", "spans": [ { "text": "the blower", "range": [ 14, 24 ] } ], "source_mask_link": null }, { "id": "left_hand_1", "label": "left hand", "spans": [ { "text": "his left hand", "range": [ 28, 41 ] } ], "source_mask_link": null }, { "id": "right_hand_1", "label": "right hand", "spans": [ { "text": "his right hand", "range": [ 56, 70 ] } ], "source_mask_link": null }, { "id": "log_1", "label": "split log", "spans": [ { "text": "a larger split log", "range": [ 82, 100 ] }, { "text": "the log", "range": [ 136, 143 ] } ], "source_mask_link": null }, { "id": "kindling_1", "label": "kindling", "spans": [ { "text": "the burning kindling", "range": [ 154, 174 ] } ], "source_mask_link": null } ], "actions": [ { "id": "action_1", "mentions": [ { "text": "holding", "ranges": [ [ 6, 13 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "type": "simple", "subject_refs": [ "person_1" ], "object_refs": [ "blower_1" ], "confidence": "high", "temporal_status": "localized", "intervals": [ { "start_s": 0.0, "end_s": 20.5, "source_start_s": 64.91, "source_end_s": 85.41 } ], "evidence": "Across the sampled frames from 0.000 through 20.000 seconds, the selected man continuously holds the black blower in his left hand.", "mask_relation": "subject_matches_selected_person_mask", "source_subject_refs": "person_1" }, { "id": "action_2", "mentions": [ { "text": "uses", "ranges": [ [ 51, 55 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "type": "simple", "subject_refs": [ "person_1" ], "object_refs": [ "right_hand_1" ], "confidence": "medium", "temporal_status": "localized", "intervals": [ { "start_s": 16.0, "end_s": 19.0, "source_start_s": 80.91, "source_end_s": 83.91 } ], "evidence": "At 16.000–18.000 seconds, the selected man visibly uses his right hand to grasp and lift the large split log.", "mask_relation": "subject_matches_selected_person_mask", "source_subject_refs": "person_1" }, { "id": "action_3", "mentions": [ { "text": "pick up", "ranges": [ [ 74, 81 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "type": "phrasal", "subject_refs": [ "person_1" ], "object_refs": [ "log_1" ], "confidence": "high", "temporal_status": "localized", "intervals": [ { "start_s": 16.0, "end_s": 19.0, "source_start_s": 80.91, "source_end_s": 83.91 } ], "evidence": "The right hand grasps the large log at 16.000 seconds; the log is raised and carried above the fire by 17.000–18.000 seconds.", "mask_relation": "subject_matches_selected_person_mask", "source_subject_refs": "person_1" }, { "id": "action_4", "mentions": [ { "text": "positions", "ranges": [ [ 115, 124 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "type": "simple", "subject_refs": [ "person_1" ], "object_refs": [ "log_1" ], "confidence": "high", "temporal_status": "localized", "intervals": [ { "start_s": 18.0, "end_s": 20.5, "source_start_s": 82.91, "source_end_s": 85.41 } ], "evidence": "At 18.000–20.000 seconds, the selected man rotates and aligns the large log horizontally over the burning kindling.", "mask_relation": "subject_matches_selected_person_mask", "source_subject_refs": "person_1" }, { "id": "action_5", "mentions": [ { "text": "places", "ranges": [ [ 129, 135 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "type": "simple", "subject_refs": [ "person_1" ], "object_refs": [ "log_1" ], "confidence": "high", "temporal_status": "localized", "intervals": [ { "start_s": 19.0, "end_s": 20.5, "source_start_s": 83.91, "source_end_s": 85.41 } ], "evidence": "At 19.000–20.000 seconds, the selected man lowers the horizontally oriented log onto the wood above the kindling; the supported action reaches the clip boundary.", "mask_relation": "subject_matches_selected_person_mask", "source_subject_refs": "person_1" } ], "notes": [ "The supplied mask consistently identifies the man; no separate source masks are supplied for the blower, hands, log, or kindling.", "The process expressed by \"burning\" is not included because its subject is the unmasked kindling rather than the selected masked person.", "Temporal intervals are sparse clip-relative support intervals based on the sampled clean RGB frames. The clip corresponds to source-video interval [64.910000, 85.410000)." ] }, "model_output_raw": { "caption": "While holding the blower in his left hand, the man uses his right hand to pick up a larger split log. He carefully positions and places the log on top of the burning kindling.", "entity_groups": [ { "id": "person_1", "label": "man", "spans": [ { "start": 28, "end": 31, "text": "his" }, { "start": 43, "end": 50, "text": "the man" }, { "start": 56, "end": 59, "text": "his" }, { "start": 102, "end": 104, "text": "He" } ], "source_mask": "selected_person" }, { "id": "blower_1", "label": "blower", "spans": [ { "start": 14, "end": 24, "text": "the blower" } ], "source_mask": null }, { "id": "left_hand_1", "label": "left hand", "spans": [ { "start": 28, "end": 41, "text": "his left hand" } ], "source_mask": null }, { "id": "right_hand_1", "label": "right hand", "spans": [ { "start": 56, "end": 70, "text": "his right hand" } ], "source_mask": null }, { "id": "log_1", "label": "split log", "spans": [ { "start": 82, "end": 100, "text": "a larger split log" }, { "start": 136, "end": 143, "text": "the log" } ], "source_mask": null }, { "id": "kindling_1", "label": "kindling", "spans": [ { "start": 154, "end": 174, "text": "the burning kindling" } ], "source_mask": null } ], "actions": [ { "id": "action_1", "mentions": [ { "text": "holding", "ranges": [ [ 6, 13 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "type": "simple", "subject_ref": "person_1", "object_refs": [ "blower_1" ], "confidence": "high", "temporal_status": "localized", "intervals": [ { "start_s": 0.0, "end_s": 20.5 } ], "evidence": "Across the sampled frames from 0.000 through 20.000 seconds, the selected man continuously holds the black blower in his left hand.", "mask_relation": "subject_matches_selected_person_mask" }, { "id": "action_2", "mentions": [ { "text": "uses", "ranges": [ [ 51, 55 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "type": "simple", "subject_ref": "person_1", "object_refs": [ "right_hand_1" ], "confidence": "medium", "temporal_status": "localized", "intervals": [ { "start_s": 16.0, "end_s": 19.0 } ], "evidence": "At 16.000–18.000 seconds, the selected man visibly uses his right hand to grasp and lift the large split log.", "mask_relation": "subject_matches_selected_person_mask" }, { "id": "action_3", "mentions": [ { "text": "pick up", "ranges": [ [ 74, 81 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "type": "phrasal", "subject_ref": "person_1", "object_refs": [ "log_1" ], "confidence": "high", "temporal_status": "localized", "intervals": [ { "start_s": 16.0, "end_s": 19.0 } ], "evidence": "The right hand grasps the large log at 16.000 seconds; the log is raised and carried above the fire by 17.000–18.000 seconds.", "mask_relation": "subject_matches_selected_person_mask" }, { "id": "action_4", "mentions": [ { "text": "positions", "ranges": [ [ 115, 124 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "type": "simple", "subject_ref": "person_1", "object_refs": [ "log_1" ], "confidence": "high", "temporal_status": "localized", "intervals": [ { "start_s": 18.0, "end_s": 20.5 } ], "evidence": "At 18.000–20.000 seconds, the selected man rotates and aligns the large log horizontally over the burning kindling.", "mask_relation": "subject_matches_selected_person_mask" }, { "id": "action_5", "mentions": [ { "text": "places", "ranges": [ [ 129, 135 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "type": "simple", "subject_ref": "person_1", "object_refs": [ "log_1" ], "confidence": "high", "temporal_status": "localized", "intervals": [ { "start_s": 19.0, "end_s": 20.5 } ], "evidence": "At 19.000–20.000 seconds, the selected man lowers the horizontally oriented log onto the wood above the kindling; the supported action reaches the clip boundary.", "mask_relation": "subject_matches_selected_person_mask" } ], "notes": [ "The supplied mask consistently identifies the man; no separate source masks are supplied for the blower, hands, log, or kindling.", "The process expressed by \"burning\" is not included because its subject is the unmasked kindling rather than the selected masked person.", "Temporal intervals are sparse clip-relative support intervals based on the sampled clean RGB frames. The clip corresponds to source-video interval [64.910000, 85.410000)." ] }, "normalization_repairs": 0 }, { "source_video_id": "v_-Sp-gOEGy2o", "caption_count_category": 13, "clip": { "caption_index": 8, "caption": "The man picks up a long, flat piece of wood with both hands. He slowly lowers it and places it across the top of the other logs in the fire pit, building up the structure.", "source_interval_sec": [ 85.41, 109.41 ], "loom_clip_id": "v_-Sp-gOEGy2o_2560_3279", "clip_frame_interval_4fps": [ 2560, 3279 ], "sample_count_1fps": 24, "sampled_frames": [ { "frame": "00000", "timestamp_sec": 85.41, "offset_sec": 0, "rgb_path": "assets/cat_13/v_-Sp-gOEGy2o_2560_3279/0000s_00000_rgb.jpg", "mask_path": "assets/cat_13/v_-Sp-gOEGy2o_2560_3279/0000s_00000_mask.png", "mask_status": "available" }, { "frame": "00028", "timestamp_sec": 86.41, "offset_sec": 1, "rgb_path": "assets/cat_13/v_-Sp-gOEGy2o_2560_3279/0001s_00028_rgb.jpg", "mask_path": "assets/cat_13/v_-Sp-gOEGy2o_2560_3279/0001s_00028_mask.png", "mask_status": "available" }, { "frame": "00060", "timestamp_sec": 87.41, "offset_sec": 2, "rgb_path": "assets/cat_13/v_-Sp-gOEGy2o_2560_3279/0002s_00060_rgb.jpg", "mask_path": "assets/cat_13/v_-Sp-gOEGy2o_2560_3279/0002s_00060_mask.png", "mask_status": "available" }, { "frame": "00088", "timestamp_sec": 88.41, "offset_sec": 3, "rgb_path": "assets/cat_13/v_-Sp-gOEGy2o_2560_3279/0003s_00088_rgb.jpg", "mask_path": "assets/cat_13/v_-Sp-gOEGy2o_2560_3279/0003s_00088_mask.png", "mask_status": "available" }, { "frame": "00120", "timestamp_sec": 89.41, "offset_sec": 4, "rgb_path": "assets/cat_13/v_-Sp-gOEGy2o_2560_3279/0004s_00120_rgb.jpg", "mask_path": "assets/cat_13/v_-Sp-gOEGy2o_2560_3279/0004s_00120_mask.png", "mask_status": "available" }, { "frame": "00148", "timestamp_sec": 90.41, "offset_sec": 5, "rgb_path": "assets/cat_13/v_-Sp-gOEGy2o_2560_3279/0005s_00148_rgb.jpg", "mask_path": "assets/cat_13/v_-Sp-gOEGy2o_2560_3279/0005s_00148_mask.png", "mask_status": "available" }, { "frame": "00180", "timestamp_sec": 91.41, "offset_sec": 6, "rgb_path": "assets/cat_13/v_-Sp-gOEGy2o_2560_3279/0006s_00180_rgb.jpg", "mask_path": "assets/cat_13/v_-Sp-gOEGy2o_2560_3279/0006s_00180_mask.png", "mask_status": "available" }, { "frame": "00208", "timestamp_sec": 92.41, "offset_sec": 7, "rgb_path": "assets/cat_13/v_-Sp-gOEGy2o_2560_3279/0007s_00208_rgb.jpg", "mask_path": "assets/cat_13/v_-Sp-gOEGy2o_2560_3279/0007s_00208_mask.png", "mask_status": "available" }, { "frame": "00240", "timestamp_sec": 93.41, "offset_sec": 8, "rgb_path": "assets/cat_13/v_-Sp-gOEGy2o_2560_3279/0008s_00240_rgb.jpg", "mask_path": "assets/cat_13/v_-Sp-gOEGy2o_2560_3279/0008s_00240_mask.png", "mask_status": "available" }, { "frame": "00268", "timestamp_sec": 94.41, "offset_sec": 9, "rgb_path": "assets/cat_13/v_-Sp-gOEGy2o_2560_3279/0009s_00268_rgb.jpg", "mask_path": "assets/cat_13/v_-Sp-gOEGy2o_2560_3279/0009s_00268_mask.png", "mask_status": "available" }, { "frame": "00300", "timestamp_sec": 95.41, "offset_sec": 10, "rgb_path": "assets/cat_13/v_-Sp-gOEGy2o_2560_3279/0010s_00300_rgb.jpg", "mask_path": "assets/cat_13/v_-Sp-gOEGy2o_2560_3279/0010s_00300_mask.png", "mask_status": "available" }, { "frame": "00328", "timestamp_sec": 96.41, "offset_sec": 11, "rgb_path": "assets/cat_13/v_-Sp-gOEGy2o_2560_3279/0011s_00328_rgb.jpg", "mask_path": "assets/cat_13/v_-Sp-gOEGy2o_2560_3279/0011s_00328_mask.png", "mask_status": "available" }, { "frame": "00360", "timestamp_sec": 97.41, "offset_sec": 12, "rgb_path": "assets/cat_13/v_-Sp-gOEGy2o_2560_3279/0012s_00360_rgb.jpg", "mask_path": "assets/cat_13/v_-Sp-gOEGy2o_2560_3279/0012s_00360_mask.png", "mask_status": "available" }, { "frame": "00388", "timestamp_sec": 98.41, "offset_sec": 13, "rgb_path": "assets/cat_13/v_-Sp-gOEGy2o_2560_3279/0013s_00388_rgb.jpg", "mask_path": "assets/cat_13/v_-Sp-gOEGy2o_2560_3279/0013s_00388_mask.png", "mask_status": "available" }, { "frame": "00420", "timestamp_sec": 99.41, "offset_sec": 14, "rgb_path": "assets/cat_13/v_-Sp-gOEGy2o_2560_3279/0014s_00420_rgb.jpg", "mask_path": "assets/cat_13/v_-Sp-gOEGy2o_2560_3279/0014s_00420_mask.png", "mask_status": "available" }, { "frame": "00448", "timestamp_sec": 100.41, "offset_sec": 15, "rgb_path": "assets/cat_13/v_-Sp-gOEGy2o_2560_3279/0015s_00448_rgb.jpg", "mask_path": "assets/cat_13/v_-Sp-gOEGy2o_2560_3279/0015s_00448_mask.png", "mask_status": "available" }, { "frame": "00480", "timestamp_sec": 101.41, "offset_sec": 16, "rgb_path": "assets/cat_13/v_-Sp-gOEGy2o_2560_3279/0016s_00480_rgb.jpg", "mask_path": "assets/cat_13/v_-Sp-gOEGy2o_2560_3279/0016s_00480_mask.png", "mask_status": "available" }, { "frame": "00508", "timestamp_sec": 102.41, "offset_sec": 17, "rgb_path": "assets/cat_13/v_-Sp-gOEGy2o_2560_3279/0017s_00508_rgb.jpg", "mask_path": "assets/cat_13/v_-Sp-gOEGy2o_2560_3279/0017s_00508_mask.png", "mask_status": "available" }, { "frame": "00540", "timestamp_sec": 103.41, "offset_sec": 18, "rgb_path": "assets/cat_13/v_-Sp-gOEGy2o_2560_3279/0018s_00540_rgb.jpg", "mask_path": "assets/cat_13/v_-Sp-gOEGy2o_2560_3279/0018s_00540_mask.png", "mask_status": "available" }, { "frame": "00568", "timestamp_sec": 104.41, "offset_sec": 19, "rgb_path": "assets/cat_13/v_-Sp-gOEGy2o_2560_3279/0019s_00568_rgb.jpg", "mask_path": "assets/cat_13/v_-Sp-gOEGy2o_2560_3279/0019s_00568_mask.png", "mask_status": "available" }, { "frame": "00600", "timestamp_sec": 105.41, "offset_sec": 20, "rgb_path": "assets/cat_13/v_-Sp-gOEGy2o_2560_3279/0020s_00600_rgb.jpg", "mask_path": "assets/cat_13/v_-Sp-gOEGy2o_2560_3279/0020s_00600_mask.png", "mask_status": "available" }, { "frame": "00628", "timestamp_sec": 106.41, "offset_sec": 21, "rgb_path": "assets/cat_13/v_-Sp-gOEGy2o_2560_3279/0021s_00628_rgb.jpg", "mask_path": "assets/cat_13/v_-Sp-gOEGy2o_2560_3279/0021s_00628_mask.png", "mask_status": "available" }, { "frame": "00660", "timestamp_sec": 107.41, "offset_sec": 22, "rgb_path": "assets/cat_13/v_-Sp-gOEGy2o_2560_3279/0022s_00660_rgb.jpg", "mask_path": "assets/cat_13/v_-Sp-gOEGy2o_2560_3279/0022s_00660_mask.png", "mask_status": "available" }, { "frame": "00688", "timestamp_sec": 108.41, "offset_sec": 23, "rgb_path": "assets/cat_13/v_-Sp-gOEGy2o_2560_3279/0023s_00688_rgb.jpg", "mask_path": "assets/cat_13/v_-Sp-gOEGy2o_2560_3279/0023s_00688_mask.png", "mask_status": "available" } ], "obj_id": [ 0 ], "anno_id": [ 72 ], "fps_inferred_from_clip_interval": 29.9583 }, "usage": { "input_tokens": 69641, "input_tokens_details": { "cache_write_tokens": 69638, "cached_tokens": 0 }, "output_tokens": 3933, "output_tokens_details": { "reasoning_tokens": 2588 }, "total_tokens": 73574 }, "model": "gpt-5.6-sol", "effort": "medium", "custom_id": "cat13_cap08_v_-Sp-gOEGy2o_2560_3279", "annotation": { "caption": "The man picks up a long, flat piece of wood with both hands. He slowly lowers it and places it across the top of the other logs in the fire pit, building up the structure.", "entity_groups": [ { "id": "entity_1", "label": "man", "spans": [ { "text": "The man", "range": [ 0, 7 ] }, { "text": "He", "range": [ 61, 63 ] } ], "source_mask_link": "selected_person" }, { "id": "entity_2", "label": "piece of wood", "spans": [ { "text": "a long, flat piece of wood", "range": [ 17, 43 ] }, { "text": "it", "range": [ 78, 80 ] }, { "text": "it", "range": [ 92, 94 ] } ], "source_mask_link": null }, { "id": "entity_3", "label": "hands", "spans": [ { "text": "both hands", "range": [ 49, 59 ] } ], "source_mask_link": null }, { "id": "entity_4", "label": "logs", "spans": [ { "text": "the other logs", "range": [ 113, 127 ] } ], "source_mask_link": null }, { "id": "entity_5", "label": "fire pit", "spans": [ { "text": "the fire pit", "range": [ 131, 143 ] } ], "source_mask_link": null }, { "id": "entity_6", "label": "structure", "spans": [ { "text": "the structure", "range": [ 157, 170 ] } ], "source_mask_link": null } ], "actions": [ { "id": "action_1", "mentions": [ { "text": "picks up", "ranges": [ [ 8, 16 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "type": "phrasal", "subject_refs": [ "entity_1" ], "object_refs": [ "entity_2" ], "confidence": "high", "temporal_status": "uncertain", "intervals": [], "evidence": "The masked man is shown handling the wood, but no sampled frame sequence clearly isolates its initial pickup.", "mask_relation": "subject matches the supplied selected-person mask", "source_subject_refs": "entity_1" }, { "id": "action_2", "mentions": [ { "text": "lowers", "ranges": [ [ 71, 77 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "type": "simple", "subject_refs": [ "entity_1" ], "object_refs": [ "entity_2" ], "confidence": "high", "temporal_status": "localized", "intervals": [ { "start_s": 8.0, "end_s": 22.0, "source_start_s": 93.41, "source_end_s": 107.41 } ], "evidence": "From sampled times 8–21 seconds, the masked man grips the long piece as it is brought progressively closer to the supporting logs.", "mask_relation": "subject matches the supplied selected-person mask", "source_subject_refs": "entity_1" }, { "id": "action_3", "mentions": [ { "text": "places", "ranges": [ [ 85, 91 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "type": "simple", "subject_refs": [ "entity_1" ], "object_refs": [ "entity_2" ], "confidence": "high", "temporal_status": "localized", "intervals": [ { "start_s": 20.0, "end_s": 24.0, "source_start_s": 105.41, "source_end_s": 109.41 } ], "evidence": "At 20–23 seconds, the wood reaches and rests across the upper logs while the masked man's hands withdraw.", "mask_relation": "subject matches the supplied selected-person mask", "source_subject_refs": "entity_1" }, { "id": "action_4", "mentions": [ { "text": "building up", "ranges": [ [ 145, 156 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "type": "phrasal", "subject_refs": [ "entity_1" ], "object_refs": [ "entity_6" ], "confidence": "high", "temporal_status": "localized", "intervals": [ { "start_s": 20.0, "end_s": 24.0, "source_start_s": 105.41, "source_end_s": 109.41 } ], "evidence": "The final sampled frames show the masked man adding the long wood piece across the existing log arrangement.", "mask_relation": "subject matches the supplied selected-person mask", "source_subject_refs": "entity_1" } ], "notes": [ "The supplied masklet consistently identifies the man, including visible portions of his arms and hands.", "No independent source masks were supplied for the wood, hands, logs, fire pit, or assembled structure.", "The pickup itself is not clearly isolated by the sparse sampled frames, so it has no temporal interval.", "The final supported intervals end at the clip boundary of 24.0 seconds, corresponding to source-video time 109.410 seconds." ] }, "model_output_raw": { "caption": "The man picks up a long, flat piece of wood with both hands. He slowly lowers it and places it across the top of the other logs in the fire pit, building up the structure.", "entity_groups": [ { "id": "entity_1", "label": "man", "spans": [ { "text": "The man", "range": [ 0, 7 ] }, { "text": "He", "range": [ 62, 64 ] } ], "source_mask": "selected_person_mask" }, { "id": "entity_2", "label": "piece of wood", "spans": [ { "text": "a long, flat piece of wood", "range": [ 17, 44 ] }, { "text": "it", "range": [ 79, 81 ] }, { "text": "it", "range": [ 93, 95 ] } ], "source_mask": null }, { "id": "entity_3", "label": "hands", "spans": [ { "text": "both hands", "range": [ 50, 60 ] } ], "source_mask": null }, { "id": "entity_4", "label": "logs", "spans": [ { "text": "the other logs", "range": [ 114, 128 ] } ], "source_mask": null }, { "id": "entity_5", "label": "fire pit", "spans": [ { "text": "the fire pit", "range": [ 132, 144 ] } ], "source_mask": null }, { "id": "entity_6", "label": "structure", "spans": [ { "text": "the structure", "range": [ 158, 171 ] } ], "source_mask": null } ], "actions": [ { "id": "action_1", "mentions": [ { "text": "picks up", "ranges": [ [ 8, 16 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "type": "phrasal", "subject": "entity_1", "objects": [ "entity_2" ], "confidence": "high", "mask_relation": "subject matches the supplied selected-person mask", "temporal_status": "uncertain", "intervals": [], "evidence": "The masked man is shown handling the wood, but no sampled frame sequence clearly isolates its initial pickup." }, { "id": "action_2", "mentions": [ { "text": "lowers", "ranges": [ [ 72, 78 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "type": "simple", "subject": "entity_1", "objects": [ "entity_2" ], "confidence": "high", "mask_relation": "subject matches the supplied selected-person mask", "temporal_status": "localized", "intervals": [ { "start_s": 8.0, "end_s": 22.0 } ], "evidence": "From sampled times 8–21 seconds, the masked man grips the long piece as it is brought progressively closer to the supporting logs." }, { "id": "action_3", "mentions": [ { "text": "places", "ranges": [ [ 86, 92 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "type": "simple", "subject": "entity_1", "objects": [ "entity_2" ], "confidence": "high", "mask_relation": "subject matches the supplied selected-person mask", "temporal_status": "localized", "intervals": [ { "start_s": 20.0, "end_s": 24.0 } ], "evidence": "At 20–23 seconds, the wood reaches and rests across the upper logs while the masked man's hands withdraw." }, { "id": "action_4", "mentions": [ { "text": "building up", "ranges": [ [ 146, 157 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "type": "phrasal", "subject": "entity_1", "objects": [ "entity_6" ], "confidence": "high", "mask_relation": "subject matches the supplied selected-person mask", "temporal_status": "localized", "intervals": [ { "start_s": 20.0, "end_s": 24.0 } ], "evidence": "The final sampled frames show the masked man adding the long wood piece across the existing log arrangement." } ], "notes": [ "The supplied masklet consistently identifies the man, including visible portions of his arms and hands.", "No independent source masks were supplied for the wood, hands, logs, fire pit, or assembled structure.", "The pickup itself is not clearly isolated by the sparse sampled frames, so it has no temporal interval.", "The final supported intervals end at the clip boundary of 24.0 seconds, corresponding to source-video time 109.410 seconds." ] }, "normalization_repairs": 11 }, { "source_video_id": "v_-Sp-gOEGy2o", "caption_count_category": 13, "clip": { "caption_index": 9, "caption": "The man sits back, holding the black blower in his left hand while resting his right hand on the stone ledge. He looks down at the established fire and speaks.", "source_interval_sec": [ 109.41, 149.41 ], "loom_clip_id": "v_-Sp-gOEGy2o_3279_4478", "clip_frame_interval_4fps": [ 3279, 4478 ], "sample_count_1fps": 40, "sampled_frames": [ { "frame": "00000", "timestamp_sec": 109.41, "offset_sec": 0, "rgb_path": "assets/cat_13/v_-Sp-gOEGy2o_3279_4478/0000s_00000_rgb.jpg", "mask_path": "assets/cat_13/v_-Sp-gOEGy2o_3279_4478/0000s_00000_mask.png", "mask_status": "available" }, { "frame": "00028", "timestamp_sec": 110.41, "offset_sec": 1, "rgb_path": "assets/cat_13/v_-Sp-gOEGy2o_3279_4478/0001s_00028_rgb.jpg", "mask_path": "assets/cat_13/v_-Sp-gOEGy2o_3279_4478/0001s_00028_mask.png", "mask_status": "available" }, { "frame": "00060", "timestamp_sec": 111.41, "offset_sec": 2, "rgb_path": "assets/cat_13/v_-Sp-gOEGy2o_3279_4478/0002s_00060_rgb.jpg", "mask_path": "assets/cat_13/v_-Sp-gOEGy2o_3279_4478/0002s_00060_mask.png", "mask_status": "available" }, { "frame": "00088", "timestamp_sec": 112.41, "offset_sec": 3, "rgb_path": "assets/cat_13/v_-Sp-gOEGy2o_3279_4478/0003s_00088_rgb.jpg", "mask_path": "assets/cat_13/v_-Sp-gOEGy2o_3279_4478/0003s_00088_mask.png", "mask_status": "available" }, { "frame": "00120", "timestamp_sec": 113.41, "offset_sec": 4, "rgb_path": "assets/cat_13/v_-Sp-gOEGy2o_3279_4478/0004s_00120_rgb.jpg", "mask_path": "assets/cat_13/v_-Sp-gOEGy2o_3279_4478/0004s_00120_mask.png", "mask_status": "available" }, { "frame": "00148", "timestamp_sec": 114.41, "offset_sec": 5, "rgb_path": "assets/cat_13/v_-Sp-gOEGy2o_3279_4478/0005s_00148_rgb.jpg", "mask_path": "assets/cat_13/v_-Sp-gOEGy2o_3279_4478/0005s_00148_mask.png", "mask_status": "available" }, { "frame": "00180", "timestamp_sec": 115.41, "offset_sec": 6, "rgb_path": "assets/cat_13/v_-Sp-gOEGy2o_3279_4478/0006s_00180_rgb.jpg", "mask_path": "assets/cat_13/v_-Sp-gOEGy2o_3279_4478/0006s_00180_mask.png", "mask_status": "available" }, { "frame": "00208", "timestamp_sec": 116.41, "offset_sec": 7, "rgb_path": "assets/cat_13/v_-Sp-gOEGy2o_3279_4478/0007s_00208_rgb.jpg", "mask_path": "assets/cat_13/v_-Sp-gOEGy2o_3279_4478/0007s_00208_mask.png", "mask_status": "available" }, { "frame": "00240", "timestamp_sec": 117.41, "offset_sec": 8, "rgb_path": "assets/cat_13/v_-Sp-gOEGy2o_3279_4478/0008s_00240_rgb.jpg", "mask_path": "assets/cat_13/v_-Sp-gOEGy2o_3279_4478/0008s_00240_mask.png", "mask_status": "available" }, { "frame": "00268", "timestamp_sec": 118.41, "offset_sec": 9, "rgb_path": "assets/cat_13/v_-Sp-gOEGy2o_3279_4478/0009s_00268_rgb.jpg", "mask_path": "assets/cat_13/v_-Sp-gOEGy2o_3279_4478/0009s_00268_mask.png", "mask_status": "available" }, { "frame": "00300", "timestamp_sec": 119.41, "offset_sec": 10, "rgb_path": "assets/cat_13/v_-Sp-gOEGy2o_3279_4478/0010s_00300_rgb.jpg", "mask_path": "assets/cat_13/v_-Sp-gOEGy2o_3279_4478/0010s_00300_mask.png", "mask_status": "available" }, { "frame": "00328", "timestamp_sec": 120.41, "offset_sec": 11, "rgb_path": "assets/cat_13/v_-Sp-gOEGy2o_3279_4478/0011s_00328_rgb.jpg", "mask_path": "assets/cat_13/v_-Sp-gOEGy2o_3279_4478/0011s_00328_mask.png", "mask_status": "available" }, { "frame": "00360", "timestamp_sec": 121.41, "offset_sec": 12, "rgb_path": "assets/cat_13/v_-Sp-gOEGy2o_3279_4478/0012s_00360_rgb.jpg", "mask_path": "assets/cat_13/v_-Sp-gOEGy2o_3279_4478/0012s_00360_mask.png", "mask_status": "available" }, { "frame": "00388", "timestamp_sec": 122.41, "offset_sec": 13, "rgb_path": "assets/cat_13/v_-Sp-gOEGy2o_3279_4478/0013s_00388_rgb.jpg", "mask_path": "assets/cat_13/v_-Sp-gOEGy2o_3279_4478/0013s_00388_mask.png", "mask_status": "available" }, { "frame": "00420", "timestamp_sec": 123.41, "offset_sec": 14, "rgb_path": "assets/cat_13/v_-Sp-gOEGy2o_3279_4478/0014s_00420_rgb.jpg", "mask_path": "assets/cat_13/v_-Sp-gOEGy2o_3279_4478/0014s_00420_mask.png", "mask_status": "available" }, { "frame": "00448", "timestamp_sec": 124.41, "offset_sec": 15, "rgb_path": "assets/cat_13/v_-Sp-gOEGy2o_3279_4478/0015s_00448_rgb.jpg", "mask_path": "assets/cat_13/v_-Sp-gOEGy2o_3279_4478/0015s_00448_mask.png", "mask_status": "available" }, { "frame": "00480", "timestamp_sec": 125.41, "offset_sec": 16, "rgb_path": "assets/cat_13/v_-Sp-gOEGy2o_3279_4478/0016s_00480_rgb.jpg", "mask_path": "assets/cat_13/v_-Sp-gOEGy2o_3279_4478/0016s_00480_mask.png", "mask_status": "available" }, { "frame": "00508", "timestamp_sec": 126.41, "offset_sec": 17, "rgb_path": "assets/cat_13/v_-Sp-gOEGy2o_3279_4478/0017s_00508_rgb.jpg", "mask_path": "assets/cat_13/v_-Sp-gOEGy2o_3279_4478/0017s_00508_mask.png", "mask_status": "available" }, { "frame": "00540", "timestamp_sec": 127.41, "offset_sec": 18, "rgb_path": "assets/cat_13/v_-Sp-gOEGy2o_3279_4478/0018s_00540_rgb.jpg", "mask_path": "assets/cat_13/v_-Sp-gOEGy2o_3279_4478/0018s_00540_mask.png", "mask_status": "available" }, { "frame": "00568", "timestamp_sec": 128.41, "offset_sec": 19, "rgb_path": "assets/cat_13/v_-Sp-gOEGy2o_3279_4478/0019s_00568_rgb.jpg", "mask_path": "assets/cat_13/v_-Sp-gOEGy2o_3279_4478/0019s_00568_mask.png", "mask_status": "available" }, { "frame": "00600", "timestamp_sec": 129.41, "offset_sec": 20, "rgb_path": "assets/cat_13/v_-Sp-gOEGy2o_3279_4478/0020s_00600_rgb.jpg", "mask_path": "assets/cat_13/v_-Sp-gOEGy2o_3279_4478/0020s_00600_mask.png", "mask_status": "available" }, { "frame": "00628", "timestamp_sec": 130.41, "offset_sec": 21, "rgb_path": "assets/cat_13/v_-Sp-gOEGy2o_3279_4478/0021s_00628_rgb.jpg", "mask_path": "assets/cat_13/v_-Sp-gOEGy2o_3279_4478/0021s_00628_mask.png", "mask_status": "available" }, { "frame": "00660", "timestamp_sec": 131.41, "offset_sec": 22, "rgb_path": "assets/cat_13/v_-Sp-gOEGy2o_3279_4478/0022s_00660_rgb.jpg", "mask_path": "assets/cat_13/v_-Sp-gOEGy2o_3279_4478/0022s_00660_mask.png", "mask_status": "available" }, { "frame": "00688", "timestamp_sec": 132.41, "offset_sec": 23, "rgb_path": "assets/cat_13/v_-Sp-gOEGy2o_3279_4478/0023s_00688_rgb.jpg", "mask_path": "assets/cat_13/v_-Sp-gOEGy2o_3279_4478/0023s_00688_mask.png", "mask_status": "available" }, { "frame": "00720", "timestamp_sec": 133.41, "offset_sec": 24, "rgb_path": "assets/cat_13/v_-Sp-gOEGy2o_3279_4478/0024s_00720_rgb.jpg", "mask_path": "assets/cat_13/v_-Sp-gOEGy2o_3279_4478/0024s_00720_mask.png", "mask_status": "available" }, { "frame": "00748", "timestamp_sec": 134.41, "offset_sec": 25, "rgb_path": "assets/cat_13/v_-Sp-gOEGy2o_3279_4478/0025s_00748_rgb.jpg", "mask_path": "assets/cat_13/v_-Sp-gOEGy2o_3279_4478/0025s_00748_mask.png", "mask_status": "available" }, { "frame": "00780", "timestamp_sec": 135.41, "offset_sec": 26, "rgb_path": "assets/cat_13/v_-Sp-gOEGy2o_3279_4478/0026s_00780_rgb.jpg", "mask_path": "assets/cat_13/v_-Sp-gOEGy2o_3279_4478/0026s_00780_mask.png", "mask_status": "available" }, { "frame": "00808", "timestamp_sec": 136.41, "offset_sec": 27, "rgb_path": "assets/cat_13/v_-Sp-gOEGy2o_3279_4478/0027s_00808_rgb.jpg", "mask_path": "assets/cat_13/v_-Sp-gOEGy2o_3279_4478/0027s_00808_mask.png", "mask_status": "available" }, { "frame": "00840", "timestamp_sec": 137.41, "offset_sec": 28, "rgb_path": "assets/cat_13/v_-Sp-gOEGy2o_3279_4478/0028s_00840_rgb.jpg", "mask_path": "assets/cat_13/v_-Sp-gOEGy2o_3279_4478/0028s_00840_mask.png", "mask_status": "available" }, { "frame": "00868", "timestamp_sec": 138.41, "offset_sec": 29, "rgb_path": "assets/cat_13/v_-Sp-gOEGy2o_3279_4478/0029s_00868_rgb.jpg", "mask_path": "assets/cat_13/v_-Sp-gOEGy2o_3279_4478/0029s_00868_mask.png", "mask_status": "available" }, { "frame": "00900", "timestamp_sec": 139.41, "offset_sec": 30, "rgb_path": "assets/cat_13/v_-Sp-gOEGy2o_3279_4478/0030s_00900_rgb.jpg", "mask_path": "assets/cat_13/v_-Sp-gOEGy2o_3279_4478/0030s_00900_mask.png", "mask_status": "available" }, { "frame": "00928", "timestamp_sec": 140.41, "offset_sec": 31, "rgb_path": "assets/cat_13/v_-Sp-gOEGy2o_3279_4478/0031s_00928_rgb.jpg", "mask_path": "assets/cat_13/v_-Sp-gOEGy2o_3279_4478/0031s_00928_mask.png", "mask_status": "available" }, { "frame": "00960", "timestamp_sec": 141.41, "offset_sec": 32, "rgb_path": "assets/cat_13/v_-Sp-gOEGy2o_3279_4478/0032s_00960_rgb.jpg", "mask_path": "assets/cat_13/v_-Sp-gOEGy2o_3279_4478/0032s_00960_mask.png", "mask_status": "available" }, { "frame": "00988", "timestamp_sec": 142.41, "offset_sec": 33, "rgb_path": "assets/cat_13/v_-Sp-gOEGy2o_3279_4478/0033s_00988_rgb.jpg", "mask_path": "assets/cat_13/v_-Sp-gOEGy2o_3279_4478/0033s_00988_mask.png", "mask_status": "available" }, { "frame": "01020", "timestamp_sec": 143.41, "offset_sec": 34, "rgb_path": "assets/cat_13/v_-Sp-gOEGy2o_3279_4478/0034s_01020_rgb.jpg", "mask_path": "assets/cat_13/v_-Sp-gOEGy2o_3279_4478/0034s_01020_mask.png", "mask_status": "available" }, { "frame": "01048", "timestamp_sec": 144.41, "offset_sec": 35, "rgb_path": "assets/cat_13/v_-Sp-gOEGy2o_3279_4478/0035s_01048_rgb.jpg", "mask_path": "assets/cat_13/v_-Sp-gOEGy2o_3279_4478/0035s_01048_mask.png", "mask_status": "available" }, { "frame": "01080", "timestamp_sec": 145.41, "offset_sec": 36, "rgb_path": "assets/cat_13/v_-Sp-gOEGy2o_3279_4478/0036s_01080_rgb.jpg", "mask_path": "assets/cat_13/v_-Sp-gOEGy2o_3279_4478/0036s_01080_mask.png", "mask_status": "available" }, { "frame": "01108", "timestamp_sec": 146.41, "offset_sec": 37, "rgb_path": "assets/cat_13/v_-Sp-gOEGy2o_3279_4478/0037s_01108_rgb.jpg", "mask_path": "assets/cat_13/v_-Sp-gOEGy2o_3279_4478/0037s_01108_mask.png", "mask_status": "available" }, { "frame": "01140", "timestamp_sec": 147.41, "offset_sec": 38, "rgb_path": "assets/cat_13/v_-Sp-gOEGy2o_3279_4478/0038s_01140_rgb.jpg", "mask_path": "assets/cat_13/v_-Sp-gOEGy2o_3279_4478/0038s_01140_mask.png", "mask_status": "available" }, { "frame": "01168", "timestamp_sec": 148.41, "offset_sec": 39, "rgb_path": "assets/cat_13/v_-Sp-gOEGy2o_3279_4478/0039s_01168_rgb.jpg", "mask_path": "assets/cat_13/v_-Sp-gOEGy2o_3279_4478/0039s_01168_mask.png", "mask_status": "available" } ], "obj_id": [ 0 ], "anno_id": [ 73 ], "fps_inferred_from_clip_interval": 29.975 }, "usage": { "input_tokens": 105844, "input_tokens_details": { "cache_write_tokens": 105841, "cached_tokens": 0 }, "output_tokens": 3533, "output_tokens_details": { "reasoning_tokens": 1872 }, "total_tokens": 109377 }, "model": "gpt-5.6-sol", "effort": "medium", "custom_id": "cat13_cap09_v_-Sp-gOEGy2o_3279_4478", "annotation": { "caption": "The man sits back, holding the black blower in his left hand while resting his right hand on the stone ledge. He looks down at the established fire and speaks.", "entity_groups": [ { "id": "man_1", "label": "man", "spans": [ { "text": "The man", "range": [ 0, 7 ] }, { "text": "his", "range": [ 47, 50 ] }, { "text": "his", "range": [ 75, 78 ] }, { "text": "He", "range": [ 110, 112 ] } ], "source_mask_link": "selected_person" }, { "id": "blower_1", "label": "blower", "spans": [ { "text": "the black blower", "range": [ 27, 43 ] } ], "source_mask_link": null }, { "id": "left_hand_1", "label": "left hand", "spans": [ { "text": "his left hand", "range": [ 47, 60 ] } ], "source_mask_link": null }, { "id": "right_hand_1", "label": "right hand", "spans": [ { "text": "his right hand", "range": [ 75, 89 ] } ], "source_mask_link": null }, { "id": "stone_ledge_1", "label": "stone ledge", "spans": [ { "text": "the stone ledge", "range": [ 93, 108 ] } ], "source_mask_link": null }, { "id": "fire_1", "label": "fire", "spans": [ { "text": "the established fire", "range": [ 127, 147 ] } ], "source_mask_link": null } ], "actions": [ { "id": "action_1", "mentions": [ { "text": "sits", "ranges": [ [ 8, 12 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "type": "posture", "subject_refs": [ "man_1" ], "object_refs": [], "confidence": "high", "temporal_status": "localized", "intervals": [ { "start_s": 1.0, "end_s": 14.0, "source_start_s": 110.41, "source_end_s": 123.41 }, { "start_s": 32.0, "end_s": 40.0, "source_start_s": 141.41, "source_end_s": 149.41 } ], "evidence": "The mask-linked man is visibly seated beside the fire in the early body views and again in the final seated body and face views.", "mask_relation": "subject matches the supplied selected-person mask", "source_subject_refs": "man_1" }, { "id": "action_2", "mentions": [ { "text": "holding", "ranges": [ [ 19, 26 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "type": "simple", "subject_refs": [ "man_1" ], "object_refs": [ "blower_1", "left_hand_1" ], "confidence": "medium", "temporal_status": "localized", "intervals": [ { "start_s": 34.0, "end_s": 35.0, "source_start_s": 143.41, "source_end_s": 144.41 } ], "evidence": "At 34.0 seconds, the selected man's left hand is visibly positioned over and holding the black blower.", "mask_relation": "subject matches the supplied selected-person mask", "source_subject_refs": "man_1" }, { "id": "action_3", "mentions": [ { "text": "resting", "ranges": [ [ 67, 74 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "type": "posture", "subject_refs": [ "man_1" ], "object_refs": [ "right_hand_1", "stone_ledge_1" ], "confidence": "high", "temporal_status": "localized", "intervals": [ { "start_s": 0.0, "end_s": 2.0, "source_start_s": 109.41, "source_end_s": 111.41 }, { "start_s": 10.0, "end_s": 14.0, "source_start_s": 119.41, "source_end_s": 123.41 }, { "start_s": 30.0, "end_s": 33.0, "source_start_s": 139.41, "source_end_s": 142.41 } ], "evidence": "The selected man's right hand is visibly supported on the stone ledge in the listed close body and fire-side views.", "mask_relation": "subject matches the supplied selected-person mask", "source_subject_refs": "man_1" }, { "id": "action_4", "mentions": [ { "text": "looks", "ranges": [ [ 113, 118 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "type": "prepositional", "subject_refs": [ "man_1" ], "object_refs": [ "fire_1" ], "confidence": "high", "temporal_status": "localized", "intervals": [ { "start_s": 3.0, "end_s": 8.0, "source_start_s": 112.41, "source_end_s": 117.41 }, { "start_s": 35.0, "end_s": 40.0, "source_start_s": 144.41, "source_end_s": 149.41 } ], "evidence": "In the face views, the mask-linked man's head and eyes are directed downward toward the fire area.", "mask_relation": "subject matches the supplied selected-person mask", "source_subject_refs": "man_1" }, { "id": "action_5", "mentions": [ { "text": "speaks", "ranges": [ [ 152, 158 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "type": "simple", "subject_refs": [ "man_1" ], "object_refs": [], "confidence": "medium", "temporal_status": "localized", "intervals": [ { "start_s": 3.0, "end_s": 7.0, "source_start_s": 112.41, "source_end_s": 116.41 }, { "start_s": 35.0, "end_s": 38.0, "source_start_s": 144.41, "source_end_s": 147.41 } ], "evidence": "Successive close face frames show visible mouth articulation by the selected man; no audio evidence was used.", "mask_relation": "subject matches the supplied selected-person mask", "source_subject_refs": "man_1" } ], "notes": [ "The supplied masklet consistently identifies the captioned man, although it becomes partial when the framing centers on the fire.", "No separate source masks were supplied for the blower, hands, stone ledge, or fire; those entity groups therefore have null source-mask links.", "Temporal intervals are sparse support intervals derived only from the sampled clean RGB frames." ] }, "model_output_raw": { "caption": "The man sits back, holding the black blower in his left hand while resting his right hand on the stone ledge. He looks down at the established fire and speaks.", "entity_groups": [ { "id": "man_1", "label": "man", "spans": [ { "text": "The man", "range": [ 0, 7 ] }, { "text": "his", "range": [ 47, 50 ] }, { "text": "his", "range": [ 75, 78 ] }, { "text": "He", "range": [ 110, 112 ] } ], "source_mask": "selected_person_masklet" }, { "id": "blower_1", "label": "blower", "spans": [ { "text": "the black blower", "range": [ 27, 43 ] } ], "source_mask": null }, { "id": "left_hand_1", "label": "left hand", "spans": [ { "text": "his left hand", "range": [ 47, 60 ] } ], "source_mask": null }, { "id": "right_hand_1", "label": "right hand", "spans": [ { "text": "his right hand", "range": [ 75, 89 ] } ], "source_mask": null }, { "id": "stone_ledge_1", "label": "stone ledge", "spans": [ { "text": "the stone ledge", "range": [ 93, 108 ] } ], "source_mask": null }, { "id": "fire_1", "label": "fire", "spans": [ { "text": "the established fire", "range": [ 127, 147 ] } ], "source_mask": null } ], "actions": [ { "id": "action_1", "mentions": [ { "text": "sits", "ranges": [ [ 8, 12 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "type": "posture", "subject": "man_1", "objects": [], "confidence": "high", "temporal_status": "localized", "intervals": [ { "start_s": 1.0, "end_s": 14.0 }, { "start_s": 32.0, "end_s": 40.0 } ], "evidence": "The mask-linked man is visibly seated beside the fire in the early body views and again in the final seated body and face views.", "mask_relation": "subject matches the supplied selected-person mask" }, { "id": "action_2", "mentions": [ { "text": "holding", "ranges": [ [ 19, 26 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "type": "simple", "subject": "man_1", "objects": [ "blower_1", "left_hand_1" ], "confidence": "medium", "temporal_status": "localized", "intervals": [ { "start_s": 34.0, "end_s": 35.0 } ], "evidence": "At 34.0 seconds, the selected man's left hand is visibly positioned over and holding the black blower.", "mask_relation": "subject matches the supplied selected-person mask" }, { "id": "action_3", "mentions": [ { "text": "resting", "ranges": [ [ 67, 74 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "type": "posture", "subject": "man_1", "objects": [ "right_hand_1", "stone_ledge_1" ], "confidence": "high", "temporal_status": "localized", "intervals": [ { "start_s": 0.0, "end_s": 2.0 }, { "start_s": 10.0, "end_s": 14.0 }, { "start_s": 30.0, "end_s": 33.0 } ], "evidence": "The selected man's right hand is visibly supported on the stone ledge in the listed close body and fire-side views.", "mask_relation": "subject matches the supplied selected-person mask" }, { "id": "action_4", "mentions": [ { "text": "looks", "ranges": [ [ 113, 118 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "type": "prepositional", "subject": "man_1", "objects": [ "fire_1" ], "confidence": "high", "temporal_status": "localized", "intervals": [ { "start_s": 3.0, "end_s": 8.0 }, { "start_s": 35.0, "end_s": 40.0 } ], "evidence": "In the face views, the mask-linked man's head and eyes are directed downward toward the fire area.", "mask_relation": "subject matches the supplied selected-person mask" }, { "id": "action_5", "mentions": [ { "text": "speaks", "ranges": [ [ 152, 158 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "type": "simple", "subject": "man_1", "objects": [], "confidence": "medium", "temporal_status": "localized", "intervals": [ { "start_s": 3.0, "end_s": 7.0 }, { "start_s": 35.0, "end_s": 38.0 } ], "evidence": "Successive close face frames show visible mouth articulation by the selected man; no audio evidence was used.", "mask_relation": "subject matches the supplied selected-person mask" } ], "notes": [ "The supplied masklet consistently identifies the captioned man, although it becomes partial when the framing centers on the fire.", "No separate source masks were supplied for the blower, hands, stone ledge, or fire; those entity groups therefore have null source-mask links.", "Temporal intervals are sparse support intervals derived only from the sampled clean RGB frames." ] }, "normalization_repairs": 0 }, { "source_video_id": "v_-Sp-gOEGy2o", "caption_count_category": 13, "clip": { "caption_index": 10, "caption": "The man leans forward again, moving his right hand from the ledge to pick up the matchbox he used earlier.", "source_interval_sec": [ 149.41, 153.99 ], "loom_clip_id": "v_-Sp-gOEGy2o_4478_4615", "clip_frame_interval_4fps": [ 4478, 4615 ], "sample_count_1fps": 5, "sampled_frames": [ { "frame": "00000", "timestamp_sec": 149.41, "offset_sec": 0, "rgb_path": "assets/cat_13/v_-Sp-gOEGy2o_4478_4615/0000s_00000_rgb.jpg", "mask_path": "assets/cat_13/v_-Sp-gOEGy2o_4478_4615/0000s_00000_mask.png", "mask_status": "available" }, { "frame": "00028", "timestamp_sec": 150.41, "offset_sec": 1, "rgb_path": "assets/cat_13/v_-Sp-gOEGy2o_4478_4615/0001s_00028_rgb.jpg", "mask_path": "assets/cat_13/v_-Sp-gOEGy2o_4478_4615/0001s_00028_mask.png", "mask_status": "available" }, { "frame": "00060", "timestamp_sec": 151.41, "offset_sec": 2, "rgb_path": "assets/cat_13/v_-Sp-gOEGy2o_4478_4615/0002s_00060_rgb.jpg", "mask_path": "assets/cat_13/v_-Sp-gOEGy2o_4478_4615/0002s_00060_mask.png", "mask_status": "available" }, { "frame": "00088", "timestamp_sec": 152.41, "offset_sec": 3, "rgb_path": "assets/cat_13/v_-Sp-gOEGy2o_4478_4615/0003s_00088_rgb.jpg", "mask_path": "assets/cat_13/v_-Sp-gOEGy2o_4478_4615/0003s_00088_mask.png", "mask_status": "available" }, { "frame": "00120", "timestamp_sec": 153.41, "offset_sec": 4, "rgb_path": "assets/cat_13/v_-Sp-gOEGy2o_4478_4615/0004s_00120_rgb.jpg", "mask_path": "assets/cat_13/v_-Sp-gOEGy2o_4478_4615/0004s_00120_mask.png", "mask_status": "available" } ], "obj_id": [ 0 ], "anno_id": [ 75 ], "fps_inferred_from_clip_interval": 29.9127 }, "usage": { "input_tokens": 26630, "input_tokens_details": { "cache_write_tokens": 26627, "cached_tokens": 0 }, "output_tokens": 3017, "output_tokens_details": { "reasoning_tokens": 1837 }, "total_tokens": 29647 }, "model": "gpt-5.6-sol", "effort": "medium", "custom_id": "cat13_cap10_v_-Sp-gOEGy2o_4478_4615", "annotation": { "caption": "The man leans forward again, moving his right hand from the ledge to pick up the matchbox he used earlier.", "entity_groups": [ { "id": "entity_1", "label": "man", "spans": [ { "text": "The man", "range": [ 0, 7 ] }, { "text": "his", "range": [ 36, 39 ] }, { "text": "he", "range": [ 90, 92 ] } ], "source_mask_link": "selected_person" }, { "id": "entity_2", "label": "right hand", "spans": [ { "text": "his right hand", "range": [ 36, 50 ] } ], "source_mask_link": null }, { "id": "entity_3", "label": "ledge", "spans": [ { "text": "the ledge", "range": [ 56, 65 ] } ], "source_mask_link": null }, { "id": "entity_4", "label": "matchbox", "spans": [ { "text": "the matchbox", "range": [ 77, 89 ] } ], "source_mask_link": null } ], "actions": [ { "id": "action_1", "mentions": [ { "text": "leans", "ranges": [ [ 8, 13 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "type": "posture", "subject_refs": [ "entity_1" ], "object_refs": [], "confidence": "high", "temporal_status": "supported_by_sampled_frames", "intervals": [ { "start_s": 0.0, "end_s": 3.0, "source_start_s": 149.41, "source_end_s": 152.41 } ], "evidence": "At clip-relative 0.0–2.0 seconds, the masked man is visibly inclined forward and downward toward the work area.", "mask_relation": "subject matches the supplied selected-person mask", "source_subject_refs": [ "entity_1" ] }, { "id": "action_2", "mentions": [ { "text": "moving", "ranges": [ [ 29, 35 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "type": "simple", "subject_refs": [ "entity_1" ], "object_refs": [ "entity_2" ], "confidence": "high", "temporal_status": "supported_by_sampled_frames", "intervals": [ { "start_s": 1.0, "end_s": 4.0, "source_start_s": 150.41, "source_end_s": 153.41 } ], "evidence": "Across samples at 1.0, 2.0, and 3.0 seconds, the masked man's right hand changes position from near his lap toward the stone ledge and matchbox.", "mask_relation": "subject matches the supplied selected-person mask", "source_subject_refs": [ "entity_1" ] }, { "id": "action_3", "mentions": [ { "text": "pick up", "ranges": [ [ 69, 76 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "type": "phrasal", "subject_refs": [ "entity_1" ], "object_refs": [ "entity_4" ], "confidence": "high", "temporal_status": "supported_by_sampled_frames", "intervals": [ { "start_s": 2.0, "end_s": 4.0, "source_start_s": 151.41, "source_end_s": 153.41 } ], "evidence": "At 2.0–3.0 seconds, the masked man's hand grasps and lifts the rectangular matchbox from the ledge.", "mask_relation": "subject matches the supplied selected-person mask", "source_subject_refs": [ "entity_1" ] }, { "id": "action_4", "mentions": [ { "text": "used", "ranges": [ [ 93, 97 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "type": "simple", "subject_refs": [ "entity_1" ], "object_refs": [ "entity_4" ], "confidence": "medium", "temporal_status": "earlier_event_not_supported_in_current_clip", "intervals": [], "evidence": "The caption places this matchbox-use event earlier; none of the supplied frames visibly supports that prior event.", "mask_relation": "subject matches the supplied selected-person mask", "source_subject_refs": [ "entity_1" ] } ], "notes": [ "The supplied masklet consistently identifies the man, although only his arm and hand remain visible near the end of the clip.", "The hand, ledge, and matchbox are required text-groundable entities but have no independent supplied instance masks.", "Temporal intervals are sparse clip-relative support intervals derived only from the sampled clean RGB frames." ] }, "model_output_raw": { "caption": "The man leans forward again, moving his right hand from the ledge to pick up the matchbox he used earlier.", "entity_groups": [ { "id": "entity_1", "entity": "man", "spans": [ { "text": "The man", "range": [ 0, 7 ] }, { "text": "his", "range": [ 36, 39 ] }, { "text": "he", "range": [ 90, 92 ] } ], "source_mask": "selected_person_mask" }, { "id": "entity_2", "entity": "right hand", "spans": [ { "text": "his right hand", "range": [ 36, 50 ] } ], "source_mask": null }, { "id": "entity_3", "entity": "ledge", "spans": [ { "text": "the ledge", "range": [ 56, 65 ] } ], "source_mask": null }, { "id": "entity_4", "entity": "matchbox", "spans": [ { "text": "the matchbox", "range": [ 77, 89 ] } ], "source_mask": null } ], "actions": [ { "id": "action_1", "mentions": [ { "text": "leans", "ranges": [ [ 8, 13 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "type": "posture", "subject_refs": [ "entity_1" ], "object_refs": [], "confidence": "high", "mask_relation": "subject matches the supplied selected-person mask", "temporal_status": "supported_by_sampled_frames", "intervals": [ { "start_s": 0.0, "end_s": 3.0 } ], "evidence": "At clip-relative 0.0–2.0 seconds, the masked man is visibly inclined forward and downward toward the work area." }, { "id": "action_2", "mentions": [ { "text": "moving", "ranges": [ [ 29, 35 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "type": "simple", "subject_refs": [ "entity_1" ], "object_refs": [ "entity_2" ], "confidence": "high", "mask_relation": "subject matches the supplied selected-person mask", "temporal_status": "supported_by_sampled_frames", "intervals": [ { "start_s": 1.0, "end_s": 4.0 } ], "evidence": "Across samples at 1.0, 2.0, and 3.0 seconds, the masked man's right hand changes position from near his lap toward the stone ledge and matchbox." }, { "id": "action_3", "mentions": [ { "text": "pick up", "ranges": [ [ 69, 76 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "type": "phrasal", "subject_refs": [ "entity_1" ], "object_refs": [ "entity_4" ], "confidence": "high", "mask_relation": "subject matches the supplied selected-person mask", "temporal_status": "supported_by_sampled_frames", "intervals": [ { "start_s": 2.0, "end_s": 4.0 } ], "evidence": "At 2.0–3.0 seconds, the masked man's hand grasps and lifts the rectangular matchbox from the ledge." }, { "id": "action_4", "mentions": [ { "text": "used", "ranges": [ [ 93, 97 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "type": "simple", "subject_refs": [ "entity_1" ], "object_refs": [ "entity_4" ], "confidence": "medium", "mask_relation": "subject matches the supplied selected-person mask", "temporal_status": "earlier_event_not_supported_in_current_clip", "intervals": [], "evidence": "The caption places this matchbox-use event earlier; none of the supplied frames visibly supports that prior event." } ], "notes": [ "The supplied masklet consistently identifies the man, although only his arm and hand remain visible near the end of the clip.", "The hand, ledge, and matchbox are required text-groundable entities but have no independent supplied instance masks.", "Temporal intervals are sparse clip-relative support intervals derived only from the sampled clean RGB frames." ] }, "normalization_repairs": 0 }, { "source_video_id": "v_-Sp-gOEGy2o", "caption_count_category": 13, "clip": { "caption_index": 11, "caption": "A person wearing a green short-sleeved shirt, glasses, and blue jeans squats by a fire pit. While holding a white box in one hand, they use the other to carefully place a log on top of a small, fledgling fire. They hold the log in place, watching as the flames underneath it begin to grow larger.", "source_interval_sec": [ 166.67, 179.67 ], "loom_clip_id": "v_-Sp-gOEGy2o_4995_5385", "clip_frame_interval_4fps": [ 4995, 5385 ], "sample_count_1fps": 13, "sampled_frames": [ { "frame": "00000", "timestamp_sec": 166.67, "offset_sec": 0, "rgb_path": "assets/cat_13/v_-Sp-gOEGy2o_4995_5385/0000s_00000_rgb.jpg", "mask_path": "assets/cat_13/v_-Sp-gOEGy2o_4995_5385/0000s_00000_mask.png", "mask_status": "available" }, { "frame": "00028", "timestamp_sec": 167.67, "offset_sec": 1, "rgb_path": "assets/cat_13/v_-Sp-gOEGy2o_4995_5385/0001s_00028_rgb.jpg", "mask_path": "assets/cat_13/v_-Sp-gOEGy2o_4995_5385/0001s_00028_mask.png", "mask_status": "available" }, { "frame": "00060", "timestamp_sec": 168.67, "offset_sec": 2, "rgb_path": "assets/cat_13/v_-Sp-gOEGy2o_4995_5385/0002s_00060_rgb.jpg", "mask_path": "assets/cat_13/v_-Sp-gOEGy2o_4995_5385/0002s_00060_mask.png", "mask_status": "available" }, { "frame": "00088", "timestamp_sec": 169.67, "offset_sec": 3, "rgb_path": "assets/cat_13/v_-Sp-gOEGy2o_4995_5385/0003s_00088_rgb.jpg", "mask_path": "assets/cat_13/v_-Sp-gOEGy2o_4995_5385/0003s_00088_mask.png", "mask_status": "available" }, { "frame": "00120", "timestamp_sec": 170.67, "offset_sec": 4, "rgb_path": "assets/cat_13/v_-Sp-gOEGy2o_4995_5385/0004s_00120_rgb.jpg", "mask_path": "assets/cat_13/v_-Sp-gOEGy2o_4995_5385/0004s_00120_mask.png", "mask_status": "available" }, { "frame": "00148", "timestamp_sec": 171.67, "offset_sec": 5, "rgb_path": "assets/cat_13/v_-Sp-gOEGy2o_4995_5385/0005s_00148_rgb.jpg", "mask_path": "assets/cat_13/v_-Sp-gOEGy2o_4995_5385/0005s_00148_mask.png", "mask_status": "available" }, { "frame": "00180", "timestamp_sec": 172.67, "offset_sec": 6, "rgb_path": "assets/cat_13/v_-Sp-gOEGy2o_4995_5385/0006s_00180_rgb.jpg", "mask_path": "assets/cat_13/v_-Sp-gOEGy2o_4995_5385/0006s_00180_mask.png", "mask_status": "available" }, { "frame": "00208", "timestamp_sec": 173.67, "offset_sec": 7, "rgb_path": "assets/cat_13/v_-Sp-gOEGy2o_4995_5385/0007s_00208_rgb.jpg", "mask_path": "assets/cat_13/v_-Sp-gOEGy2o_4995_5385/0007s_00208_mask.png", "mask_status": "available" }, { "frame": "00240", "timestamp_sec": 174.67, "offset_sec": 8, "rgb_path": "assets/cat_13/v_-Sp-gOEGy2o_4995_5385/0008s_00240_rgb.jpg", "mask_path": "assets/cat_13/v_-Sp-gOEGy2o_4995_5385/0008s_00240_mask.png", "mask_status": "available" }, { "frame": "00268", "timestamp_sec": 175.67, "offset_sec": 9, "rgb_path": "assets/cat_13/v_-Sp-gOEGy2o_4995_5385/0009s_00268_rgb.jpg", "mask_path": "assets/cat_13/v_-Sp-gOEGy2o_4995_5385/0009s_00268_mask.png", "mask_status": "available" }, { "frame": "00300", "timestamp_sec": 176.67, "offset_sec": 10, "rgb_path": "assets/cat_13/v_-Sp-gOEGy2o_4995_5385/0010s_00300_rgb.jpg", "mask_path": "assets/cat_13/v_-Sp-gOEGy2o_4995_5385/0010s_00300_mask.png", "mask_status": "available" }, { "frame": "00328", "timestamp_sec": 177.67, "offset_sec": 11, "rgb_path": "assets/cat_13/v_-Sp-gOEGy2o_4995_5385/0011s_00328_rgb.jpg", "mask_path": "assets/cat_13/v_-Sp-gOEGy2o_4995_5385/0011s_00328_mask.png", "mask_status": "available" }, { "frame": "00360", "timestamp_sec": 178.67, "offset_sec": 12, "rgb_path": "assets/cat_13/v_-Sp-gOEGy2o_4995_5385/0012s_00360_rgb.jpg", "mask_path": "assets/cat_13/v_-Sp-gOEGy2o_4995_5385/0012s_00360_mask.png", "mask_status": "available" } ], "obj_id": [ 0 ], "anno_id": [ 76 ], "fps_inferred_from_clip_interval": 30.0 }, "usage": { "input_tokens": 44778, "input_tokens_details": { "cache_write_tokens": 44775, "cached_tokens": 0 }, "output_tokens": 6028, "output_tokens_details": { "reasoning_tokens": 3564 }, "total_tokens": 50806 }, "model": "gpt-5.6-sol", "effort": "medium", "custom_id": "cat13_cap11_v_-Sp-gOEGy2o_4995_5385", "annotation": { "caption": "A person wearing a green short-sleeved shirt, glasses, and blue jeans squats by a fire pit. While holding a white box in one hand, they use the other to carefully place a log on top of a small, fledgling fire. They hold the log in place, watching as the flames underneath it begin to grow larger.", "entity_groups": [ { "id": "E1", "label": "person", "spans": [ { "text": "A person", "range": [ 0, 8 ] }, { "text": "they", "range": [ 131, 135 ] }, { "text": "They", "range": [ 210, 214 ] } ], "source_mask_link": "selected_person" }, { "id": "E2", "label": "shirt", "spans": [ { "text": "a green short-sleeved shirt", "range": [ 17, 44 ] } ], "source_mask_link": null }, { "id": "E3", "label": "glasses", "spans": [ { "text": "glasses", "range": [ 46, 53 ] } ], "source_mask_link": null }, { "id": "E4", "label": "jeans", "spans": [ { "text": "blue jeans", "range": [ 59, 69 ] } ], "source_mask_link": null }, { "id": "E5", "label": "fire pit", "spans": [ { "text": "a fire pit", "range": [ 80, 90 ] } ], "source_mask_link": null }, { "id": "E6", "label": "box", "spans": [ { "text": "a white box", "range": [ 106, 117 ] } ], "source_mask_link": null }, { "id": "E7", "label": "hand", "spans": [ { "text": "one hand", "range": [ 121, 129 ] } ], "source_mask_link": null }, { "id": "E8", "label": "hand", "spans": [ { "text": "the other", "range": [ 140, 149 ] } ], "source_mask_link": null }, { "id": "E9", "label": "log", "spans": [ { "text": "a log", "range": [ 169, 174 ] }, { "text": "the log", "range": [ 220, 227 ] }, { "text": "it", "range": [ 272, 274 ] } ], "source_mask_link": null }, { "id": "E10", "label": "fire", "spans": [ { "text": "a small, fledgling fire", "range": [ 185, 208 ] }, { "text": "the flames", "range": [ 250, 260 ] } ], "source_mask_link": null } ], "actions": [ { "id": "A1", "mentions": [ { "text": "wearing", "ranges": [ [ 9, 16 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "type": "simple", "subject_refs": [ "E1" ], "object_refs": [ "E2", "E3", "E4" ], "confidence": "high", "temporal_status": "localized_from_sparse_samples", "intervals": [ { "start_s": 1.0, "end_s": 13.0, "source_start_s": 167.67, "source_end_s": 179.67 } ], "evidence": "Samples 1–12 show the masked person continuously wearing the green shirt and blue jeans; the glasses are visible in the wider later samples.", "mask_relation": "subject_matches_supplied_selected_person_mask", "source_subject_refs": [ "E1" ] }, { "id": "A2", "mentions": [ { "text": "squats", "ranges": [ [ 70, 76 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "type": "posture", "subject_refs": [ "E1" ], "object_refs": [], "confidence": "high", "temporal_status": "localized_from_sparse_samples", "intervals": [ { "start_s": 1.0, "end_s": 13.0, "source_start_s": 167.67, "source_end_s": 179.67 } ], "evidence": "The masked person maintains a low, folded-leg squatting posture beside the fire pit throughout samples 1–12.", "mask_relation": "subject_matches_supplied_selected_person_mask", "source_subject_refs": [ "E1" ] }, { "id": "A3", "mentions": [ { "text": "holding", "ranges": [ [ 98, 105 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "type": "simple", "subject_refs": [ "E1" ], "object_refs": [ "E6" ], "confidence": "high", "temporal_status": "localized_from_sparse_samples", "intervals": [ { "start_s": 1.0, "end_s": 13.0, "source_start_s": 167.67, "source_end_s": 179.67 } ], "evidence": "Samples 1–12 visibly show the masked person retaining the white box in one hand.", "mask_relation": "subject_matches_supplied_selected_person_mask", "source_subject_refs": [ "E1" ] }, { "id": "A4", "mentions": [ { "text": "use", "ranges": [ [ 136, 139 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "type": "simple", "subject_refs": [ "E1" ], "object_refs": [ "E8" ], "confidence": "high", "temporal_status": "localized_from_sparse_samples", "intervals": [ { "start_s": 5.0, "end_s": 9.0, "source_start_s": 171.67, "source_end_s": 175.67 } ], "evidence": "In samples 5–8, the masked person uses the free hand to guide the long log while the other hand retains the box.", "mask_relation": "subject_matches_supplied_selected_person_mask", "source_subject_refs": [ "E1" ] }, { "id": "A5", "mentions": [ { "text": "place", "ranges": [ [ 163, 168 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "type": "simple", "subject_refs": [ "E1" ], "object_refs": [ "E9" ], "confidence": "high", "temporal_status": "localized_from_sparse_samples", "intervals": [ { "start_s": 5.0, "end_s": 9.0, "source_start_s": 171.67, "source_end_s": 175.67 } ], "evidence": "Samples 5–8 show the masked person guiding and lowering the long log across the small fire.", "mask_relation": "subject_matches_supplied_selected_person_mask", "source_subject_refs": [ "E1" ] }, { "id": "A6", "mentions": [ { "text": "hold", "ranges": [ [ 215, 219 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "type": "simple", "subject_refs": [ "E1" ], "object_refs": [ "E9" ], "confidence": "high", "temporal_status": "localized_from_sparse_samples", "intervals": [ { "start_s": 9.0, "end_s": 13.0, "source_start_s": 175.67, "source_end_s": 179.67 } ], "evidence": "Samples 9–12 show the log resting over the fire while the masked person's hand remains on it to steady it.", "mask_relation": "subject_matches_supplied_selected_person_mask", "source_subject_refs": [ "E1" ] }, { "id": "A7", "mentions": [ { "text": "watching", "ranges": [ [ 238, 246 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "type": "simple", "subject_refs": [ "E1" ], "object_refs": [ "E10" ], "confidence": "medium", "temporal_status": "localized_from_sparse_samples", "intervals": [ { "start_s": 5.0, "end_s": 6.0, "source_start_s": 171.67, "source_end_s": 172.67 }, { "start_s": 7.0, "end_s": 13.0, "source_start_s": 173.67, "source_end_s": 179.67 } ], "evidence": "The masked person's head and gaze are directed toward the log and fire in sample 5 and again in samples 7–12; sample 6 shows a glance away.", "mask_relation": "subject_matches_supplied_selected_person_mask", "source_subject_refs": [ "E1" ] } ], "notes": [ "The supplied masklet identifies only the person group E1; all other entity groups lack independent source masks and therefore have null links.", "Sample 0 has an empty person mask and does not provide support for a masked-person action.", "The fire and its flames are grouped as one visual fire instance.", "The caption's growth event expressed by \"grow\" is omitted because its subject is the unmasked flames, not the selected masked person.", "\"begin\" is an aspectual predicate and is excluded from the action span." ] }, "model_output_raw": { "caption": "A person wearing a green short-sleeved shirt, glasses, and blue jeans squats by a fire pit. While holding a white box in one hand, they use the other to carefully place a log on top of a small, fledgling fire. They hold the log in place, watching as the flames underneath it begin to grow larger.", "entity_groups": [ { "id": "E1", "label": "person", "spans": [ { "text": "A person", "range": [ 0, 8 ] }, { "text": "they", "range": [ 131, 135 ] }, { "text": "They", "range": [ 210, 214 ] } ], "source_mask_link": "selected_person_mask" }, { "id": "E2", "label": "shirt", "spans": [ { "text": "a green short-sleeved shirt", "range": [ 17, 44 ] } ], "source_mask_link": null }, { "id": "E3", "label": "glasses", "spans": [ { "text": "glasses", "range": [ 46, 53 ] } ], "source_mask_link": null }, { "id": "E4", "label": "jeans", "spans": [ { "text": "blue jeans", "range": [ 59, 69 ] } ], "source_mask_link": null }, { "id": "E5", "label": "fire pit", "spans": [ { "text": "a fire pit", "range": [ 80, 90 ] } ], "source_mask_link": null }, { "id": "E6", "label": "box", "spans": [ { "text": "a white box", "range": [ 106, 117 ] } ], "source_mask_link": null }, { "id": "E7", "label": "hand", "instance": 1, "spans": [ { "text": "one hand", "range": [ 121, 129 ] } ], "source_mask_link": null }, { "id": "E8", "label": "hand", "instance": 2, "spans": [ { "text": "the other", "range": [ 140, 149 ] } ], "source_mask_link": null }, { "id": "E9", "label": "log", "spans": [ { "text": "a log", "range": [ 169, 174 ] }, { "text": "the log", "range": [ 220, 227 ] }, { "text": "it", "range": [ 272, 274 ] } ], "source_mask_link": null }, { "id": "E10", "label": "fire", "spans": [ { "text": "a small, fledgling fire", "range": [ 185, 208 ] }, { "text": "the flames", "range": [ 250, 260 ] } ], "source_mask_link": null } ], "actions": [ { "id": "A1", "mentions": [ { "text": "wearing", "occurrence": 1, "ranges": [ [ 9, 16 ] ], "particle": { "text": null, "occurrence": 1 } } ], "type": "simple", "subject_refs": [ "E1" ], "object_refs": [ "E2", "E3", "E4" ], "confidence": "high", "mask_relation": "subject_matches_supplied_selected_person_mask", "temporal_status": "localized_from_sparse_samples", "intervals": [ { "start_s": 1.0, "end_s": 13.0, "source_start_s": 167.67, "source_end_s": 179.67 } ], "evidence": "Samples 1–12 show the masked person continuously wearing the green shirt and blue jeans; the glasses are visible in the wider later samples." }, { "id": "A2", "mentions": [ { "text": "squats", "occurrence": 1, "ranges": [ [ 70, 76 ] ], "particle": { "text": null, "occurrence": 1 } } ], "type": "posture", "subject_refs": [ "E1" ], "object_refs": [], "confidence": "high", "mask_relation": "subject_matches_supplied_selected_person_mask", "temporal_status": "localized_from_sparse_samples", "intervals": [ { "start_s": 1.0, "end_s": 13.0, "source_start_s": 167.67, "source_end_s": 179.67 } ], "evidence": "The masked person maintains a low, folded-leg squatting posture beside the fire pit throughout samples 1–12." }, { "id": "A3", "mentions": [ { "text": "holding", "occurrence": 1, "ranges": [ [ 98, 105 ] ], "particle": { "text": null, "occurrence": 1 } } ], "type": "simple", "subject_refs": [ "E1" ], "object_refs": [ "E6" ], "confidence": "high", "mask_relation": "subject_matches_supplied_selected_person_mask", "temporal_status": "localized_from_sparse_samples", "intervals": [ { "start_s": 1.0, "end_s": 13.0, "source_start_s": 167.67, "source_end_s": 179.67 } ], "evidence": "Samples 1–12 visibly show the masked person retaining the white box in one hand." }, { "id": "A4", "mentions": [ { "text": "use", "occurrence": 1, "ranges": [ [ 136, 139 ] ], "particle": { "text": null, "occurrence": 1 } } ], "type": "simple", "subject_refs": [ "E1" ], "object_refs": [ "E8" ], "confidence": "high", "mask_relation": "subject_matches_supplied_selected_person_mask", "temporal_status": "localized_from_sparse_samples", "intervals": [ { "start_s": 5.0, "end_s": 9.0, "source_start_s": 171.67, "source_end_s": 175.67 } ], "evidence": "In samples 5–8, the masked person uses the free hand to guide the long log while the other hand retains the box." }, { "id": "A5", "mentions": [ { "text": "place", "occurrence": 1, "ranges": [ [ 163, 168 ] ], "particle": { "text": null, "occurrence": 1 } } ], "type": "simple", "subject_refs": [ "E1" ], "object_refs": [ "E9" ], "confidence": "high", "mask_relation": "subject_matches_supplied_selected_person_mask", "temporal_status": "localized_from_sparse_samples", "intervals": [ { "start_s": 5.0, "end_s": 9.0, "source_start_s": 171.67, "source_end_s": 175.67 } ], "evidence": "Samples 5–8 show the masked person guiding and lowering the long log across the small fire." }, { "id": "A6", "mentions": [ { "text": "hold", "occurrence": 1, "ranges": [ [ 215, 219 ] ], "particle": { "text": null, "occurrence": 1 } } ], "type": "simple", "subject_refs": [ "E1" ], "object_refs": [ "E9" ], "confidence": "high", "mask_relation": "subject_matches_supplied_selected_person_mask", "temporal_status": "localized_from_sparse_samples", "intervals": [ { "start_s": 9.0, "end_s": 13.0, "source_start_s": 175.67, "source_end_s": 179.67 } ], "evidence": "Samples 9–12 show the log resting over the fire while the masked person's hand remains on it to steady it." }, { "id": "A7", "mentions": [ { "text": "watching", "occurrence": 1, "ranges": [ [ 238, 246 ] ], "particle": { "text": null, "occurrence": 1 } } ], "type": "simple", "subject_refs": [ "E1" ], "object_refs": [ "E10" ], "confidence": "medium", "mask_relation": "subject_matches_supplied_selected_person_mask", "temporal_status": "localized_from_sparse_samples", "intervals": [ { "start_s": 5.0, "end_s": 6.0, "source_start_s": 171.67, "source_end_s": 172.67 }, { "start_s": 7.0, "end_s": 13.0, "source_start_s": 173.67, "source_end_s": 179.67 } ], "evidence": "The masked person's head and gaze are directed toward the log and fire in sample 5 and again in samples 7–12; sample 6 shows a glance away." } ], "notes": [ "The supplied masklet identifies only the person group E1; all other entity groups lack independent source masks and therefore have null links.", "Sample 0 has an empty person mask and does not provide support for a masked-person action.", "The fire and its flames are grouped as one visual fire instance.", "The caption's growth event expressed by \"grow\" is omitted because its subject is the unmasked flames, not the selected masked person.", "\"begin\" is an aspectual predicate and is excluded from the action span." ] }, "normalization_repairs": 0 }, { "source_video_id": "v_-Sp-gOEGy2o", "caption_count_category": 13, "clip": { "caption_index": 12, "caption": "The person lets go of the log and picks up a black, cylindrical bellows with their right hand, while their left hand holds the white box on their lap. They aim the bellows at the base of the fire and repeatedly pump it to stoke the flames. Throughout this period, they consistently work the bellows to help the fire grow stronger.", "source_interval_sec": [ 179.67, 209.67 ], "loom_clip_id": "v_-Sp-gOEGy2o_5385_6284", "clip_frame_interval_4fps": [ 5385, 6284 ], "sample_count_1fps": 30, "sampled_frames": [ { "frame": "00000", "timestamp_sec": 179.67, "offset_sec": 0, "rgb_path": "assets/cat_13/v_-Sp-gOEGy2o_5385_6284/0000s_00000_rgb.jpg", "mask_path": "assets/cat_13/v_-Sp-gOEGy2o_5385_6284/0000s_00000_mask.png", "mask_status": "available" }, { "frame": "00028", "timestamp_sec": 180.67, "offset_sec": 1, "rgb_path": "assets/cat_13/v_-Sp-gOEGy2o_5385_6284/0001s_00028_rgb.jpg", "mask_path": "assets/cat_13/v_-Sp-gOEGy2o_5385_6284/0001s_00028_mask.png", "mask_status": "available" }, { "frame": "00060", "timestamp_sec": 181.67, "offset_sec": 2, "rgb_path": "assets/cat_13/v_-Sp-gOEGy2o_5385_6284/0002s_00060_rgb.jpg", "mask_path": "assets/cat_13/v_-Sp-gOEGy2o_5385_6284/0002s_00060_mask.png", "mask_status": "available" }, { "frame": "00088", "timestamp_sec": 182.67, "offset_sec": 3, "rgb_path": "assets/cat_13/v_-Sp-gOEGy2o_5385_6284/0003s_00088_rgb.jpg", "mask_path": "assets/cat_13/v_-Sp-gOEGy2o_5385_6284/0003s_00088_mask.png", "mask_status": "available" }, { "frame": "00120", "timestamp_sec": 183.67, "offset_sec": 4, "rgb_path": "assets/cat_13/v_-Sp-gOEGy2o_5385_6284/0004s_00120_rgb.jpg", "mask_path": "assets/cat_13/v_-Sp-gOEGy2o_5385_6284/0004s_00120_mask.png", "mask_status": "available" }, { "frame": "00148", "timestamp_sec": 184.67, "offset_sec": 5, "rgb_path": "assets/cat_13/v_-Sp-gOEGy2o_5385_6284/0005s_00148_rgb.jpg", "mask_path": "assets/cat_13/v_-Sp-gOEGy2o_5385_6284/0005s_00148_mask.png", "mask_status": "available" }, { "frame": "00180", "timestamp_sec": 185.67, "offset_sec": 6, "rgb_path": "assets/cat_13/v_-Sp-gOEGy2o_5385_6284/0006s_00180_rgb.jpg", "mask_path": "assets/cat_13/v_-Sp-gOEGy2o_5385_6284/0006s_00180_mask.png", "mask_status": "available" }, { "frame": "00208", "timestamp_sec": 186.67, "offset_sec": 7, "rgb_path": "assets/cat_13/v_-Sp-gOEGy2o_5385_6284/0007s_00208_rgb.jpg", "mask_path": "assets/cat_13/v_-Sp-gOEGy2o_5385_6284/0007s_00208_mask.png", "mask_status": "available" }, { "frame": "00240", "timestamp_sec": 187.67, "offset_sec": 8, "rgb_path": "assets/cat_13/v_-Sp-gOEGy2o_5385_6284/0008s_00240_rgb.jpg", "mask_path": "assets/cat_13/v_-Sp-gOEGy2o_5385_6284/0008s_00240_mask.png", "mask_status": "available" }, { "frame": "00268", "timestamp_sec": 188.67, "offset_sec": 9, "rgb_path": "assets/cat_13/v_-Sp-gOEGy2o_5385_6284/0009s_00268_rgb.jpg", "mask_path": "assets/cat_13/v_-Sp-gOEGy2o_5385_6284/0009s_00268_mask.png", "mask_status": "available" }, { "frame": "00300", "timestamp_sec": 189.67, "offset_sec": 10, "rgb_path": "assets/cat_13/v_-Sp-gOEGy2o_5385_6284/0010s_00300_rgb.jpg", "mask_path": "assets/cat_13/v_-Sp-gOEGy2o_5385_6284/0010s_00300_mask.png", "mask_status": "available" }, { "frame": "00328", "timestamp_sec": 190.67, "offset_sec": 11, "rgb_path": "assets/cat_13/v_-Sp-gOEGy2o_5385_6284/0011s_00328_rgb.jpg", "mask_path": "assets/cat_13/v_-Sp-gOEGy2o_5385_6284/0011s_00328_mask.png", "mask_status": "available" }, { "frame": "00360", "timestamp_sec": 191.67, "offset_sec": 12, "rgb_path": "assets/cat_13/v_-Sp-gOEGy2o_5385_6284/0012s_00360_rgb.jpg", "mask_path": "assets/cat_13/v_-Sp-gOEGy2o_5385_6284/0012s_00360_mask.png", "mask_status": "available" }, { "frame": "00388", "timestamp_sec": 192.67, "offset_sec": 13, "rgb_path": "assets/cat_13/v_-Sp-gOEGy2o_5385_6284/0013s_00388_rgb.jpg", "mask_path": "assets/cat_13/v_-Sp-gOEGy2o_5385_6284/0013s_00388_mask.png", "mask_status": "available" }, { "frame": "00420", "timestamp_sec": 193.67, "offset_sec": 14, "rgb_path": "assets/cat_13/v_-Sp-gOEGy2o_5385_6284/0014s_00420_rgb.jpg", "mask_path": "assets/cat_13/v_-Sp-gOEGy2o_5385_6284/0014s_00420_mask.png", "mask_status": "available" }, { "frame": "00448", "timestamp_sec": 194.67, "offset_sec": 15, "rgb_path": "assets/cat_13/v_-Sp-gOEGy2o_5385_6284/0015s_00448_rgb.jpg", "mask_path": "assets/cat_13/v_-Sp-gOEGy2o_5385_6284/0015s_00448_mask.png", "mask_status": "available" }, { "frame": "00480", "timestamp_sec": 195.67, "offset_sec": 16, "rgb_path": "assets/cat_13/v_-Sp-gOEGy2o_5385_6284/0016s_00480_rgb.jpg", "mask_path": "assets/cat_13/v_-Sp-gOEGy2o_5385_6284/0016s_00480_mask.png", "mask_status": "available" }, { "frame": "00508", "timestamp_sec": 196.67, "offset_sec": 17, "rgb_path": "assets/cat_13/v_-Sp-gOEGy2o_5385_6284/0017s_00508_rgb.jpg", "mask_path": "assets/cat_13/v_-Sp-gOEGy2o_5385_6284/0017s_00508_mask.png", "mask_status": "available" }, { "frame": "00540", "timestamp_sec": 197.67, "offset_sec": 18, "rgb_path": "assets/cat_13/v_-Sp-gOEGy2o_5385_6284/0018s_00540_rgb.jpg", "mask_path": "assets/cat_13/v_-Sp-gOEGy2o_5385_6284/0018s_00540_mask.png", "mask_status": "available" }, { "frame": "00568", "timestamp_sec": 198.67, "offset_sec": 19, "rgb_path": "assets/cat_13/v_-Sp-gOEGy2o_5385_6284/0019s_00568_rgb.jpg", "mask_path": "assets/cat_13/v_-Sp-gOEGy2o_5385_6284/0019s_00568_mask.png", "mask_status": "available" }, { "frame": "00600", "timestamp_sec": 199.67, "offset_sec": 20, "rgb_path": "assets/cat_13/v_-Sp-gOEGy2o_5385_6284/0020s_00600_rgb.jpg", "mask_path": "assets/cat_13/v_-Sp-gOEGy2o_5385_6284/0020s_00600_mask.png", "mask_status": "available" }, { "frame": "00628", "timestamp_sec": 200.67, "offset_sec": 21, "rgb_path": "assets/cat_13/v_-Sp-gOEGy2o_5385_6284/0021s_00628_rgb.jpg", "mask_path": "assets/cat_13/v_-Sp-gOEGy2o_5385_6284/0021s_00628_mask.png", "mask_status": "available" }, { "frame": "00660", "timestamp_sec": 201.67, "offset_sec": 22, "rgb_path": "assets/cat_13/v_-Sp-gOEGy2o_5385_6284/0022s_00660_rgb.jpg", "mask_path": "assets/cat_13/v_-Sp-gOEGy2o_5385_6284/0022s_00660_mask.png", "mask_status": "available" }, { "frame": "00688", "timestamp_sec": 202.67, "offset_sec": 23, "rgb_path": "assets/cat_13/v_-Sp-gOEGy2o_5385_6284/0023s_00688_rgb.jpg", "mask_path": "assets/cat_13/v_-Sp-gOEGy2o_5385_6284/0023s_00688_mask.png", "mask_status": "available" }, { "frame": "00720", "timestamp_sec": 203.67, "offset_sec": 24, "rgb_path": "assets/cat_13/v_-Sp-gOEGy2o_5385_6284/0024s_00720_rgb.jpg", "mask_path": "assets/cat_13/v_-Sp-gOEGy2o_5385_6284/0024s_00720_mask.png", "mask_status": "available" }, { "frame": "00748", "timestamp_sec": 204.67, "offset_sec": 25, "rgb_path": "assets/cat_13/v_-Sp-gOEGy2o_5385_6284/0025s_00748_rgb.jpg", "mask_path": "assets/cat_13/v_-Sp-gOEGy2o_5385_6284/0025s_00748_mask.png", "mask_status": "available" }, { "frame": "00780", "timestamp_sec": 205.67, "offset_sec": 26, "rgb_path": "assets/cat_13/v_-Sp-gOEGy2o_5385_6284/0026s_00780_rgb.jpg", "mask_path": "assets/cat_13/v_-Sp-gOEGy2o_5385_6284/0026s_00780_mask.png", "mask_status": "available" }, { "frame": "00808", "timestamp_sec": 206.67, "offset_sec": 27, "rgb_path": "assets/cat_13/v_-Sp-gOEGy2o_5385_6284/0027s_00808_rgb.jpg", "mask_path": "assets/cat_13/v_-Sp-gOEGy2o_5385_6284/0027s_00808_mask.png", "mask_status": "available" }, { "frame": "00840", "timestamp_sec": 207.67, "offset_sec": 28, "rgb_path": "assets/cat_13/v_-Sp-gOEGy2o_5385_6284/0028s_00840_rgb.jpg", "mask_path": "assets/cat_13/v_-Sp-gOEGy2o_5385_6284/0028s_00840_mask.png", "mask_status": "available" }, { "frame": "00868", "timestamp_sec": 208.67, "offset_sec": 29, "rgb_path": "assets/cat_13/v_-Sp-gOEGy2o_5385_6284/0029s_00868_rgb.jpg", "mask_path": "assets/cat_13/v_-Sp-gOEGy2o_5385_6284/0029s_00868_mask.png", "mask_status": "available" } ], "obj_id": [ 0 ], "anno_id": [ 77 ], "fps_inferred_from_clip_interval": 29.9667 }, "usage": { "input_tokens": 83251, "input_tokens_details": { "cache_write_tokens": 83248, "cached_tokens": 0 }, "output_tokens": 4176, "output_tokens_details": { "reasoning_tokens": 2070 }, "total_tokens": 87427 }, "model": "gpt-5.6-sol", "effort": "medium", "custom_id": "cat13_cap12_v_-Sp-gOEGy2o_5385_6284", "annotation": { "caption": "The person lets go of the log and picks up a black, cylindrical bellows with their right hand, while their left hand holds the white box on their lap. They aim the bellows at the base of the fire and repeatedly pump it to stoke the flames. Throughout this period, they consistently work the bellows to help the fire grow stronger.", "entity_groups": [ { "id": "person_1", "label": "person", "spans": [ { "text": "The person", "range": [ 0, 10 ] }, { "text": "their", "range": [ 77, 82 ] }, { "text": "their", "range": [ 101, 106 ] }, { "text": "their", "range": [ 140, 145 ] }, { "text": "They", "range": [ 151, 155 ] }, { "text": "they", "range": [ 264, 268 ] } ], "source_mask_link": "selected_person" }, { "id": "log_1", "label": "log", "spans": [ { "text": "the log", "range": [ 22, 29 ] } ], "source_mask_link": null }, { "id": "bellows_1", "label": "bellows", "spans": [ { "text": "a black, cylindrical bellows", "range": [ 43, 71 ] }, { "text": "the bellows", "range": [ 160, 171 ] }, { "text": "it", "range": [ 216, 218 ] }, { "text": "the bellows", "range": [ 287, 298 ] } ], "source_mask_link": null }, { "id": "right_hand_1", "label": "right hand", "spans": [ { "text": "their right hand", "range": [ 77, 93 ] } ], "source_mask_link": null }, { "id": "left_hand_1", "label": "left hand", "spans": [ { "text": "their left hand", "range": [ 101, 116 ] } ], "source_mask_link": null }, { "id": "box_1", "label": "box", "spans": [ { "text": "the white box", "range": [ 123, 136 ] } ], "source_mask_link": null }, { "id": "lap_1", "label": "lap", "spans": [ { "text": "their lap", "range": [ 140, 149 ] } ], "source_mask_link": null }, { "id": "fire_1", "label": "fire", "spans": [ { "text": "the fire", "range": [ 187, 195 ] }, { "text": "the flames", "range": [ 228, 238 ] }, { "text": "the fire", "range": [ 307, 315 ] } ], "source_mask_link": null } ], "actions": [ { "id": "action_1", "mentions": [ { "text": "lets go", "ranges": [ [ 11, 18 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "type": "phrasal", "subject_refs": [ "person_1" ], "object_refs": [ "log_1" ], "confidence": "high", "temporal_status": "not_observed_in_sampled_frames", "intervals": [], "evidence": "The selected person is visible throughout, but the first sampled frame already shows the fire setup and does not capture the release transition.", "mask_relation": "subject_matches_selected_person_mask", "source_subject_refs": "person_1" }, { "id": "action_2", "mentions": [ { "text": "picks up", "ranges": [ [ 34, 42 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "type": "phrasal", "subject_refs": [ "person_1" ], "object_refs": [ "bellows_1" ], "confidence": "high", "temporal_status": "not_observed_in_sampled_frames", "intervals": [], "evidence": "At 0.000 seconds the selected person is already holding and positioning the bellows; no sampled frame captures the pickup transition.", "mask_relation": "subject_matches_selected_person_mask", "source_subject_refs": "person_1" }, { "id": "action_3", "mentions": [ { "text": "holds", "ranges": [ [ 117, 122 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "type": "simple", "subject_refs": [ "person_1" ], "object_refs": [ "box_1" ], "confidence": "high", "temporal_status": "supported", "intervals": [ { "start_s": 0.0, "end_s": 30.0, "source_start_s": 179.67, "source_end_s": 209.67 } ], "evidence": "Across samples 0.000–29.000 seconds, the mask-linked person visibly keeps the white rectangular box in one hand.", "mask_relation": "subject_matches_selected_person_mask", "source_subject_refs": "person_1" }, { "id": "action_4", "mentions": [ { "text": "aim", "ranges": [ [ 156, 159 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "type": "prepositional", "subject_refs": [ "person_1" ], "object_refs": [ "bellows_1", "fire_1" ], "confidence": "high", "temporal_status": "supported", "intervals": [ { "start_s": 0.0, "end_s": 30.0, "source_start_s": 179.67, "source_end_s": 209.67 } ], "evidence": "The selected person keeps the dark bellows directed toward the base of the visible fire throughout the sampled sequence.", "mask_relation": "subject_matches_selected_person_mask", "source_subject_refs": "person_1" }, { "id": "action_5", "mentions": [ { "text": "pump", "ranges": [ [ 211, 215 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } }, { "text": "work", "ranges": [ [ 282, 286 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "type": "simple", "subject_refs": [ "person_1" ], "object_refs": [ "bellows_1" ], "confidence": "high", "temporal_status": "supported", "intervals": [ { "start_s": 0.0, "end_s": 30.0, "source_start_s": 179.67, "source_end_s": 209.67 } ], "evidence": "Successive clean frames show the selected person's extended hand operating the bellows at the fire while the flames remain active and become stronger.", "mask_relation": "subject_matches_selected_person_mask", "source_subject_refs": "person_1" }, { "id": "action_6", "mentions": [ { "text": "stoke", "ranges": [ [ 222, 227 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "type": "simple", "subject_refs": [ "person_1" ], "object_refs": [ "fire_1" ], "confidence": "medium", "temporal_status": "supported", "intervals": [ { "start_s": 0.0, "end_s": 30.0, "source_start_s": 179.67, "source_end_s": 209.67 } ], "evidence": "The mask-linked person operates the bellows directly beside the fire, and the clean frames show sustained, increasingly prominent flames.", "mask_relation": "subject_matches_selected_person_mask", "source_subject_refs": "person_1" } ], "notes": [ "The supplied masklet consistently identifies the single person described by “The person,” “They,” and “they.”", "Only the person has a supplied source mask. The log, bellows, hands, box, lap, and fire/flames are text-groundable entities but have null source-mask links.", "The fire process expressed by “grow” is omitted from actions because its subject is the unmasked fire, not the selected masked person.", "The release and pickup events occur before the first sampled frame or between unsampled video moments; the sampled sequence begins with the person already using the bellows." ] }, "model_output_raw": { "caption": "The person lets go of the log and picks up a black, cylindrical bellows with their right hand, while their left hand holds the white box on their lap. They aim the bellows at the base of the fire and repeatedly pump it to stoke the flames. Throughout this period, they consistently work the bellows to help the fire grow stronger.", "entity_groups": [ { "id": "person_1", "label": "person", "spans": [ { "text": "The person", "range": [ 0, 10 ] }, { "text": "their", "range": [ 77, 82 ] }, { "text": "their", "range": [ 101, 106 ] }, { "text": "their", "range": [ 140, 145 ] }, { "text": "They", "range": [ 151, 155 ] }, { "text": "they", "range": [ 264, 268 ] } ], "source_mask": "selected_person_masklet" }, { "id": "log_1", "label": "log", "spans": [ { "text": "the log", "range": [ 22, 29 ] } ], "source_mask": null }, { "id": "bellows_1", "label": "bellows", "spans": [ { "text": "a black, cylindrical bellows", "range": [ 43, 71 ] }, { "text": "the bellows", "range": [ 160, 171 ] }, { "text": "it", "range": [ 216, 218 ] }, { "text": "the bellows", "range": [ 287, 298 ] } ], "source_mask": null }, { "id": "right_hand_1", "label": "right hand", "spans": [ { "text": "their right hand", "range": [ 77, 93 ] } ], "source_mask": null }, { "id": "left_hand_1", "label": "left hand", "spans": [ { "text": "their left hand", "range": [ 101, 116 ] } ], "source_mask": null }, { "id": "box_1", "label": "box", "spans": [ { "text": "the white box", "range": [ 123, 136 ] } ], "source_mask": null }, { "id": "lap_1", "label": "lap", "spans": [ { "text": "their lap", "range": [ 140, 149 ] } ], "source_mask": null }, { "id": "fire_1", "label": "fire", "spans": [ { "text": "the fire", "range": [ 187, 195 ] }, { "text": "the flames", "range": [ 228, 238 ] }, { "text": "the fire", "range": [ 307, 315 ] } ], "source_mask": null } ], "actions": [ { "id": "action_1", "mentions": [ { "text": "lets go", "ranges": [ [ 11, 18 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "type": "phrasal", "subject": "person_1", "objects": [ "log_1" ], "confidence": "high", "mask_relation": "subject_matches_selected_person_mask", "temporal_status": "not_observed_in_sampled_frames", "intervals": [], "evidence": "The selected person is visible throughout, but the first sampled frame already shows the fire setup and does not capture the release transition." }, { "id": "action_2", "mentions": [ { "text": "picks up", "ranges": [ [ 34, 42 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "type": "phrasal", "subject": "person_1", "objects": [ "bellows_1" ], "confidence": "high", "mask_relation": "subject_matches_selected_person_mask", "temporal_status": "not_observed_in_sampled_frames", "intervals": [], "evidence": "At 0.000 seconds the selected person is already holding and positioning the bellows; no sampled frame captures the pickup transition." }, { "id": "action_3", "mentions": [ { "text": "holds", "ranges": [ [ 117, 122 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "type": "simple", "subject": "person_1", "objects": [ "box_1" ], "confidence": "high", "mask_relation": "subject_matches_selected_person_mask", "temporal_status": "supported", "intervals": [ { "clip_relative": [ 0.0, 30.0 ], "source_video": [ 179.67, 209.67 ] } ], "evidence": "Across samples 0.000–29.000 seconds, the mask-linked person visibly keeps the white rectangular box in one hand." }, { "id": "action_4", "mentions": [ { "text": "aim", "ranges": [ [ 156, 159 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "type": "prepositional", "subject": "person_1", "objects": [ "bellows_1", "fire_1" ], "confidence": "high", "mask_relation": "subject_matches_selected_person_mask", "temporal_status": "supported", "intervals": [ { "clip_relative": [ 0.0, 30.0 ], "source_video": [ 179.67, 209.67 ] } ], "evidence": "The selected person keeps the dark bellows directed toward the base of the visible fire throughout the sampled sequence." }, { "id": "action_5", "mentions": [ { "text": "pump", "ranges": [ [ 211, 215 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } }, { "text": "work", "ranges": [ [ 282, 286 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "type": "simple", "subject": "person_1", "objects": [ "bellows_1" ], "confidence": "high", "mask_relation": "subject_matches_selected_person_mask", "temporal_status": "supported", "intervals": [ { "clip_relative": [ 0.0, 30.0 ], "source_video": [ 179.67, 209.67 ] } ], "evidence": "Successive clean frames show the selected person's extended hand operating the bellows at the fire while the flames remain active and become stronger." }, { "id": "action_6", "mentions": [ { "text": "stoke", "ranges": [ [ 222, 227 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "type": "simple", "subject": "person_1", "objects": [ "fire_1" ], "confidence": "medium", "mask_relation": "subject_matches_selected_person_mask", "temporal_status": "supported", "intervals": [ { "clip_relative": [ 0.0, 30.0 ], "source_video": [ 179.67, 209.67 ] } ], "evidence": "The mask-linked person operates the bellows directly beside the fire, and the clean frames show sustained, increasingly prominent flames." } ], "notes": [ "The supplied masklet consistently identifies the single person described by “The person,” “They,” and “they.”", "Only the person has a supplied source mask. The log, bellows, hands, box, lap, and fire/flames are text-groundable entities but have null source-mask links.", "The fire process expressed by “grow” is omitted from actions because its subject is the unmasked fire, not the selected masked person.", "The release and pickup events occur before the first sampled frame or between unsampled video moments; the sampled sequence begins with the person already using the bellows." ] }, "normalization_repairs": 0 }, { "source_video_id": "v_-Sp-gOEGy2o", "caption_count_category": 13, "clip": { "caption_index": 13, "caption": "The person stops using the bellows and lowers it, observing the now well-established fire. They briefly bring the bellows back up towards their lap before the scene abruptly darkens. They remain seated by the fire pit, holding the white box and the bellows.", "source_interval_sec": [ 209.67, 219.62 ], "loom_clip_id": "v_-Sp-gOEGy2o_6284_6582", "clip_frame_interval_4fps": [ 6284, 6582 ], "sample_count_1fps": 10, "sampled_frames": [ { "frame": "00000", "timestamp_sec": 209.67, "offset_sec": 0, "rgb_path": "assets/cat_13/v_-Sp-gOEGy2o_6284_6582/0000s_00000_rgb.jpg", "mask_path": "assets/cat_13/v_-Sp-gOEGy2o_6284_6582/0000s_00000_mask.png", "mask_status": "available" }, { "frame": "00028", "timestamp_sec": 210.67, "offset_sec": 1, "rgb_path": "assets/cat_13/v_-Sp-gOEGy2o_6284_6582/0001s_00028_rgb.jpg", "mask_path": "assets/cat_13/v_-Sp-gOEGy2o_6284_6582/0001s_00028_mask.png", "mask_status": "available" }, { "frame": "00060", "timestamp_sec": 211.67, "offset_sec": 2, "rgb_path": "assets/cat_13/v_-Sp-gOEGy2o_6284_6582/0002s_00060_rgb.jpg", "mask_path": "assets/cat_13/v_-Sp-gOEGy2o_6284_6582/0002s_00060_mask.png", "mask_status": "available" }, { "frame": "00088", "timestamp_sec": 212.67, "offset_sec": 3, "rgb_path": "assets/cat_13/v_-Sp-gOEGy2o_6284_6582/0003s_00088_rgb.jpg", "mask_path": "assets/cat_13/v_-Sp-gOEGy2o_6284_6582/0003s_00088_mask.png", "mask_status": "available" }, { "frame": "00120", "timestamp_sec": 213.67, "offset_sec": 4, "rgb_path": "assets/cat_13/v_-Sp-gOEGy2o_6284_6582/0004s_00120_rgb.jpg", "mask_path": "assets/cat_13/v_-Sp-gOEGy2o_6284_6582/0004s_00120_mask.png", "mask_status": "available" }, { "frame": "00148", "timestamp_sec": 214.67, "offset_sec": 5, "rgb_path": "assets/cat_13/v_-Sp-gOEGy2o_6284_6582/0005s_00148_rgb.jpg", "mask_path": "assets/cat_13/v_-Sp-gOEGy2o_6284_6582/0005s_00148_mask.png", "mask_status": "available" }, { "frame": "00180", "timestamp_sec": 215.67, "offset_sec": 6, "rgb_path": "assets/cat_13/v_-Sp-gOEGy2o_6284_6582/0006s_00180_rgb.jpg", "mask_path": "assets/cat_13/v_-Sp-gOEGy2o_6284_6582/0006s_00180_mask.png", "mask_status": "available" }, { "frame": "00208", "timestamp_sec": 216.67, "offset_sec": 7, "rgb_path": "assets/cat_13/v_-Sp-gOEGy2o_6284_6582/0007s_00208_rgb.jpg", "mask_path": "assets/cat_13/v_-Sp-gOEGy2o_6284_6582/0007s_00208_mask.png", "mask_status": "available" }, { "frame": "00240", "timestamp_sec": 217.67, "offset_sec": 8, "rgb_path": "assets/cat_13/v_-Sp-gOEGy2o_6284_6582/0008s_00240_rgb.jpg", "mask_path": "assets/cat_13/v_-Sp-gOEGy2o_6284_6582/0008s_00240_mask.png", "mask_status": "available" }, { "frame": "00268", "timestamp_sec": 218.67, "offset_sec": 9, "rgb_path": "assets/cat_13/v_-Sp-gOEGy2o_6284_6582/0009s_00268_rgb.jpg", "mask_path": "assets/cat_13/v_-Sp-gOEGy2o_6284_6582/0009s_00268_mask.png", "mask_status": "available" } ], "obj_id": [ 0 ], "anno_id": [ 79 ], "fps_inferred_from_clip_interval": 29.9497 }, "usage": { "input_tokens": 37973, "input_tokens_details": { "cache_write_tokens": 37970, "cached_tokens": 0 }, "output_tokens": 4438, "output_tokens_details": { "reasoning_tokens": 2532 }, "total_tokens": 42411 }, "model": "gpt-5.6-sol", "effort": "medium", "custom_id": "cat13_cap13_v_-Sp-gOEGy2o_6284_6582", "annotation": { "caption": "The person stops using the bellows and lowers it, observing the now well-established fire. They briefly bring the bellows back up towards their lap before the scene abruptly darkens. They remain seated by the fire pit, holding the white box and the bellows.", "entity_groups": [ { "id": "person_1", "label": "person", "spans": [ { "text": "The person", "range": [ 0, 10 ] }, { "text": "They", "range": [ 91, 95 ] }, { "text": "their", "range": [ 138, 143 ] }, { "text": "They", "range": [ 183, 187 ] } ], "source_mask_link": "selected_person" }, { "id": "bellows_1", "label": "bellows", "spans": [ { "text": "the bellows", "range": [ 23, 34 ] }, { "text": "it", "range": [ 46, 48 ] }, { "text": "the bellows", "range": [ 110, 121 ] }, { "text": "the bellows", "range": [ 245, 256 ] } ], "source_mask_link": null }, { "id": "fire_1", "label": "fire", "spans": [ { "text": "the now well-established fire", "range": [ 60, 89 ] } ], "source_mask_link": null }, { "id": "lap_1", "label": "lap", "spans": [ { "text": "their lap", "range": [ 138, 147 ] } ], "source_mask_link": null }, { "id": "fire_pit_1", "label": "fire pit", "spans": [ { "text": "the fire pit", "range": [ 205, 217 ] } ], "source_mask_link": null }, { "id": "box_1", "label": "box", "spans": [ { "text": "the white box", "range": [ 227, 240 ] } ], "source_mask_link": null } ], "actions": [ { "id": "action_1", "mentions": [ { "text": "using", "ranges": [ [ 17, 22 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "type": "simple", "subject_refs": [ "person_1" ], "object_refs": [ "bellows_1" ], "confidence": "medium", "temporal_status": "localized", "intervals": [ { "start_s": 0.0, "end_s": 2.0, "source_start_s": 209.67, "source_end_s": 211.67 } ], "evidence": "At 0.000–1.000 seconds, the masked person holds the black bellows down beside the active fire, consistent with the final moments of using it.", "mask_relation": "subject_matches_supplied_selected_person_mask", "source_subject_refs": [ "person_1" ] }, { "id": "action_2", "mentions": [ { "text": "lowers", "ranges": [ [ 39, 45 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "type": "simple", "subject_refs": [ "person_1" ], "object_refs": [ "bellows_1" ], "confidence": "medium", "temporal_status": "localized", "intervals": [ { "start_s": 0.0, "end_s": 3.0, "source_start_s": 209.67, "source_end_s": 212.67 } ], "evidence": "Across the 0.000–2.000 second frames, the masked person's bellows-holding hand withdraws from the fire area toward a lower resting position.", "mask_relation": "subject_matches_supplied_selected_person_mask", "source_subject_refs": [ "person_1" ] }, { "id": "action_3", "mentions": [ { "text": "observing", "ranges": [ [ 50, 59 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "type": "simple", "subject_refs": [ "person_1" ], "object_refs": [ "fire_1" ], "confidence": "high", "temporal_status": "localized", "intervals": [ { "start_s": 0.0, "end_s": 5.0, "source_start_s": 209.67, "source_end_s": 214.67 } ], "evidence": "From 0.000 through 4.000 seconds, the masked person remains oriented downward toward the visibly burning fire.", "mask_relation": "subject_matches_supplied_selected_person_mask", "source_subject_refs": [ "person_1" ] }, { "id": "action_4", "mentions": [ { "text": "bring", "ranges": [ [ 104, 109 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "type": "simple", "subject_refs": [ "person_1" ], "object_refs": [ "bellows_1" ], "confidence": "high", "temporal_status": "localized", "intervals": [ { "start_s": 3.0, "end_s": 5.0, "source_start_s": 212.67, "source_end_s": 214.67 } ], "evidence": "At 3.000–4.000 seconds, the masked person raises the black bellows back toward his lap and upper body.", "mask_relation": "subject_matches_supplied_selected_person_mask", "source_subject_refs": [ "person_1" ] }, { "id": "action_5", "mentions": [ { "text": "seated", "ranges": [ [ 195, 201 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "type": "posture", "subject_refs": [ "person_1" ], "object_refs": [], "confidence": "high", "temporal_status": "localized", "intervals": [ { "start_s": 0.0, "end_s": 9.950000000000017, "source_start_s": 209.67, "source_end_s": 219.62 } ], "evidence": "Every sampled frame from 0.000 through 9.000 seconds shows the masked person maintaining a seated posture beside the fire pit.", "mask_relation": "subject_matches_supplied_selected_person_mask", "source_subject_refs": [ "person_1" ] }, { "id": "action_6", "mentions": [ { "text": "holding", "ranges": [ [ 219, 226 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "type": "simple", "subject_refs": [ "person_1" ], "object_refs": [ "box_1", "bellows_1" ], "confidence": "high", "temporal_status": "localized", "intervals": [ { "start_s": 0.0, "end_s": 9.950000000000017, "source_start_s": 209.67, "source_end_s": 219.62 } ], "evidence": "The masked person visibly retains the small white box in one hand and the black bellows in the other or against his lap throughout the sampled sequence.", "mask_relation": "subject_matches_supplied_selected_person_mask", "source_subject_refs": [ "person_1" ] } ], "notes": [ "The supplied mask consistently identifies the sole person, so all included actions have a mask-linked subject.", "No separate masks were supplied for the bellows, fire, lap, fire pit, or white box; their source-mask links are null.", "The abstract phrase \"the scene\" is not an entity group. Its darkening predicate is omitted because its subject is not the supplied masked person and it describes a scene/view transition.", "\"stops\" and \"remain\" are excluded as aspectual framing predicates; the embedded action \"using\" and posture \"seated\" are retained.", "In \"bring the bellows back up,\" \"back\" and \"up\" express direction and are excluded from the action span." ] }, "model_output_raw": { "caption": "The person stops using the bellows and lowers it, observing the now well-established fire. They briefly bring the bellows back up towards their lap before the scene abruptly darkens. They remain seated by the fire pit, holding the white box and the bellows.", "entity_groups": [ { "id": "person_1", "label": "person", "spans": [ { "text": "The person", "range": [ 0, 10 ] }, { "text": "They", "range": [ 91, 95 ] }, { "text": "their", "range": [ 138, 143 ] }, { "text": "They", "range": [ 183, 187 ] } ], "source_mask_link": "selected_person_masklet" }, { "id": "bellows_1", "label": "bellows", "spans": [ { "text": "the bellows", "range": [ 23, 34 ] }, { "text": "it", "range": [ 46, 48 ] }, { "text": "the bellows", "range": [ 110, 121 ] }, { "text": "the bellows", "range": [ 245, 256 ] } ], "source_mask_link": null }, { "id": "fire_1", "label": "fire", "spans": [ { "text": "the now well-established fire", "range": [ 60, 89 ] } ], "source_mask_link": null }, { "id": "lap_1", "label": "lap", "spans": [ { "text": "their lap", "range": [ 138, 147 ] } ], "source_mask_link": null }, { "id": "fire_pit_1", "label": "fire pit", "spans": [ { "text": "the fire pit", "range": [ 205, 217 ] } ], "source_mask_link": null }, { "id": "box_1", "label": "box", "spans": [ { "text": "the white box", "range": [ 227, 240 ] } ], "source_mask_link": null } ], "actions": [ { "id": "action_1", "mentions": [ { "text": "using", "occurrence": 1, "ranges": [ [ 17, 22 ] ], "particle": { "text": null, "occurrence": 1 } } ], "type": "simple", "subject_refs": [ "person_1" ], "object_refs": [ "bellows_1" ], "confidence": "medium", "temporal_status": "localized", "intervals": [ { "start_s": 0.0, "end_s": 2.0 } ], "evidence": "At 0.000–1.000 seconds, the masked person holds the black bellows down beside the active fire, consistent with the final moments of using it.", "mask_relation": "subject_matches_supplied_selected_person_mask" }, { "id": "action_2", "mentions": [ { "text": "lowers", "occurrence": 1, "ranges": [ [ 39, 45 ] ], "particle": { "text": null, "occurrence": 1 } } ], "type": "simple", "subject_refs": [ "person_1" ], "object_refs": [ "bellows_1" ], "confidence": "medium", "temporal_status": "localized", "intervals": [ { "start_s": 0.0, "end_s": 3.0 } ], "evidence": "Across the 0.000–2.000 second frames, the masked person's bellows-holding hand withdraws from the fire area toward a lower resting position.", "mask_relation": "subject_matches_supplied_selected_person_mask" }, { "id": "action_3", "mentions": [ { "text": "observing", "occurrence": 1, "ranges": [ [ 50, 59 ] ], "particle": { "text": null, "occurrence": 1 } } ], "type": "simple", "subject_refs": [ "person_1" ], "object_refs": [ "fire_1" ], "confidence": "high", "temporal_status": "localized", "intervals": [ { "start_s": 0.0, "end_s": 5.0 } ], "evidence": "From 0.000 through 4.000 seconds, the masked person remains oriented downward toward the visibly burning fire.", "mask_relation": "subject_matches_supplied_selected_person_mask" }, { "id": "action_4", "mentions": [ { "text": "bring", "occurrence": 1, "ranges": [ [ 104, 109 ] ], "particle": { "text": null, "occurrence": 1 } } ], "type": "simple", "subject_refs": [ "person_1" ], "object_refs": [ "bellows_1" ], "confidence": "high", "temporal_status": "localized", "intervals": [ { "start_s": 3.0, "end_s": 5.0 } ], "evidence": "At 3.000–4.000 seconds, the masked person raises the black bellows back toward his lap and upper body.", "mask_relation": "subject_matches_supplied_selected_person_mask" }, { "id": "action_5", "mentions": [ { "text": "seated", "occurrence": 1, "ranges": [ [ 195, 201 ] ], "particle": { "text": null, "occurrence": 1 } } ], "type": "posture", "subject_refs": [ "person_1" ], "object_refs": [], "confidence": "high", "temporal_status": "localized", "intervals": [ { "start_s": 0.0, "end_s": 9.95 } ], "evidence": "Every sampled frame from 0.000 through 9.000 seconds shows the masked person maintaining a seated posture beside the fire pit.", "mask_relation": "subject_matches_supplied_selected_person_mask" }, { "id": "action_6", "mentions": [ { "text": "holding", "occurrence": 1, "ranges": [ [ 219, 226 ] ], "particle": { "text": null, "occurrence": 1 } } ], "type": "simple", "subject_refs": [ "person_1" ], "object_refs": [ "box_1", "bellows_1" ], "confidence": "high", "temporal_status": "localized", "intervals": [ { "start_s": 0.0, "end_s": 9.95 } ], "evidence": "The masked person visibly retains the small white box in one hand and the black bellows in the other or against his lap throughout the sampled sequence.", "mask_relation": "subject_matches_supplied_selected_person_mask" } ], "notes": [ "The supplied mask consistently identifies the sole person, so all included actions have a mask-linked subject.", "No separate masks were supplied for the bellows, fire, lap, fire pit, or white box; their source-mask links are null.", "The abstract phrase \"the scene\" is not an entity group. Its darkening predicate is omitted because its subject is not the supplied masked person and it describes a scene/view transition.", "\"stops\" and \"remain\" are excluded as aspectual framing predicates; the embedded action \"using\" and posture \"seated\" are retained.", "In \"bring the bellows back up,\" \"back\" and \"up\" express direction and are excluded from the action span." ] }, "normalization_repairs": 0 }, { "source_video_id": "v_9A0BwCxAKAg", "caption_count_category": 14, "clip": { "caption_index": 1, "caption": "A person with dark hair, a goatee, and an earring in his left ear stands on the left side of the frame. He is wearing a black short-sleeved collared shirt and is speaking while looking slightly to his left.", "source_interval_sec": [ 0.0, 6.5 ], "loom_clip_id": "v_9A0BwCxAKAg_1_195", "clip_frame_interval_4fps": [ 1, 195 ], "sample_count_1fps": 7, "sampled_frames": [ { "frame": "00000", "timestamp_sec": 0.0, "offset_sec": 0, "rgb_path": "assets/cat_14/v_9A0BwCxAKAg_1_195/0000s_00000_rgb.jpg", "mask_path": "assets/cat_14/v_9A0BwCxAKAg_1_195/0000s_00000_mask.png", "mask_status": "available" }, { "frame": "00028", "timestamp_sec": 1.0, "offset_sec": 1, "rgb_path": "assets/cat_14/v_9A0BwCxAKAg_1_195/0001s_00028_rgb.jpg", "mask_path": "assets/cat_14/v_9A0BwCxAKAg_1_195/0001s_00028_mask.png", "mask_status": "available" }, { "frame": "00060", "timestamp_sec": 2.0, "offset_sec": 2, "rgb_path": "assets/cat_14/v_9A0BwCxAKAg_1_195/0002s_00060_rgb.jpg", "mask_path": "assets/cat_14/v_9A0BwCxAKAg_1_195/0002s_00060_mask.png", "mask_status": "available" }, { "frame": "00088", "timestamp_sec": 3.0, "offset_sec": 3, "rgb_path": "assets/cat_14/v_9A0BwCxAKAg_1_195/0003s_00088_rgb.jpg", "mask_path": "assets/cat_14/v_9A0BwCxAKAg_1_195/0003s_00088_mask.png", "mask_status": "available" }, { "frame": "00120", "timestamp_sec": 4.0, "offset_sec": 4, "rgb_path": "assets/cat_14/v_9A0BwCxAKAg_1_195/0004s_00120_rgb.jpg", "mask_path": "assets/cat_14/v_9A0BwCxAKAg_1_195/0004s_00120_mask.png", "mask_status": "available" }, { "frame": "00148", "timestamp_sec": 5.0, "offset_sec": 5, "rgb_path": "assets/cat_14/v_9A0BwCxAKAg_1_195/0005s_00148_rgb.jpg", "mask_path": "assets/cat_14/v_9A0BwCxAKAg_1_195/0005s_00148_mask.png", "mask_status": "available" }, { "frame": "00180", "timestamp_sec": 6.0, "offset_sec": 6, "rgb_path": "assets/cat_14/v_9A0BwCxAKAg_1_195/0006s_00180_rgb.jpg", "mask_path": "assets/cat_14/v_9A0BwCxAKAg_1_195/0006s_00180_mask.png", "mask_status": "available" } ], "obj_id": [ 0 ], "anno_id": [ 1267 ], "fps_inferred_from_clip_interval": 29.8462 }, "usage": { "input_tokens": 20764, "input_tokens_details": { "cache_write_tokens": 20761, "cached_tokens": 0 }, "output_tokens": 3577, "output_tokens_details": { "reasoning_tokens": 2070 }, "total_tokens": 24341 }, "model": "gpt-5.6-sol", "effort": "medium", "custom_id": "cat14_cap01_v_9A0BwCxAKAg_1_195", "annotation": { "caption": "A person with dark hair, a goatee, and an earring in his left ear stands on the left side of the frame. He is wearing a black short-sleeved collared shirt and is speaking while looking slightly to his left.", "entity_groups": [ { "id": "entity_1", "label": "person", "spans": [ { "text": "A person", "range": [ 0, 8 ] }, { "text": "his", "range": [ 53, 56 ] }, { "text": "He", "range": [ 104, 106 ] }, { "text": "his", "range": [ 197, 200 ] } ], "source_mask_link": "selected_person" }, { "id": "entity_2", "label": "hair", "spans": [ { "text": "dark hair", "range": [ 14, 23 ] } ], "source_mask_link": null }, { "id": "entity_3", "label": "goatee", "spans": [ { "text": "a goatee", "range": [ 25, 33 ] } ], "source_mask_link": null }, { "id": "entity_4", "label": "earring", "spans": [ { "text": "an earring", "range": [ 39, 49 ] } ], "source_mask_link": null }, { "id": "entity_5", "label": "ear", "spans": [ { "text": "his left ear", "range": [ 53, 65 ] } ], "source_mask_link": null }, { "id": "entity_6", "label": "shirt", "spans": [ { "text": "a black short-sleeved collared shirt", "range": [ 118, 154 ] } ], "source_mask_link": null } ], "actions": [ { "id": "action_1", "mentions": [ { "text": "stands", "ranges": [ [ 66, 72 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "type": "posture", "subject_refs": [ "entity_1" ], "object_refs": [], "confidence": "medium", "temporal_status": "localized", "intervals": [ { "start_s": 0.0, "end_s": 6.5, "source_start_s": 0.0, "source_end_s": 6.5 } ], "evidence": "The masked person maintains an upright posture at the left side in every sampled frame; the lower body is cropped.", "mask_relation": "subject matches the supplied selected-person mask", "source_subject_refs": "entity_1" }, { "id": "action_2", "mentions": [ { "text": "is wearing", "ranges": [ [ 107, 117 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "type": "simple", "subject_refs": [ "entity_1" ], "object_refs": [ "entity_6" ], "confidence": "high", "temporal_status": "localized", "intervals": [ { "start_s": 0.0, "end_s": 6.5, "source_start_s": 0.0, "source_end_s": 6.5 } ], "evidence": "The masked person visibly wears the same black collared shirt throughout all sampled frames.", "mask_relation": "subject matches the supplied selected-person mask", "source_subject_refs": "entity_1" }, { "id": "action_3", "mentions": [ { "text": "is speaking", "ranges": [ [ 159, 170 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "type": "simple", "subject_refs": [ "entity_1" ], "object_refs": [], "confidence": "medium", "temporal_status": "localized_sparse", "intervals": [ { "start_s": 0.0, "end_s": 1.0, "source_start_s": 0.0, "source_end_s": 1.0 }, { "start_s": 2.0, "end_s": 3.0, "source_start_s": 2.0, "source_end_s": 3.0 }, { "start_s": 5.0, "end_s": 6.5, "source_start_s": 5.0, "source_end_s": 6.5 } ], "evidence": "Visible open-mouth configurations support speech at samples 0, 2, 5, and 6; still frames without clear mouth movement are not used as support.", "mask_relation": "subject matches the supplied selected-person mask", "source_subject_refs": "entity_1" }, { "id": "action_4", "mentions": [ { "text": "looking", "ranges": [ [ 177, 184 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "type": "prepositional", "subject_refs": [ "entity_1" ], "object_refs": [], "confidence": "medium", "temporal_status": "localized", "intervals": [ { "start_s": 0.0, "end_s": 6.5, "source_start_s": 0.0, "source_end_s": 6.5 } ], "evidence": "Across the sampled frames, the masked person's head and gaze remain angled toward his left; the directional phrase is excluded from the action span.", "mask_relation": "subject matches the supplied selected-person mask", "source_subject_refs": "entity_1" } ], "notes": [ "The supplied masklet consistently identifies the single person in all seven sampled frames.", "No separate source masks are supplied for the hair, goatee, earring, ear, or shirt, so those entity groups have null source-mask links.", "The phrase \"the frame\" is excluded as a meta-reference to the image/video medium.", "Speech support is sparse because isolated clean frames can show mouth configurations but cannot establish continuous vocalization." ] }, "model_output_raw": { "caption": "A person with dark hair, a goatee, and an earring in his left ear stands on the left side of the frame. He is wearing a black short-sleeved collared shirt and is speaking while looking slightly to his left.", "entity_groups": [ { "id": "entity_1", "label": "person", "spans": [ { "text": "A person", "range": [ 0, 8 ] }, { "text": "his", "range": [ 53, 56 ] }, { "text": "He", "range": [ 104, 106 ] }, { "text": "his", "range": [ 197, 200 ] } ], "source_mask": "selected_person_masklet" }, { "id": "entity_2", "label": "hair", "spans": [ { "text": "dark hair", "range": [ 14, 23 ] } ], "source_mask": null }, { "id": "entity_3", "label": "goatee", "spans": [ { "text": "a goatee", "range": [ 25, 33 ] } ], "source_mask": null }, { "id": "entity_4", "label": "earring", "spans": [ { "text": "an earring", "range": [ 39, 49 ] } ], "source_mask": null }, { "id": "entity_5", "label": "ear", "spans": [ { "text": "his left ear", "range": [ 53, 65 ] } ], "source_mask": null }, { "id": "entity_6", "label": "shirt", "spans": [ { "text": "a black short-sleeved collared shirt", "range": [ 118, 154 ] } ], "source_mask": null } ], "actions": [ { "mentions": [ { "text": "stands", "ranges": [ [ 66, 72 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "type": "posture", "subject": "entity_1", "objects": [], "confidence": "medium", "mask_relation": "subject matches the supplied selected-person mask", "temporal_status": "localized", "intervals": [ { "start_s": 0.0, "end_s": 6.5, "source_start_s": 0.0, "source_end_s": 6.5 } ], "evidence": "The masked person maintains an upright posture at the left side in every sampled frame; the lower body is cropped." }, { "mentions": [ { "text": "is wearing", "ranges": [ [ 107, 117 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "type": "simple", "subject": "entity_1", "objects": [ "entity_6" ], "confidence": "high", "mask_relation": "subject matches the supplied selected-person mask", "temporal_status": "localized", "intervals": [ { "start_s": 0.0, "end_s": 6.5, "source_start_s": 0.0, "source_end_s": 6.5 } ], "evidence": "The masked person visibly wears the same black collared shirt throughout all sampled frames." }, { "mentions": [ { "text": "is speaking", "ranges": [ [ 159, 170 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "type": "simple", "subject": "entity_1", "objects": [], "confidence": "medium", "mask_relation": "subject matches the supplied selected-person mask", "temporal_status": "localized_sparse", "intervals": [ { "start_s": 0.0, "end_s": 1.0, "source_start_s": 0.0, "source_end_s": 1.0 }, { "start_s": 2.0, "end_s": 3.0, "source_start_s": 2.0, "source_end_s": 3.0 }, { "start_s": 5.0, "end_s": 6.5, "source_start_s": 5.0, "source_end_s": 6.5 } ], "evidence": "Visible open-mouth configurations support speech at samples 0, 2, 5, and 6; still frames without clear mouth movement are not used as support." }, { "mentions": [ { "text": "looking", "ranges": [ [ 177, 184 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "type": "prepositional", "subject": "entity_1", "objects": [], "confidence": "medium", "mask_relation": "subject matches the supplied selected-person mask", "temporal_status": "localized", "intervals": [ { "start_s": 0.0, "end_s": 6.5, "source_start_s": 0.0, "source_end_s": 6.5 } ], "evidence": "Across the sampled frames, the masked person's head and gaze remain angled toward his left; the directional phrase is excluded from the action span." } ], "notes": [ "The supplied masklet consistently identifies the single person in all seven sampled frames.", "No separate source masks are supplied for the hair, goatee, earring, ear, or shirt, so those entity groups have null source-mask links.", "The phrase \"the frame\" is excluded as a meta-reference to the image/video medium.", "Speech support is sparse because isolated clean frames can show mouth configurations but cannot establish continuous vocalization." ] }, "normalization_repairs": 0 }, { "source_video_id": "v_9A0BwCxAKAg", "caption_count_category": 14, "clip": { "caption_index": 2, "caption": "The person turns his body to his right to face a large window pane. He gestures with his left hand, pointing up and then across the window, before turning back to his left.", "source_interval_sec": [ 6.5, 15.0 ], "loom_clip_id": "v_9A0BwCxAKAg_195_450", "clip_frame_interval_4fps": [ 195, 450 ], "sample_count_1fps": 9, "sampled_frames": [ { "frame": "00000", "timestamp_sec": 6.5, "offset_sec": 0, "rgb_path": "assets/cat_14/v_9A0BwCxAKAg_195_450/0000s_00000_rgb.jpg", "mask_path": "assets/cat_14/v_9A0BwCxAKAg_195_450/0000s_00000_mask.png", "mask_status": "available" }, { "frame": "00028", "timestamp_sec": 7.5, "offset_sec": 1, "rgb_path": "assets/cat_14/v_9A0BwCxAKAg_195_450/0001s_00028_rgb.jpg", "mask_path": "assets/cat_14/v_9A0BwCxAKAg_195_450/0001s_00028_mask.png", "mask_status": "available" }, { "frame": "00060", "timestamp_sec": 8.5, "offset_sec": 2, "rgb_path": "assets/cat_14/v_9A0BwCxAKAg_195_450/0002s_00060_rgb.jpg", "mask_path": "assets/cat_14/v_9A0BwCxAKAg_195_450/0002s_00060_mask.png", "mask_status": "available" }, { "frame": "00088", "timestamp_sec": 9.5, "offset_sec": 3, "rgb_path": "assets/cat_14/v_9A0BwCxAKAg_195_450/0003s_00088_rgb.jpg", "mask_path": "assets/cat_14/v_9A0BwCxAKAg_195_450/0003s_00088_mask.png", "mask_status": "available" }, { "frame": "00120", "timestamp_sec": 10.5, "offset_sec": 4, "rgb_path": "assets/cat_14/v_9A0BwCxAKAg_195_450/0004s_00120_rgb.jpg", "mask_path": "assets/cat_14/v_9A0BwCxAKAg_195_450/0004s_00120_mask.png", "mask_status": "available" }, { "frame": "00148", "timestamp_sec": 11.5, "offset_sec": 5, "rgb_path": "assets/cat_14/v_9A0BwCxAKAg_195_450/0005s_00148_rgb.jpg", "mask_path": "assets/cat_14/v_9A0BwCxAKAg_195_450/0005s_00148_mask.png", "mask_status": "available" }, { "frame": "00180", "timestamp_sec": 12.5, "offset_sec": 6, "rgb_path": "assets/cat_14/v_9A0BwCxAKAg_195_450/0006s_00180_rgb.jpg", "mask_path": "assets/cat_14/v_9A0BwCxAKAg_195_450/0006s_00180_mask.png", "mask_status": "available" }, { "frame": "00208", "timestamp_sec": 13.5, "offset_sec": 7, "rgb_path": "assets/cat_14/v_9A0BwCxAKAg_195_450/0007s_00208_rgb.jpg", "mask_path": "assets/cat_14/v_9A0BwCxAKAg_195_450/0007s_00208_mask.png", "mask_status": "available" }, { "frame": "00240", "timestamp_sec": 14.5, "offset_sec": 8, "rgb_path": "assets/cat_14/v_9A0BwCxAKAg_195_450/0008s_00240_rgb.jpg", "mask_path": "assets/cat_14/v_9A0BwCxAKAg_195_450/0008s_00240_mask.png", "mask_status": "available" } ], "obj_id": [ 0 ], "anno_id": [ 1266 ], "fps_inferred_from_clip_interval": 30.0 }, "usage": { "input_tokens": 22302, "input_tokens_details": { "cache_write_tokens": 22299, "cached_tokens": 0 }, "output_tokens": 4055, "output_tokens_details": { "reasoning_tokens": 2588 }, "total_tokens": 26357 }, "model": "gpt-5.6-sol", "effort": "medium", "custom_id": "cat14_cap02_v_9A0BwCxAKAg_195_450", "annotation": { "caption": "The person turns his body to his right to face a large window pane. He gestures with his left hand, pointing up and then across the window, before turning back to his left.", "entity_groups": [ { "id": "person_1", "label": "person", "spans": [ { "text": "The person", "range": [ 0, 10 ] }, { "text": "his", "range": [ 17, 20 ] }, { "text": "his", "range": [ 29, 32 ] }, { "text": "He", "range": [ 68, 70 ] }, { "text": "his", "range": [ 85, 88 ] }, { "text": "his", "range": [ 163, 166 ] } ], "source_mask_link": "selected_person" }, { "id": "body_1", "label": "body", "spans": [ { "text": "his body", "range": [ 17, 25 ] } ], "source_mask_link": null }, { "id": "window_pane_1", "label": "window pane", "spans": [ { "text": "a large window pane", "range": [ 47, 66 ] }, { "text": "the window", "range": [ 128, 138 ] } ], "source_mask_link": null }, { "id": "left_hand_1", "label": "left hand", "spans": [ { "text": "his left hand", "range": [ 85, 98 ] } ], "source_mask_link": null } ], "actions": [ { "id": "action_1", "mentions": [ { "text": "turns", "ranges": [ [ 11, 16 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "type": "simple", "subject_refs": [ "person_1" ], "object_refs": [ "body_1" ], "confidence": "medium", "temporal_status": "localized_sparse_support", "intervals": [ { "start_s": 0.0, "end_s": 1.0, "source_start_s": 6.5, "source_end_s": 7.5 } ], "evidence": "At clip-relative 0.0 s (source 6.5 s) the selected person is oriented partly toward the viewer; by 1.0 s (source 7.5 s) his torso and head are oriented toward the window.", "mask_relation": "subject matches the supplied selected-person mask", "source_subject_refs": [ "person_1" ] }, { "id": "action_2", "mentions": [ { "text": "face", "ranges": [ [ 42, 46 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "type": "posture", "subject_refs": [ "person_1" ], "object_refs": [ "window_pane_1" ], "confidence": "high", "temporal_status": "localized_sparse_support", "intervals": [ { "start_s": 1.0, "end_s": 7.0, "source_start_s": 7.5, "source_end_s": 13.5 } ], "evidence": "Across samples from 1.0 through 6.0 s (source 7.5–12.5 s), the selected person's head and torso remain oriented toward the large window.", "mask_relation": "subject matches the supplied selected-person mask", "source_subject_refs": [ "person_1" ] }, { "id": "action_3", "mentions": [ { "text": "gestures", "ranges": [ [ 71, 79 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } }, { "text": "pointing", "ranges": [ [ 100, 108 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "type": "simple", "subject_refs": [ "person_1" ], "object_refs": [ "window_pane_1" ], "confidence": "high", "temporal_status": "localized_sparse_support", "intervals": [ { "start_s": 1.0, "end_s": 7.0, "source_start_s": 7.5, "source_end_s": 13.5 } ], "evidence": "Samples from 1.0 through 6.0 s show the mask-linked person repeatedly raising and extending his left arm and hand upward and across the window.", "mask_relation": "subject matches the supplied selected-person mask", "source_subject_refs": [ "person_1" ] }, { "id": "action_4", "mentions": [ { "text": "turning", "ranges": [ [ 147, 154 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "type": "simple", "subject_refs": [ "person_1" ], "object_refs": [], "confidence": "medium", "temporal_status": "localized_sparse_support", "intervals": [ { "start_s": 7.0, "end_s": 8.5, "source_start_s": 13.5, "source_end_s": 15.0 } ], "evidence": "Between the 7.0 and 8.0 s samples (source 13.5–14.5 s), the selected person's torso and head change away from the window toward his left; the final sampled pose supports continuation to the clip boundary.", "mask_relation": "subject matches the supplied selected-person mask", "source_subject_refs": [ "person_1" ] } ], "notes": "The supplied mask consistently identifies the captioned person. No separate source masks were supplied for the person's body, left hand, or the window pane, so those entity groups have null source-mask links. Directional references such as 'his right' and 'his left' are spatial concepts and are not entity groups. 'gestures' and 'pointing' are grouped as two mentions of the same elaborated hand-gesture event." }, "model_output_raw": { "caption": "The person turns his body to his right to face a large window pane. He gestures with his left hand, pointing up and then across the window, before turning back to his left.", "entity_groups": [ { "id": "person_1", "entity": "person", "spans": [ { "text": "The person", "range": [ 0, 10 ] }, { "text": "his", "range": [ 17, 20 ] }, { "text": "his", "range": [ 29, 32 ] }, { "text": "He", "range": [ 68, 70 ] }, { "text": "his", "range": [ 85, 88 ] }, { "text": "his", "range": [ 163, 166 ] } ], "source_mask": "selected_person_mask" }, { "id": "body_1", "entity": "body", "spans": [ { "text": "his body", "range": [ 17, 25 ] } ], "source_mask": null }, { "id": "window_pane_1", "entity": "window pane", "spans": [ { "text": "a large window pane", "range": [ 47, 66 ] }, { "text": "the window", "range": [ 128, 138 ] } ], "source_mask": null }, { "id": "left_hand_1", "entity": "left hand", "spans": [ { "text": "his left hand", "range": [ 85, 98 ] } ], "source_mask": null } ], "actions": [ { "id": "action_1", "mentions": [ { "text": "turns", "occurrence": 1, "ranges": [ [ 11, 16 ] ], "particle": { "text": null, "occurrence": 1 } } ], "type": "simple", "subject_refs": [ "person_1" ], "object_refs": [ "body_1" ], "confidence": "medium", "temporal_status": "localized_sparse_support", "intervals": [ { "start_s": 0.0, "end_s": 1.0 } ], "evidence": "At clip-relative 0.0 s (source 6.5 s) the selected person is oriented partly toward the viewer; by 1.0 s (source 7.5 s) his torso and head are oriented toward the window.", "mask_relation": "subject matches the supplied selected-person mask" }, { "id": "action_2", "mentions": [ { "text": "face", "occurrence": 1, "ranges": [ [ 42, 46 ] ], "particle": { "text": null, "occurrence": 1 } } ], "type": "posture", "subject_refs": [ "person_1" ], "object_refs": [ "window_pane_1" ], "confidence": "high", "temporal_status": "localized_sparse_support", "intervals": [ { "start_s": 1.0, "end_s": 7.0 } ], "evidence": "Across samples from 1.0 through 6.0 s (source 7.5–12.5 s), the selected person's head and torso remain oriented toward the large window.", "mask_relation": "subject matches the supplied selected-person mask" }, { "id": "action_3", "mentions": [ { "text": "gestures", "occurrence": 1, "ranges": [ [ 71, 79 ] ], "particle": { "text": null, "occurrence": 1 } }, { "text": "pointing", "occurrence": 1, "ranges": [ [ 100, 108 ] ], "particle": { "text": null, "occurrence": 1 } } ], "type": "simple", "subject_refs": [ "person_1" ], "object_refs": [ "window_pane_1" ], "confidence": "high", "temporal_status": "localized_sparse_support", "intervals": [ { "start_s": 1.0, "end_s": 7.0 } ], "evidence": "Samples from 1.0 through 6.0 s show the mask-linked person repeatedly raising and extending his left arm and hand upward and across the window.", "mask_relation": "subject matches the supplied selected-person mask" }, { "id": "action_4", "mentions": [ { "text": "turning", "occurrence": 1, "ranges": [ [ 147, 154 ] ], "particle": { "text": null, "occurrence": 1 } } ], "type": "simple", "subject_refs": [ "person_1" ], "object_refs": [], "confidence": "medium", "temporal_status": "localized_sparse_support", "intervals": [ { "start_s": 7.0, "end_s": 8.5 } ], "evidence": "Between the 7.0 and 8.0 s samples (source 13.5–14.5 s), the selected person's torso and head change away from the window toward his left; the final sampled pose supports continuation to the clip boundary.", "mask_relation": "subject matches the supplied selected-person mask" } ], "notes": "The supplied mask consistently identifies the captioned person. No separate source masks were supplied for the person's body, left hand, or the window pane, so those entity groups have null source-mask links. Directional references such as 'his right' and 'his left' are spatial concepts and are not entity groups. 'gestures' and 'pointing' are grouped as two mentions of the same elaborated hand-gesture event." }, "normalization_repairs": 0 }, { "source_video_id": "v_9A0BwCxAKAg", "caption_count_category": 14, "clip": { "caption_index": 3, "caption": "He picks up a yellow sponge squeegee and turns back to the window. Using his right hand, he wets the glass pane with the sponge, moving it up and down and then in a large circular motion.", "source_interval_sec": [ 15.0, 33.5 ], "loom_clip_id": "v_9A0BwCxAKAg_450_1004", "clip_frame_interval_4fps": [ 450, 1004 ], "sample_count_1fps": 19, "sampled_frames": [ { "frame": "00000", "timestamp_sec": 15.0, "offset_sec": 0, "rgb_path": "assets/cat_14/v_9A0BwCxAKAg_450_1004/0000s_00000_rgb.jpg", "mask_path": "assets/cat_14/v_9A0BwCxAKAg_450_1004/0000s_00000_mask.png", "mask_status": "available" }, { "frame": "00028", "timestamp_sec": 16.0, "offset_sec": 1, "rgb_path": "assets/cat_14/v_9A0BwCxAKAg_450_1004/0001s_00028_rgb.jpg", "mask_path": "assets/cat_14/v_9A0BwCxAKAg_450_1004/0001s_00028_mask.png", "mask_status": "available" }, { "frame": "00060", "timestamp_sec": 17.0, "offset_sec": 2, "rgb_path": "assets/cat_14/v_9A0BwCxAKAg_450_1004/0002s_00060_rgb.jpg", "mask_path": "assets/cat_14/v_9A0BwCxAKAg_450_1004/0002s_00060_mask.png", "mask_status": "available" }, { "frame": "00088", "timestamp_sec": 18.0, "offset_sec": 3, "rgb_path": "assets/cat_14/v_9A0BwCxAKAg_450_1004/0003s_00088_rgb.jpg", "mask_path": "assets/cat_14/v_9A0BwCxAKAg_450_1004/0003s_00088_mask.png", "mask_status": "available" }, { "frame": "00120", "timestamp_sec": 19.0, "offset_sec": 4, "rgb_path": "assets/cat_14/v_9A0BwCxAKAg_450_1004/0004s_00120_rgb.jpg", "mask_path": "assets/cat_14/v_9A0BwCxAKAg_450_1004/0004s_00120_mask.png", "mask_status": "available" }, { "frame": "00148", "timestamp_sec": 20.0, "offset_sec": 5, "rgb_path": "assets/cat_14/v_9A0BwCxAKAg_450_1004/0005s_00148_rgb.jpg", "mask_path": "assets/cat_14/v_9A0BwCxAKAg_450_1004/0005s_00148_mask.png", "mask_status": "available" }, { "frame": "00180", "timestamp_sec": 21.0, "offset_sec": 6, "rgb_path": "assets/cat_14/v_9A0BwCxAKAg_450_1004/0006s_00180_rgb.jpg", "mask_path": "assets/cat_14/v_9A0BwCxAKAg_450_1004/0006s_00180_mask.png", "mask_status": "available" }, { "frame": "00208", "timestamp_sec": 22.0, "offset_sec": 7, "rgb_path": "assets/cat_14/v_9A0BwCxAKAg_450_1004/0007s_00208_rgb.jpg", "mask_path": "assets/cat_14/v_9A0BwCxAKAg_450_1004/0007s_00208_mask.png", "mask_status": "available" }, { "frame": "00240", "timestamp_sec": 23.0, "offset_sec": 8, "rgb_path": "assets/cat_14/v_9A0BwCxAKAg_450_1004/0008s_00240_rgb.jpg", "mask_path": "assets/cat_14/v_9A0BwCxAKAg_450_1004/0008s_00240_mask.png", "mask_status": "available" }, { "frame": "00268", "timestamp_sec": 24.0, "offset_sec": 9, "rgb_path": "assets/cat_14/v_9A0BwCxAKAg_450_1004/0009s_00268_rgb.jpg", "mask_path": "assets/cat_14/v_9A0BwCxAKAg_450_1004/0009s_00268_mask.png", "mask_status": "available" }, { "frame": "00300", "timestamp_sec": 25.0, "offset_sec": 10, "rgb_path": "assets/cat_14/v_9A0BwCxAKAg_450_1004/0010s_00300_rgb.jpg", "mask_path": "assets/cat_14/v_9A0BwCxAKAg_450_1004/0010s_00300_mask.png", "mask_status": "available" }, { "frame": "00328", "timestamp_sec": 26.0, "offset_sec": 11, "rgb_path": "assets/cat_14/v_9A0BwCxAKAg_450_1004/0011s_00328_rgb.jpg", "mask_path": "assets/cat_14/v_9A0BwCxAKAg_450_1004/0011s_00328_mask.png", "mask_status": "available" }, { "frame": "00360", "timestamp_sec": 27.0, "offset_sec": 12, "rgb_path": "assets/cat_14/v_9A0BwCxAKAg_450_1004/0012s_00360_rgb.jpg", "mask_path": "assets/cat_14/v_9A0BwCxAKAg_450_1004/0012s_00360_mask.png", "mask_status": "available" }, { "frame": "00388", "timestamp_sec": 28.0, "offset_sec": 13, "rgb_path": "assets/cat_14/v_9A0BwCxAKAg_450_1004/0013s_00388_rgb.jpg", "mask_path": "assets/cat_14/v_9A0BwCxAKAg_450_1004/0013s_00388_mask.png", "mask_status": "available" }, { "frame": "00420", "timestamp_sec": 29.0, "offset_sec": 14, "rgb_path": "assets/cat_14/v_9A0BwCxAKAg_450_1004/0014s_00420_rgb.jpg", "mask_path": "assets/cat_14/v_9A0BwCxAKAg_450_1004/0014s_00420_mask.png", "mask_status": "available" }, { "frame": "00448", "timestamp_sec": 30.0, "offset_sec": 15, "rgb_path": "assets/cat_14/v_9A0BwCxAKAg_450_1004/0015s_00448_rgb.jpg", "mask_path": "assets/cat_14/v_9A0BwCxAKAg_450_1004/0015s_00448_mask.png", "mask_status": "available" }, { "frame": "00480", "timestamp_sec": 31.0, "offset_sec": 16, "rgb_path": "assets/cat_14/v_9A0BwCxAKAg_450_1004/0016s_00480_rgb.jpg", "mask_path": "assets/cat_14/v_9A0BwCxAKAg_450_1004/0016s_00480_mask.png", "mask_status": "available" }, { "frame": "00508", "timestamp_sec": 32.0, "offset_sec": 17, "rgb_path": "assets/cat_14/v_9A0BwCxAKAg_450_1004/0017s_00508_rgb.jpg", "mask_path": "assets/cat_14/v_9A0BwCxAKAg_450_1004/0017s_00508_mask.png", "mask_status": "available" }, { "frame": "00540", "timestamp_sec": 33.0, "offset_sec": 18, "rgb_path": "assets/cat_14/v_9A0BwCxAKAg_450_1004/0018s_00540_rgb.jpg", "mask_path": "assets/cat_14/v_9A0BwCxAKAg_450_1004/0018s_00540_mask.png", "mask_status": "available" } ], "obj_id": [ 0 ], "anno_id": [ 1271 ], "fps_inferred_from_clip_interval": 29.9459 }, "usage": { "input_tokens": 30058, "input_tokens_details": { "cache_write_tokens": 30055, "cached_tokens": 0 }, "output_tokens": 4454, "output_tokens_details": { "reasoning_tokens": 2588 }, "total_tokens": 34512 }, "model": "gpt-5.6-sol", "effort": "medium", "custom_id": "cat14_cap03_v_9A0BwCxAKAg_450_1004", "annotation": { "caption": "He picks up a yellow sponge squeegee and turns back to the window. Using his right hand, he wets the glass pane with the sponge, moving it up and down and then in a large circular motion.", "entity_groups": [ { "id": "person_1", "label": "person", "spans": [ { "text": "He", "range": [ 0, 2 ] }, { "text": "his", "range": [ 73, 76 ] }, { "text": "he", "range": [ 89, 91 ] } ], "source_mask_link": "selected_person" }, { "id": "sponge_squeegee_1", "label": "sponge squeegee", "spans": [ { "text": "a yellow sponge squeegee", "range": [ 12, 36 ] }, { "text": "the sponge", "range": [ 117, 127 ] }, { "text": "it", "range": [ 136, 138 ] } ], "source_mask_link": null }, { "id": "window_1", "label": "window", "spans": [ { "text": "the window", "range": [ 55, 65 ] } ], "source_mask_link": null }, { "id": "right_hand_1", "label": "right hand", "spans": [ { "text": "his right hand", "range": [ 73, 87 ] } ], "source_mask_link": null }, { "id": "glass_pane_1", "label": "glass pane", "spans": [ { "text": "the glass pane", "range": [ 97, 111 ] } ], "source_mask_link": null } ], "actions": [ { "id": "action_1", "mentions": [ { "text": "picks up", "ranges": [ [ 3, 11 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "type": "phrasal", "subject_refs": [ "person_1" ], "object_refs": [], "confidence": "medium", "temporal_status": "sparsely_supported", "intervals": [ { "start_s": 0.0, "end_s": 1.0, "source_start_s": 15.0, "source_end_s": 16.0 } ], "evidence": "At 0.000 s the tool is not visible in the person's raised hand area; by 1.000 s the masked person is holding the yellow squeegee against the window.", "mask_relation": "The action subject matches the supplied selected-person mask.", "source_subject_refs": "person_1" }, { "id": "action_2", "mentions": [ { "text": "turns", "ranges": [ [ 41, 46 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "type": "simple", "subject_refs": [ "person_1" ], "object_refs": [], "confidence": "medium", "temporal_status": "uncertain", "intervals": [], "evidence": "The sampled frames show the masked person oriented toward the window, but no sampled transition directly shows the turn.", "mask_relation": "The action subject matches the supplied selected-person mask.", "source_subject_refs": "person_1" }, { "id": "action_3", "mentions": [ { "text": "Using", "ranges": [ [ 67, 72 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "type": "simple", "subject_refs": [ "person_1" ], "object_refs": [], "confidence": "high", "temporal_status": "supported", "intervals": [ { "start_s": 1.0, "end_s": 17.0, "source_start_s": 16.0, "source_end_s": 32.0 }, { "start_s": 18.0, "end_s": 18.5, "source_start_s": 33.0, "source_end_s": 33.5 } ], "evidence": "The masked person repeatedly holds and guides the yellow squeegee with the right hand across the pane; the tool is not visible at the 17.000 s sample but reappears at 18.000 s.", "mask_relation": "The action subject matches the supplied selected-person mask.", "source_subject_refs": "person_1" }, { "id": "action_4", "mentions": [ { "text": "wets", "ranges": [ [ 92, 96 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "type": "simple", "subject_refs": [ "person_1" ], "object_refs": [], "confidence": "high", "temporal_status": "supported", "intervals": [ { "start_s": 1.0, "end_s": 17.0, "source_start_s": 16.0, "source_end_s": 32.0 }, { "start_s": 18.0, "end_s": 18.5, "source_start_s": 33.0, "source_end_s": 33.5 } ], "evidence": "The masked person applies the sponge squeegee over the glass, and wet streaks become visible across the pane.", "mask_relation": "The action subject matches the supplied selected-person mask.", "source_subject_refs": "person_1" }, { "id": "action_5", "mentions": [ { "text": "moving", "ranges": [ [ 129, 135 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "type": "simple", "subject_refs": [ "person_1" ], "object_refs": [], "confidence": "high", "temporal_status": "supported", "intervals": [ { "start_s": 1.0, "end_s": 17.0, "source_start_s": 16.0, "source_end_s": 32.0 }, { "start_s": 18.0, "end_s": 18.5, "source_start_s": 33.0, "source_end_s": 33.5 } ], "evidence": "Across consecutive samples, the masked person's squeegee changes between high, middle, low, and lateral positions; the final sample shows a broad curved wet trace.", "mask_relation": "The action subject matches the supplied selected-person mask.", "source_subject_refs": "person_1" } ], "notes": [ "The supplied masklet consistently identifies the single person referred to by “He,” “his,” and “he.”", "No separate source masks were supplied for the sponge squeegee, window, right hand, or glass pane.", "The window and its glass pane are kept as separate entity groups because the caption names the whole structure and the pane as a tangible part.", "The phrase “a large circular motion” names manner/trajectory rather than a separate concrete entity or light-verb action.", "Directional modifiers “back,” “up,” and “down” are excluded from the action spans." ] }, "model_output_raw": { "caption": "He picks up a yellow sponge squeegee and turns back to the window. Using his right hand, he wets the glass pane with the sponge, moving it up and down and then in a large circular motion.", "entity_groups": [ { "id": "person_1", "label": "person", "spans": [ { "text": "He", "range": [ 0, 2 ] }, { "text": "his", "range": [ 73, 76 ] }, { "text": "he", "range": [ 89, 91 ] } ], "source_mask": "selected_person" }, { "id": "sponge_squeegee_1", "label": "sponge squeegee", "spans": [ { "text": "a yellow sponge squeegee", "range": [ 12, 36 ] }, { "text": "the sponge", "range": [ 117, 127 ] }, { "text": "it", "range": [ 136, 138 ] } ], "source_mask": null }, { "id": "window_1", "label": "window", "spans": [ { "text": "the window", "range": [ 55, 65 ] } ], "source_mask": null }, { "id": "right_hand_1", "label": "right hand", "spans": [ { "text": "his right hand", "range": [ 73, 87 ] } ], "source_mask": null }, { "id": "glass_pane_1", "label": "glass pane", "spans": [ { "text": "the glass pane", "range": [ 97, 111 ] } ], "source_mask": null } ], "actions": [ { "id": "action_1", "mentions": [ { "text": "picks up", "ranges": [ [ 3, 11 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "type": "phrasal", "subject_ref": "person_1", "object_refs": [ { "entity_id": "sponge_squeegee_1", "role": "patient" } ], "confidence": "medium", "temporal_status": "sparsely_supported", "intervals": [ { "clip_relative": [ 0.0, 1.0 ], "source_video": [ 15.0, 16.0 ] } ], "evidence": "At 0.000 s the tool is not visible in the person's raised hand area; by 1.000 s the masked person is holding the yellow squeegee against the window.", "mask_relation": "The action subject matches the supplied selected-person mask." }, { "id": "action_2", "mentions": [ { "text": "turns", "ranges": [ [ 41, 46 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "type": "simple", "subject_ref": "person_1", "object_refs": [ { "entity_id": "window_1", "role": "goal" } ], "confidence": "medium", "temporal_status": "uncertain", "intervals": [], "evidence": "The sampled frames show the masked person oriented toward the window, but no sampled transition directly shows the turn.", "mask_relation": "The action subject matches the supplied selected-person mask." }, { "id": "action_3", "mentions": [ { "text": "Using", "ranges": [ [ 67, 72 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "type": "simple", "subject_ref": "person_1", "object_refs": [ { "entity_id": "right_hand_1", "role": "instrument" } ], "confidence": "high", "temporal_status": "supported", "intervals": [ { "clip_relative": [ 1.0, 17.0 ], "source_video": [ 16.0, 32.0 ] }, { "clip_relative": [ 18.0, 18.5 ], "source_video": [ 33.0, 33.5 ] } ], "evidence": "The masked person repeatedly holds and guides the yellow squeegee with the right hand across the pane; the tool is not visible at the 17.000 s sample but reappears at 18.000 s.", "mask_relation": "The action subject matches the supplied selected-person mask." }, { "id": "action_4", "mentions": [ { "text": "wets", "ranges": [ [ 92, 96 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "type": "simple", "subject_ref": "person_1", "object_refs": [ { "entity_id": "glass_pane_1", "role": "patient" }, { "entity_id": "sponge_squeegee_1", "role": "instrument" } ], "confidence": "high", "temporal_status": "supported", "intervals": [ { "clip_relative": [ 1.0, 17.0 ], "source_video": [ 16.0, 32.0 ] }, { "clip_relative": [ 18.0, 18.5 ], "source_video": [ 33.0, 33.5 ] } ], "evidence": "The masked person applies the sponge squeegee over the glass, and wet streaks become visible across the pane.", "mask_relation": "The action subject matches the supplied selected-person mask." }, { "id": "action_5", "mentions": [ { "text": "moving", "ranges": [ [ 129, 135 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "type": "simple", "subject_ref": "person_1", "object_refs": [ { "entity_id": "sponge_squeegee_1", "role": "patient" } ], "confidence": "high", "temporal_status": "supported", "intervals": [ { "clip_relative": [ 1.0, 17.0 ], "source_video": [ 16.0, 32.0 ] }, { "clip_relative": [ 18.0, 18.5 ], "source_video": [ 33.0, 33.5 ] } ], "evidence": "Across consecutive samples, the masked person's squeegee changes between high, middle, low, and lateral positions; the final sample shows a broad curved wet trace.", "mask_relation": "The action subject matches the supplied selected-person mask." } ], "notes": [ "The supplied masklet consistently identifies the single person referred to by “He,” “his,” and “he.”", "No separate source masks were supplied for the sponge squeegee, window, right hand, or glass pane.", "The window and its glass pane are kept as separate entity groups because the caption names the whole structure and the pane as a tangible part.", "The phrase “a large circular motion” names manner/trajectory rather than a separate concrete entity or light-verb action.", "Directional modifiers “back,” “up,” and “down” are excluded from the action spans." ] }, "normalization_repairs": 0 }, { "source_video_id": "v_9A0BwCxAKAg", "caption_count_category": 14, "clip": { "caption_index": 4, "caption": "The person sets down the sponge squeegee and bends down, disappearing from the lower part of the frame. He straightens back up holding a dark cloth in his hand and looks at the window.", "source_interval_sec": [ 33.5, 47.5 ], "loom_clip_id": "v_9A0BwCxAKAg_1004_1424", "clip_frame_interval_4fps": [ 1004, 1424 ], "sample_count_1fps": 14, "sampled_frames": [ { "frame": "00000", "timestamp_sec": 33.5, "offset_sec": 0, "rgb_path": "assets/cat_14/v_9A0BwCxAKAg_1004_1424/0000s_00000_rgb.jpg", "mask_path": "assets/cat_14/v_9A0BwCxAKAg_1004_1424/0000s_00000_mask.png", "mask_status": "available" }, { "frame": "00028", "timestamp_sec": 34.5, "offset_sec": 1, "rgb_path": "assets/cat_14/v_9A0BwCxAKAg_1004_1424/0001s_00028_rgb.jpg", "mask_path": "assets/cat_14/v_9A0BwCxAKAg_1004_1424/0001s_00028_mask.png", "mask_status": "available" }, { "frame": "00060", "timestamp_sec": 35.5, "offset_sec": 2, "rgb_path": "assets/cat_14/v_9A0BwCxAKAg_1004_1424/0002s_00060_rgb.jpg", "mask_path": "assets/cat_14/v_9A0BwCxAKAg_1004_1424/0002s_00060_mask.png", "mask_status": "available" }, { "frame": "00088", "timestamp_sec": 36.5, "offset_sec": 3, "rgb_path": "assets/cat_14/v_9A0BwCxAKAg_1004_1424/0003s_00088_rgb.jpg", "mask_path": "assets/cat_14/v_9A0BwCxAKAg_1004_1424/0003s_00088_mask.png", "mask_status": "available" }, { "frame": "00120", "timestamp_sec": 37.5, "offset_sec": 4, "rgb_path": "assets/cat_14/v_9A0BwCxAKAg_1004_1424/0004s_00120_rgb.jpg", "mask_path": "assets/cat_14/v_9A0BwCxAKAg_1004_1424/0004s_00120_mask.png", "mask_status": "available" }, { "frame": "00148", "timestamp_sec": 38.5, "offset_sec": 5, "rgb_path": "assets/cat_14/v_9A0BwCxAKAg_1004_1424/0005s_00148_rgb.jpg", "mask_path": "assets/cat_14/v_9A0BwCxAKAg_1004_1424/0005s_00148_mask.png", "mask_status": "available" }, { "frame": "00180", "timestamp_sec": 39.5, "offset_sec": 6, "rgb_path": "assets/cat_14/v_9A0BwCxAKAg_1004_1424/0006s_00180_rgb.jpg", "mask_path": "assets/cat_14/v_9A0BwCxAKAg_1004_1424/0006s_00180_mask.png", "mask_status": "available" }, { "frame": "00208", "timestamp_sec": 40.5, "offset_sec": 7, "rgb_path": "assets/cat_14/v_9A0BwCxAKAg_1004_1424/0007s_00208_rgb.jpg", "mask_path": "assets/cat_14/v_9A0BwCxAKAg_1004_1424/0007s_00208_mask.png", "mask_status": "available" }, { "frame": "00240", "timestamp_sec": 41.5, "offset_sec": 8, "rgb_path": "assets/cat_14/v_9A0BwCxAKAg_1004_1424/0008s_00240_rgb.jpg", "mask_path": "assets/cat_14/v_9A0BwCxAKAg_1004_1424/0008s_00240_mask.png", "mask_status": "available" }, { "frame": "00268", "timestamp_sec": 42.5, "offset_sec": 9, "rgb_path": "assets/cat_14/v_9A0BwCxAKAg_1004_1424/0009s_00268_rgb.jpg", "mask_path": "assets/cat_14/v_9A0BwCxAKAg_1004_1424/0009s_00268_mask.png", "mask_status": "available" }, { "frame": "00300", "timestamp_sec": 43.5, "offset_sec": 10, "rgb_path": "assets/cat_14/v_9A0BwCxAKAg_1004_1424/0010s_00300_rgb.jpg", "mask_path": "assets/cat_14/v_9A0BwCxAKAg_1004_1424/0010s_00300_mask.png", "mask_status": "available" }, { "frame": "00328", "timestamp_sec": 44.5, "offset_sec": 11, "rgb_path": "assets/cat_14/v_9A0BwCxAKAg_1004_1424/0011s_00328_rgb.jpg", "mask_path": "assets/cat_14/v_9A0BwCxAKAg_1004_1424/0011s_00328_mask.png", "mask_status": "available" }, { "frame": "00360", "timestamp_sec": 45.5, "offset_sec": 12, "rgb_path": "assets/cat_14/v_9A0BwCxAKAg_1004_1424/0012s_00360_rgb.jpg", "mask_path": "assets/cat_14/v_9A0BwCxAKAg_1004_1424/0012s_00360_mask.png", "mask_status": "available" }, { "frame": "00388", "timestamp_sec": 46.5, "offset_sec": 13, "rgb_path": "assets/cat_14/v_9A0BwCxAKAg_1004_1424/0013s_00388_rgb.jpg", "mask_path": "assets/cat_14/v_9A0BwCxAKAg_1004_1424/0013s_00388_mask.png", "mask_status": "available" } ], "obj_id": [ 0 ], "anno_id": [ 1264 ], "fps_inferred_from_clip_interval": 30.0 }, "usage": { "input_tokens": 26180, "input_tokens_details": { "cache_write_tokens": 26177, "cached_tokens": 0 }, "output_tokens": 5214, "output_tokens_details": { "reasoning_tokens": 3439 }, "total_tokens": 31394 }, "model": "gpt-5.6-sol", "effort": "medium", "custom_id": "cat14_cap04_v_9A0BwCxAKAg_1004_1424", "annotation": { "caption": "The person sets down the sponge squeegee and bends down, disappearing from the lower part of the frame. He straightens back up holding a dark cloth in his hand and looks at the window.", "entity_groups": [ { "id": "person_1", "label": "person", "spans": [ { "text": "The person", "range": [ 0, 10 ] }, { "text": "He", "range": [ 104, 106 ] }, { "text": "his", "range": [ 151, 154 ] } ], "source_mask_link": "selected_person" }, { "id": "squeegee_1", "label": "sponge squeegee", "spans": [ { "text": "the sponge squeegee", "range": [ 21, 40 ] } ], "source_mask_link": null }, { "id": "cloth_1", "label": "cloth", "spans": [ { "text": "a dark cloth", "range": [ 135, 147 ] } ], "source_mask_link": null }, { "id": "hand_1", "label": "hand", "spans": [ { "text": "his hand", "range": [ 151, 159 ] } ], "source_mask_link": null }, { "id": "window_1", "label": "window", "spans": [ { "text": "the window", "range": [ 173, 183 ] } ], "source_mask_link": null } ], "actions": [ { "id": "action_1", "mentions": [ { "text": "sets down", "ranges": [ [ 11, 20 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "type": "phrasal", "subject_refs": [ "person_1" ], "object_refs": [ "squeegee_1" ], "confidence": "high", "temporal_status": "uncertain", "intervals": [], "evidence": "At 0.000 s the masked person still has a raised cleaning tool; by 1.000 s the arm and tool are lowered or out of view, but the actual placement is not directly visible in a sampled frame.", "mask_relation": "subject_matches_selected_person_mask", "source_subject_refs": [ "person_1" ] }, { "id": "action_2", "mentions": [ { "text": "bends", "ranges": [ [ 45, 50 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "type": "simple", "subject_refs": [ "person_1" ], "object_refs": [], "confidence": "high", "temporal_status": "localized", "intervals": [ { "start_s": 7.0, "end_s": 11.0, "source_start_s": 40.5, "source_end_s": 44.5 } ], "evidence": "Samples from 7.000 through 10.000 s show the masked person's torso and head progressively lowered into a deeply bent posture.", "mask_relation": "subject_matches_selected_person_mask", "source_subject_refs": [ "person_1" ] }, { "id": "action_3", "mentions": [ { "text": "disappearing", "ranges": [ [ 57, 69 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "type": "simple", "subject_refs": [ "person_1" ], "object_refs": [], "confidence": "medium", "temporal_status": "localized", "intervals": [ { "start_s": 8.0, "end_s": 11.0, "source_start_s": 41.5, "source_end_s": 44.5 } ], "evidence": "At 8.000–10.000 s the masked person bends into the lower edge of the image, leaving progressively less of the body clearly visible.", "mask_relation": "subject_matches_selected_person_mask", "source_subject_refs": [ "person_1" ] }, { "id": "action_4", "mentions": [ { "text": "straightens up", "ranges": [ [ 107, 118 ], [ 124, 126 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "type": "phrasal", "subject_refs": [ "person_1" ], "object_refs": [], "confidence": "high", "temporal_status": "localized", "intervals": [ { "start_s": 10.0, "end_s": 12.0, "source_start_s": 43.5, "source_end_s": 45.5 } ], "evidence": "The masked person is deeply bent at 10.000 s and is substantially upright by 11.000 s.", "mask_relation": "subject_matches_selected_person_mask", "source_subject_refs": [ "person_1" ] }, { "id": "action_5", "mentions": [ { "text": "holding", "ranges": [ [ 127, 134 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "type": "simple", "subject_refs": [ "person_1" ], "object_refs": [ "cloth_1" ], "confidence": "high", "temporal_status": "uncertain", "intervals": [], "evidence": "The person's hand is near the lower-left area in the final samples, but the dark cloth cannot be confidently distinguished from the dark clothing and background.", "mask_relation": "subject_matches_selected_person_mask", "source_subject_refs": [ "person_1" ] }, { "id": "action_6", "mentions": [ { "text": "looks", "ranges": [ [ 164, 169 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "type": "prepositional", "subject_refs": [ "person_1" ], "object_refs": [ "window_1" ], "confidence": "high", "temporal_status": "localized", "intervals": [ { "start_s": 11.0, "end_s": 14.0, "source_start_s": 44.5, "source_end_s": 47.5 } ], "evidence": "From 11.000 through the final 13.000 s sample, the masked person's head is oriented toward the window; support reaches the clip boundary at 14.000 s.", "mask_relation": "subject_matches_selected_person_mask", "source_subject_refs": [ "person_1" ] } ], "notes": [ "The supplied mask consistently identifies the single person, so all annotated actions have a mask-linked subject.", "No separate masks were supplied for the sponge squeegee, cloth, hand, or window.", "\"the lower part\" is an abstract spatial reference and \"the frame\" is a meta-reference to the image medium, so neither is an entity group.", "Directional \"down\" after \"bends\" and the adverb \"back\" in \"straightens back up\" are excluded from their action spans." ] }, "model_output_raw": { "caption": "The person sets down the sponge squeegee and bends down, disappearing from the lower part of the frame. He straightens back up holding a dark cloth in his hand and looks at the window.", "entity_groups": [ { "id": "person_1", "entity_type": "person", "spans": [ { "text": "The person", "range": [ 0, 10 ] }, { "text": "He", "range": [ 104, 106 ] }, { "text": "his", "range": [ 151, 154 ] } ], "source_mask_link": "selected_person" }, { "id": "squeegee_1", "entity_type": "sponge squeegee", "spans": [ { "text": "the sponge squeegee", "range": [ 21, 40 ] } ], "source_mask_link": null }, { "id": "cloth_1", "entity_type": "cloth", "spans": [ { "text": "a dark cloth", "range": [ 135, 147 ] } ], "source_mask_link": null }, { "id": "hand_1", "entity_type": "hand", "spans": [ { "text": "his hand", "range": [ 151, 159 ] } ], "source_mask_link": null }, { "id": "window_1", "entity_type": "window", "spans": [ { "text": "the window", "range": [ 173, 183 ] } ], "source_mask_link": null } ], "actions": [ { "id": "action_1", "mentions": [ { "text": "sets down", "ranges": [ [ 11, 20 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "type": "phrasal", "subject_refs": [ "person_1" ], "object_refs": [ "squeegee_1" ], "confidence": "high", "mask_relation": "subject_matches_selected_person_mask", "temporal_status": "uncertain", "intervals": [], "evidence": "At 0.000 s the masked person still has a raised cleaning tool; by 1.000 s the arm and tool are lowered or out of view, but the actual placement is not directly visible in a sampled frame." }, { "id": "action_2", "mentions": [ { "text": "bends", "ranges": [ [ 45, 50 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "type": "simple", "subject_refs": [ "person_1" ], "object_refs": [], "confidence": "high", "mask_relation": "subject_matches_selected_person_mask", "temporal_status": "localized", "intervals": [ { "start_s": 7.0, "end_s": 11.0, "source_start_s": 40.5, "source_end_s": 44.5 } ], "evidence": "Samples from 7.000 through 10.000 s show the masked person's torso and head progressively lowered into a deeply bent posture." }, { "id": "action_3", "mentions": [ { "text": "disappearing", "ranges": [ [ 57, 69 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "type": "simple", "subject_refs": [ "person_1" ], "object_refs": [], "confidence": "medium", "mask_relation": "subject_matches_selected_person_mask", "temporal_status": "localized", "intervals": [ { "start_s": 8.0, "end_s": 11.0, "source_start_s": 41.5, "source_end_s": 44.5 } ], "evidence": "At 8.000–10.000 s the masked person bends into the lower edge of the image, leaving progressively less of the body clearly visible." }, { "id": "action_4", "mentions": [ { "text": "straightens up", "ranges": [ [ 107, 118 ], [ 124, 126 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "type": "phrasal", "subject_refs": [ "person_1" ], "object_refs": [], "confidence": "high", "mask_relation": "subject_matches_selected_person_mask", "temporal_status": "localized", "intervals": [ { "start_s": 10.0, "end_s": 12.0, "source_start_s": 43.5, "source_end_s": 45.5 } ], "evidence": "The masked person is deeply bent at 10.000 s and is substantially upright by 11.000 s." }, { "id": "action_5", "mentions": [ { "text": "holding", "ranges": [ [ 127, 134 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "type": "simple", "subject_refs": [ "person_1" ], "object_refs": [ "cloth_1" ], "confidence": "high", "mask_relation": "subject_matches_selected_person_mask", "temporal_status": "uncertain", "intervals": [], "evidence": "The person's hand is near the lower-left area in the final samples, but the dark cloth cannot be confidently distinguished from the dark clothing and background." }, { "id": "action_6", "mentions": [ { "text": "looks", "ranges": [ [ 164, 169 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "type": "prepositional", "subject_refs": [ "person_1" ], "object_refs": [ "window_1" ], "confidence": "high", "mask_relation": "subject_matches_selected_person_mask", "temporal_status": "localized", "intervals": [ { "start_s": 11.0, "end_s": 14.0, "source_start_s": 44.5, "source_end_s": 47.5 } ], "evidence": "From 11.000 through the final 13.000 s sample, the masked person's head is oriented toward the window; support reaches the clip boundary at 14.000 s." } ], "notes": [ "The supplied mask consistently identifies the single person, so all annotated actions have a mask-linked subject.", "No separate masks were supplied for the sponge squeegee, cloth, hand, or window.", "\"the lower part\" is an abstract spatial reference and \"the frame\" is a meta-reference to the image medium, so neither is an entity group.", "Directional \"down\" after \"bends\" and the adverb \"back\" in \"straightens back up\" are excluded from their action spans." ] }, "normalization_repairs": 0 }, { "source_video_id": "v_9A0BwCxAKAg", "caption_count_category": 14, "clip": { "caption_index": 5, "caption": "He uses the cloth to wipe the top and left wooden frame of the window. He then puts the cloth down, picks up a metal squeegee, and begins wiping the top edge of the glass.", "source_interval_sec": [ 47.5, 69.0 ], "loom_clip_id": "v_9A0BwCxAKAg_1424_2068", "clip_frame_interval_4fps": [ 1424, 2068 ], "sample_count_1fps": 22, "sampled_frames": [ { "frame": "00000", "timestamp_sec": 47.5, "offset_sec": 0, "rgb_path": "assets/cat_14/v_9A0BwCxAKAg_1424_2068/0000s_00000_rgb.jpg", "mask_path": "assets/cat_14/v_9A0BwCxAKAg_1424_2068/0000s_00000_mask.png", "mask_status": "available" }, { "frame": "00028", "timestamp_sec": 48.5, "offset_sec": 1, "rgb_path": "assets/cat_14/v_9A0BwCxAKAg_1424_2068/0001s_00028_rgb.jpg", "mask_path": "assets/cat_14/v_9A0BwCxAKAg_1424_2068/0001s_00028_mask.png", "mask_status": "available" }, { "frame": "00060", "timestamp_sec": 49.5, "offset_sec": 2, "rgb_path": "assets/cat_14/v_9A0BwCxAKAg_1424_2068/0002s_00060_rgb.jpg", "mask_path": "assets/cat_14/v_9A0BwCxAKAg_1424_2068/0002s_00060_mask.png", "mask_status": "available" }, { "frame": "00088", "timestamp_sec": 50.5, "offset_sec": 3, "rgb_path": "assets/cat_14/v_9A0BwCxAKAg_1424_2068/0003s_00088_rgb.jpg", "mask_path": "assets/cat_14/v_9A0BwCxAKAg_1424_2068/0003s_00088_mask.png", "mask_status": "available" }, { "frame": "00120", "timestamp_sec": 51.5, "offset_sec": 4, "rgb_path": "assets/cat_14/v_9A0BwCxAKAg_1424_2068/0004s_00120_rgb.jpg", "mask_path": "assets/cat_14/v_9A0BwCxAKAg_1424_2068/0004s_00120_mask.png", "mask_status": "available" }, { "frame": "00148", "timestamp_sec": 52.5, "offset_sec": 5, "rgb_path": "assets/cat_14/v_9A0BwCxAKAg_1424_2068/0005s_00148_rgb.jpg", "mask_path": "assets/cat_14/v_9A0BwCxAKAg_1424_2068/0005s_00148_mask.png", "mask_status": "available" }, { "frame": "00180", "timestamp_sec": 53.5, "offset_sec": 6, "rgb_path": "assets/cat_14/v_9A0BwCxAKAg_1424_2068/0006s_00180_rgb.jpg", "mask_path": "assets/cat_14/v_9A0BwCxAKAg_1424_2068/0006s_00180_mask.png", "mask_status": "available" }, { "frame": "00208", "timestamp_sec": 54.5, "offset_sec": 7, "rgb_path": "assets/cat_14/v_9A0BwCxAKAg_1424_2068/0007s_00208_rgb.jpg", "mask_path": "assets/cat_14/v_9A0BwCxAKAg_1424_2068/0007s_00208_mask.png", "mask_status": "available" }, { "frame": "00240", "timestamp_sec": 55.5, "offset_sec": 8, "rgb_path": "assets/cat_14/v_9A0BwCxAKAg_1424_2068/0008s_00240_rgb.jpg", "mask_path": "assets/cat_14/v_9A0BwCxAKAg_1424_2068/0008s_00240_mask.png", "mask_status": "available" }, { "frame": "00268", "timestamp_sec": 56.5, "offset_sec": 9, "rgb_path": "assets/cat_14/v_9A0BwCxAKAg_1424_2068/0009s_00268_rgb.jpg", "mask_path": "assets/cat_14/v_9A0BwCxAKAg_1424_2068/0009s_00268_mask.png", "mask_status": "available" }, { "frame": "00300", "timestamp_sec": 57.5, "offset_sec": 10, "rgb_path": "assets/cat_14/v_9A0BwCxAKAg_1424_2068/0010s_00300_rgb.jpg", "mask_path": "assets/cat_14/v_9A0BwCxAKAg_1424_2068/0010s_00300_mask.png", "mask_status": "available" }, { "frame": "00328", "timestamp_sec": 58.5, "offset_sec": 11, "rgb_path": "assets/cat_14/v_9A0BwCxAKAg_1424_2068/0011s_00328_rgb.jpg", "mask_path": "assets/cat_14/v_9A0BwCxAKAg_1424_2068/0011s_00328_mask.png", "mask_status": "available" }, { "frame": "00360", "timestamp_sec": 59.5, "offset_sec": 12, "rgb_path": "assets/cat_14/v_9A0BwCxAKAg_1424_2068/0012s_00360_rgb.jpg", "mask_path": "assets/cat_14/v_9A0BwCxAKAg_1424_2068/0012s_00360_mask.png", "mask_status": "available" }, { "frame": "00388", "timestamp_sec": 60.5, "offset_sec": 13, "rgb_path": "assets/cat_14/v_9A0BwCxAKAg_1424_2068/0013s_00388_rgb.jpg", "mask_path": "assets/cat_14/v_9A0BwCxAKAg_1424_2068/0013s_00388_mask.png", "mask_status": "available" }, { "frame": "00420", "timestamp_sec": 61.5, "offset_sec": 14, "rgb_path": "assets/cat_14/v_9A0BwCxAKAg_1424_2068/0014s_00420_rgb.jpg", "mask_path": "assets/cat_14/v_9A0BwCxAKAg_1424_2068/0014s_00420_mask.png", "mask_status": "available" }, { "frame": "00448", "timestamp_sec": 62.5, "offset_sec": 15, "rgb_path": "assets/cat_14/v_9A0BwCxAKAg_1424_2068/0015s_00448_rgb.jpg", "mask_path": "assets/cat_14/v_9A0BwCxAKAg_1424_2068/0015s_00448_mask.png", "mask_status": "available" }, { "frame": "00480", "timestamp_sec": 63.5, "offset_sec": 16, "rgb_path": "assets/cat_14/v_9A0BwCxAKAg_1424_2068/0016s_00480_rgb.jpg", "mask_path": "assets/cat_14/v_9A0BwCxAKAg_1424_2068/0016s_00480_mask.png", "mask_status": "available" }, { "frame": "00508", "timestamp_sec": 64.5, "offset_sec": 17, "rgb_path": "assets/cat_14/v_9A0BwCxAKAg_1424_2068/0017s_00508_rgb.jpg", "mask_path": "assets/cat_14/v_9A0BwCxAKAg_1424_2068/0017s_00508_mask.png", "mask_status": "available" }, { "frame": "00540", "timestamp_sec": 65.5, "offset_sec": 18, "rgb_path": "assets/cat_14/v_9A0BwCxAKAg_1424_2068/0018s_00540_rgb.jpg", "mask_path": "assets/cat_14/v_9A0BwCxAKAg_1424_2068/0018s_00540_mask.png", "mask_status": "available" }, { "frame": "00568", "timestamp_sec": 66.5, "offset_sec": 19, "rgb_path": "assets/cat_14/v_9A0BwCxAKAg_1424_2068/0019s_00568_rgb.jpg", "mask_path": "assets/cat_14/v_9A0BwCxAKAg_1424_2068/0019s_00568_mask.png", "mask_status": "available" }, { "frame": "00600", "timestamp_sec": 67.5, "offset_sec": 20, "rgb_path": "assets/cat_14/v_9A0BwCxAKAg_1424_2068/0020s_00600_rgb.jpg", "mask_path": "assets/cat_14/v_9A0BwCxAKAg_1424_2068/0020s_00600_mask.png", "mask_status": "available" }, { "frame": "00628", "timestamp_sec": 68.5, "offset_sec": 21, "rgb_path": "assets/cat_14/v_9A0BwCxAKAg_1424_2068/0021s_00628_rgb.jpg", "mask_path": "assets/cat_14/v_9A0BwCxAKAg_1424_2068/0021s_00628_mask.png", "mask_status": "available" } ], "obj_id": [ 0 ], "anno_id": [ 1265 ], "fps_inferred_from_clip_interval": 29.9535 }, "usage": { "input_tokens": 32381, "input_tokens_details": { "cache_write_tokens": 32378, "cached_tokens": 0 }, "output_tokens": 4762, "output_tokens_details": { "reasoning_tokens": 3099 }, "total_tokens": 37143 }, "model": "gpt-5.6-sol", "effort": "medium", "custom_id": "cat14_cap05_v_9A0BwCxAKAg_1424_2068", "annotation": { "caption": "He uses the cloth to wipe the top and left wooden frame of the window. He then puts the cloth down, picks up a metal squeegee, and begins wiping the top edge of the glass.", "entity_groups": [ { "id": "e1", "label": "person", "spans": [ { "text": "He", "range": [ 0, 2 ] }, { "text": "He", "range": [ 71, 73 ] } ], "source_mask_link": "selected_person" }, { "id": "e2", "label": "cloth", "spans": [ { "text": "the cloth", "range": [ 8, 17 ] }, { "text": "the cloth", "range": [ 84, 93 ] } ], "source_mask_link": null }, { "id": "e3", "label": "window frame", "spans": [ { "text": "the top and left wooden frame", "range": [ 26, 55 ] } ], "source_mask_link": null }, { "id": "e4", "label": "window", "spans": [ { "text": "the window", "range": [ 59, 69 ] } ], "source_mask_link": null }, { "id": "e5", "label": "squeegee", "spans": [ { "text": "a metal squeegee", "range": [ 109, 125 ] } ], "source_mask_link": null }, { "id": "e6", "label": "glass", "spans": [ { "text": "the glass", "range": [ 161, 170 ] } ], "source_mask_link": null } ], "actions": [ { "id": "a1", "mentions": [ { "text": "uses", "ranges": [ [ 3, 7 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "type": "simple", "subject_refs": [ "e1" ], "object_refs": [ "e2" ], "confidence": "medium", "temporal_status": "localized", "intervals": [ { "start_s": 1.0, "end_s": 5.0, "source_start_s": 48.5, "source_end_s": 52.5 } ], "evidence": "At clip-relative 1–4 seconds, the masked person holds a dark cloth at the upper window frame and moves it along the frame.", "mask_relation": "subject matches the supplied selected-person mask", "source_subject_refs": "e1" }, { "id": "a2", "mentions": [ { "text": "wipe", "ranges": [ [ 21, 25 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "type": "simple", "subject_refs": [ "e1" ], "object_refs": [ "e3" ], "confidence": "high", "temporal_status": "localized", "intervals": [ { "start_s": 1.0, "end_s": 5.0, "source_start_s": 48.5, "source_end_s": 52.5 } ], "evidence": "The masked person repeatedly reaches across the top and upper side of the wooden window frame with the cloth at 1–4 seconds.", "mask_relation": "subject matches the supplied selected-person mask", "source_subject_refs": "e1" }, { "id": "a3", "mentions": [ { "text": "puts", "ranges": [ [ 79, 83 ] ], "occurrence": 1, "particle": { "text": "down", "occurrence": 1, "ranges": [ [ 94, 98 ] ] } } ], "type": "phrasal", "subject_refs": [ "e1" ], "object_refs": [ "e2" ], "confidence": "medium", "temporal_status": "localized", "intervals": [ { "start_s": 5.0, "end_s": 9.0, "source_start_s": 52.5, "source_end_s": 56.5 } ], "evidence": "From 5–8 seconds, the masked person lowers both hands away from the frame and bends toward the lower work area; the cloth itself is poorly resolved.", "mask_relation": "subject matches the supplied selected-person mask", "source_subject_refs": "e1" }, { "id": "a4", "mentions": [ { "text": "picks up", "ranges": [ [ 100, 108 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "type": "phrasal", "subject_refs": [ "e1" ], "object_refs": [ "e5" ], "confidence": "medium", "temporal_status": "localized", "intervals": [ { "start_s": 8.0, "end_s": 12.0, "source_start_s": 55.5, "source_end_s": 59.5 } ], "evidence": "At 8–11 seconds, the masked person bends and manipulates something below the window; by 12 seconds the squeegee has been raised into view.", "mask_relation": "subject matches the supplied selected-person mask", "source_subject_refs": "e1" }, { "id": "a5", "mentions": [ { "text": "wiping", "ranges": [ [ 138, 144 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "type": "simple", "subject_refs": [ "e1" ], "object_refs": [ "e6" ], "confidence": "high", "temporal_status": "localized", "intervals": [ { "start_s": 12.0, "end_s": 21.5, "source_start_s": 59.5, "source_end_s": 69.0 } ], "evidence": "From 12 seconds through the clip boundary, the masked person moves the metal squeegee along the upper glass and then across the glass surface.", "mask_relation": "subject matches the supplied selected-person mask", "source_subject_refs": "e1" } ], "notes": [ "The supplied masklet consistently identifies the sole visible person and links both occurrences of “He” to entity group e1.", "No separate source masks were supplied for the cloth, window frame, window, squeegee, or glass.", "The spatial part phrase “the top edge” is not a separate entity group; “the glass” is the grounded object of the final wiping action.", "The aspectual verb “begins” is excluded from the final action span; only “wiping” is selected." ] }, "model_output_raw": { "caption": "He uses the cloth to wipe the top and left wooden frame of the window. He then puts the cloth down, picks up a metal squeegee, and begins wiping the top edge of the glass.", "entity_groups": [ { "id": "e1", "entity": "person", "spans": [ { "text": "He", "range": [ 0, 2 ] }, { "text": "He", "range": [ 71, 73 ] } ], "source_mask_link": "selected_person_masklet" }, { "id": "e2", "entity": "cloth", "spans": [ { "text": "the cloth", "range": [ 8, 17 ] }, { "text": "the cloth", "range": [ 84, 93 ] } ], "source_mask_link": null }, { "id": "e3", "entity": "window frame", "spans": [ { "text": "the top and left wooden frame", "range": [ 26, 55 ] } ], "source_mask_link": null }, { "id": "e4", "entity": "window", "spans": [ { "text": "the window", "range": [ 59, 69 ] } ], "source_mask_link": null }, { "id": "e5", "entity": "squeegee", "spans": [ { "text": "a metal squeegee", "range": [ 109, 125 ] } ], "source_mask_link": null }, { "id": "e6", "entity": "glass", "spans": [ { "text": "the glass", "range": [ 161, 170 ] } ], "source_mask_link": null } ], "actions": [ { "id": "a1", "mentions": [ { "text": "uses", "ranges": [ [ 3, 7 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "type": "simple", "subject_entity_group_id": "e1", "object_entity_group_ids": [ "e2" ], "confidence": "medium", "temporal_status": "localized", "intervals": [ { "clip_relative_s": [ 1.0, 5.0 ], "source_video_s": [ 48.5, 52.5 ] } ], "mask_relation": "subject matches the supplied selected-person mask", "evidence": "At clip-relative 1–4 seconds, the masked person holds a dark cloth at the upper window frame and moves it along the frame." }, { "id": "a2", "mentions": [ { "text": "wipe", "ranges": [ [ 21, 25 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "type": "simple", "subject_entity_group_id": "e1", "object_entity_group_ids": [ "e3" ], "instrument_entity_group_id": "e2", "confidence": "high", "temporal_status": "localized", "intervals": [ { "clip_relative_s": [ 1.0, 5.0 ], "source_video_s": [ 48.5, 52.5 ] } ], "mask_relation": "subject matches the supplied selected-person mask", "evidence": "The masked person repeatedly reaches across the top and upper side of the wooden window frame with the cloth at 1–4 seconds." }, { "id": "a3", "mentions": [ { "text": "puts", "ranges": [ [ 79, 83 ] ], "occurrence": 1, "particle": { "text": "down", "occurrence": 1, "ranges": [ [ 94, 98 ] ] } } ], "type": "phrasal", "subject_entity_group_id": "e1", "object_entity_group_ids": [ "e2" ], "confidence": "medium", "temporal_status": "localized", "intervals": [ { "clip_relative_s": [ 5.0, 9.0 ], "source_video_s": [ 52.5, 56.5 ] } ], "mask_relation": "subject matches the supplied selected-person mask", "evidence": "From 5–8 seconds, the masked person lowers both hands away from the frame and bends toward the lower work area; the cloth itself is poorly resolved." }, { "id": "a4", "mentions": [ { "text": "picks up", "ranges": [ [ 100, 108 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "type": "phrasal", "subject_entity_group_id": "e1", "object_entity_group_ids": [ "e5" ], "confidence": "medium", "temporal_status": "localized", "intervals": [ { "clip_relative_s": [ 8.0, 12.0 ], "source_video_s": [ 55.5, 59.5 ] } ], "mask_relation": "subject matches the supplied selected-person mask", "evidence": "At 8–11 seconds, the masked person bends and manipulates something below the window; by 12 seconds the squeegee has been raised into view." }, { "id": "a5", "mentions": [ { "text": "wiping", "ranges": [ [ 138, 144 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "type": "simple", "subject_entity_group_id": "e1", "object_entity_group_ids": [ "e6" ], "instrument_entity_group_id": "e5", "confidence": "high", "temporal_status": "localized", "intervals": [ { "clip_relative_s": [ 12.0, 21.5 ], "source_video_s": [ 59.5, 69.0 ] } ], "mask_relation": "subject matches the supplied selected-person mask", "evidence": "From 12 seconds through the clip boundary, the masked person moves the metal squeegee along the upper glass and then across the glass surface." } ], "notes": [ "The supplied masklet consistently identifies the sole visible person and links both occurrences of “He” to entity group e1.", "No separate source masks were supplied for the cloth, window frame, window, squeegee, or glass.", "The spatial part phrase “the top edge” is not a separate entity group; “the glass” is the grounded object of the final wiping action.", "The aspectual verb “begins” is excluded from the final action span; only “wiping” is selected." ] }, "normalization_repairs": 0 }, { "source_video_id": "v_9A0BwCxAKAg", "caption_count_category": 14, "clip": { "caption_index": 6, "caption": "Continuing to hold the squeegee, he makes several horizontal passes from left to right across the top section of the window. He then switches to cleaning the glass with downward vertical strokes.", "source_interval_sec": [ 69.0, 103.0 ], "loom_clip_id": "v_9A0BwCxAKAg_2068_3087", "clip_frame_interval_4fps": [ 2068, 3087 ], "sample_count_1fps": 34, "sampled_frames": [ { "frame": "00000", "timestamp_sec": 69.0, "offset_sec": 0, "rgb_path": "assets/cat_14/v_9A0BwCxAKAg_2068_3087/0000s_00000_rgb.jpg", "mask_path": "assets/cat_14/v_9A0BwCxAKAg_2068_3087/0000s_00000_mask.png", "mask_status": "available" }, { "frame": "00028", "timestamp_sec": 70.0, "offset_sec": 1, "rgb_path": "assets/cat_14/v_9A0BwCxAKAg_2068_3087/0001s_00028_rgb.jpg", "mask_path": "assets/cat_14/v_9A0BwCxAKAg_2068_3087/0001s_00028_mask.png", "mask_status": "available" }, { "frame": "00060", "timestamp_sec": 71.0, "offset_sec": 2, "rgb_path": "assets/cat_14/v_9A0BwCxAKAg_2068_3087/0002s_00060_rgb.jpg", "mask_path": "assets/cat_14/v_9A0BwCxAKAg_2068_3087/0002s_00060_mask.png", "mask_status": "available" }, { "frame": "00088", "timestamp_sec": 72.0, "offset_sec": 3, "rgb_path": "assets/cat_14/v_9A0BwCxAKAg_2068_3087/0003s_00088_rgb.jpg", "mask_path": "assets/cat_14/v_9A0BwCxAKAg_2068_3087/0003s_00088_mask.png", "mask_status": "available" }, { "frame": "00120", "timestamp_sec": 73.0, "offset_sec": 4, "rgb_path": "assets/cat_14/v_9A0BwCxAKAg_2068_3087/0004s_00120_rgb.jpg", "mask_path": "assets/cat_14/v_9A0BwCxAKAg_2068_3087/0004s_00120_mask.png", "mask_status": "available" }, { "frame": "00148", "timestamp_sec": 74.0, "offset_sec": 5, "rgb_path": "assets/cat_14/v_9A0BwCxAKAg_2068_3087/0005s_00148_rgb.jpg", "mask_path": "assets/cat_14/v_9A0BwCxAKAg_2068_3087/0005s_00148_mask.png", "mask_status": "available" }, { "frame": "00180", "timestamp_sec": 75.0, "offset_sec": 6, "rgb_path": "assets/cat_14/v_9A0BwCxAKAg_2068_3087/0006s_00180_rgb.jpg", "mask_path": "assets/cat_14/v_9A0BwCxAKAg_2068_3087/0006s_00180_mask.png", "mask_status": "available" }, { "frame": "00208", "timestamp_sec": 76.0, "offset_sec": 7, "rgb_path": "assets/cat_14/v_9A0BwCxAKAg_2068_3087/0007s_00208_rgb.jpg", "mask_path": "assets/cat_14/v_9A0BwCxAKAg_2068_3087/0007s_00208_mask.png", "mask_status": "available" }, { "frame": "00240", "timestamp_sec": 77.0, "offset_sec": 8, "rgb_path": "assets/cat_14/v_9A0BwCxAKAg_2068_3087/0008s_00240_rgb.jpg", "mask_path": "assets/cat_14/v_9A0BwCxAKAg_2068_3087/0008s_00240_mask.png", "mask_status": "available" }, { "frame": "00268", "timestamp_sec": 78.0, "offset_sec": 9, "rgb_path": "assets/cat_14/v_9A0BwCxAKAg_2068_3087/0009s_00268_rgb.jpg", "mask_path": "assets/cat_14/v_9A0BwCxAKAg_2068_3087/0009s_00268_mask.png", "mask_status": "available" }, { "frame": "00300", "timestamp_sec": 79.0, "offset_sec": 10, "rgb_path": "assets/cat_14/v_9A0BwCxAKAg_2068_3087/0010s_00300_rgb.jpg", "mask_path": "assets/cat_14/v_9A0BwCxAKAg_2068_3087/0010s_00300_mask.png", "mask_status": "available" }, { "frame": "00328", "timestamp_sec": 80.0, "offset_sec": 11, "rgb_path": "assets/cat_14/v_9A0BwCxAKAg_2068_3087/0011s_00328_rgb.jpg", "mask_path": "assets/cat_14/v_9A0BwCxAKAg_2068_3087/0011s_00328_mask.png", "mask_status": "available" }, { "frame": "00360", "timestamp_sec": 81.0, "offset_sec": 12, "rgb_path": "assets/cat_14/v_9A0BwCxAKAg_2068_3087/0012s_00360_rgb.jpg", "mask_path": "assets/cat_14/v_9A0BwCxAKAg_2068_3087/0012s_00360_mask.png", "mask_status": "available" }, { "frame": "00388", "timestamp_sec": 82.0, "offset_sec": 13, "rgb_path": "assets/cat_14/v_9A0BwCxAKAg_2068_3087/0013s_00388_rgb.jpg", "mask_path": "assets/cat_14/v_9A0BwCxAKAg_2068_3087/0013s_00388_mask.png", "mask_status": "available" }, { "frame": "00420", "timestamp_sec": 83.0, "offset_sec": 14, "rgb_path": "assets/cat_14/v_9A0BwCxAKAg_2068_3087/0014s_00420_rgb.jpg", "mask_path": "assets/cat_14/v_9A0BwCxAKAg_2068_3087/0014s_00420_mask.png", "mask_status": "available" }, { "frame": "00448", "timestamp_sec": 84.0, "offset_sec": 15, "rgb_path": "assets/cat_14/v_9A0BwCxAKAg_2068_3087/0015s_00448_rgb.jpg", "mask_path": "assets/cat_14/v_9A0BwCxAKAg_2068_3087/0015s_00448_mask.png", "mask_status": "available" }, { "frame": "00480", "timestamp_sec": 85.0, "offset_sec": 16, "rgb_path": "assets/cat_14/v_9A0BwCxAKAg_2068_3087/0016s_00480_rgb.jpg", "mask_path": "assets/cat_14/v_9A0BwCxAKAg_2068_3087/0016s_00480_mask.png", "mask_status": "available" }, { "frame": "00508", "timestamp_sec": 86.0, "offset_sec": 17, "rgb_path": "assets/cat_14/v_9A0BwCxAKAg_2068_3087/0017s_00508_rgb.jpg", "mask_path": "assets/cat_14/v_9A0BwCxAKAg_2068_3087/0017s_00508_mask.png", "mask_status": "available" }, { "frame": "00540", "timestamp_sec": 87.0, "offset_sec": 18, "rgb_path": "assets/cat_14/v_9A0BwCxAKAg_2068_3087/0018s_00540_rgb.jpg", "mask_path": "assets/cat_14/v_9A0BwCxAKAg_2068_3087/0018s_00540_mask.png", "mask_status": "available" }, { "frame": "00568", "timestamp_sec": 88.0, "offset_sec": 19, "rgb_path": "assets/cat_14/v_9A0BwCxAKAg_2068_3087/0019s_00568_rgb.jpg", "mask_path": "assets/cat_14/v_9A0BwCxAKAg_2068_3087/0019s_00568_mask.png", "mask_status": "available" }, { "frame": "00600", "timestamp_sec": 89.0, "offset_sec": 20, "rgb_path": "assets/cat_14/v_9A0BwCxAKAg_2068_3087/0020s_00600_rgb.jpg", "mask_path": "assets/cat_14/v_9A0BwCxAKAg_2068_3087/0020s_00600_mask.png", "mask_status": "available" }, { "frame": "00628", "timestamp_sec": 90.0, "offset_sec": 21, "rgb_path": "assets/cat_14/v_9A0BwCxAKAg_2068_3087/0021s_00628_rgb.jpg", "mask_path": "assets/cat_14/v_9A0BwCxAKAg_2068_3087/0021s_00628_mask.png", "mask_status": "available" }, { "frame": "00660", "timestamp_sec": 91.0, "offset_sec": 22, "rgb_path": "assets/cat_14/v_9A0BwCxAKAg_2068_3087/0022s_00660_rgb.jpg", "mask_path": "assets/cat_14/v_9A0BwCxAKAg_2068_3087/0022s_00660_mask.png", "mask_status": "available" }, { "frame": "00688", "timestamp_sec": 92.0, "offset_sec": 23, "rgb_path": "assets/cat_14/v_9A0BwCxAKAg_2068_3087/0023s_00688_rgb.jpg", "mask_path": "assets/cat_14/v_9A0BwCxAKAg_2068_3087/0023s_00688_mask.png", "mask_status": "available" }, { "frame": "00720", "timestamp_sec": 93.0, "offset_sec": 24, "rgb_path": "assets/cat_14/v_9A0BwCxAKAg_2068_3087/0024s_00720_rgb.jpg", "mask_path": "assets/cat_14/v_9A0BwCxAKAg_2068_3087/0024s_00720_mask.png", "mask_status": "available" }, { "frame": "00748", "timestamp_sec": 94.0, "offset_sec": 25, "rgb_path": "assets/cat_14/v_9A0BwCxAKAg_2068_3087/0025s_00748_rgb.jpg", "mask_path": "assets/cat_14/v_9A0BwCxAKAg_2068_3087/0025s_00748_mask.png", "mask_status": "available" }, { "frame": "00780", "timestamp_sec": 95.0, "offset_sec": 26, "rgb_path": "assets/cat_14/v_9A0BwCxAKAg_2068_3087/0026s_00780_rgb.jpg", "mask_path": "assets/cat_14/v_9A0BwCxAKAg_2068_3087/0026s_00780_mask.png", "mask_status": "available" }, { "frame": "00808", "timestamp_sec": 96.0, "offset_sec": 27, "rgb_path": "assets/cat_14/v_9A0BwCxAKAg_2068_3087/0027s_00808_rgb.jpg", "mask_path": "assets/cat_14/v_9A0BwCxAKAg_2068_3087/0027s_00808_mask.png", "mask_status": "available" }, { "frame": "00840", "timestamp_sec": 97.0, "offset_sec": 28, "rgb_path": "assets/cat_14/v_9A0BwCxAKAg_2068_3087/0028s_00840_rgb.jpg", "mask_path": "assets/cat_14/v_9A0BwCxAKAg_2068_3087/0028s_00840_mask.png", "mask_status": "available" }, { "frame": "00868", "timestamp_sec": 98.0, "offset_sec": 29, "rgb_path": "assets/cat_14/v_9A0BwCxAKAg_2068_3087/0029s_00868_rgb.jpg", "mask_path": "assets/cat_14/v_9A0BwCxAKAg_2068_3087/0029s_00868_mask.png", "mask_status": "available" }, { "frame": "00900", "timestamp_sec": 99.0, "offset_sec": 30, "rgb_path": "assets/cat_14/v_9A0BwCxAKAg_2068_3087/0030s_00900_rgb.jpg", "mask_path": "assets/cat_14/v_9A0BwCxAKAg_2068_3087/0030s_00900_mask.png", "mask_status": "available" }, { "frame": "00928", "timestamp_sec": 100.0, "offset_sec": 31, "rgb_path": "assets/cat_14/v_9A0BwCxAKAg_2068_3087/0031s_00928_rgb.jpg", "mask_path": "assets/cat_14/v_9A0BwCxAKAg_2068_3087/0031s_00928_mask.png", "mask_status": "available" }, { "frame": "00960", "timestamp_sec": 101.0, "offset_sec": 32, "rgb_path": "assets/cat_14/v_9A0BwCxAKAg_2068_3087/0032s_00960_rgb.jpg", "mask_path": "assets/cat_14/v_9A0BwCxAKAg_2068_3087/0032s_00960_mask.png", "mask_status": "available" }, { "frame": "00988", "timestamp_sec": 102.0, "offset_sec": 33, "rgb_path": "assets/cat_14/v_9A0BwCxAKAg_2068_3087/0033s_00988_rgb.jpg", "mask_path": "assets/cat_14/v_9A0BwCxAKAg_2068_3087/0033s_00988_mask.png", "mask_status": "available" } ], "obj_id": [ 0 ], "anno_id": [ 1268 ], "fps_inferred_from_clip_interval": 29.9706 }, "usage": { "input_tokens": 41678, "input_tokens_details": { "cache_write_tokens": 41675, "cached_tokens": 0 }, "output_tokens": 4369, "output_tokens_details": { "reasoning_tokens": 2971 }, "total_tokens": 46047 }, "model": "gpt-5.6-sol", "effort": "medium", "custom_id": "cat14_cap06_v_9A0BwCxAKAg_2068_3087", "annotation": { "caption": "Continuing to hold the squeegee, he makes several horizontal passes from left to right across the top section of the window. He then switches to cleaning the glass with downward vertical strokes.", "entity_groups": [ { "id": "person_1", "label": "person", "spans": [ { "text": "he", "range": [ 33, 35 ] }, { "text": "He", "range": [ 125, 127 ] } ], "source_mask_link": "selected_person" }, { "id": "squeegee_1", "label": "squeegee", "spans": [ { "text": "the squeegee", "range": [ 19, 31 ] } ], "source_mask_link": null }, { "id": "window_glass_1", "label": "window glass", "spans": [ { "text": "the window", "range": [ 113, 123 ] }, { "text": "the glass", "range": [ 154, 163 ] } ], "source_mask_link": null } ], "actions": [ { "id": "action_1", "mentions": [ { "text": "hold", "ranges": [ [ 14, 18 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "type": "simple", "subject_refs": [ "person_1" ], "object_refs": [ "squeegee_1" ], "confidence": "high", "temporal_status": "localized_sparse", "intervals": [ { "start_s": 0.0, "end_s": 2.0, "source_start_s": 69.0, "source_end_s": 71.0 }, { "start_s": 5.0, "end_s": 11.0, "source_start_s": 74.0, "source_end_s": 80.0 }, { "start_s": 22.0, "end_s": 27.0, "source_start_s": 91.0, "source_end_s": 96.0 }, { "start_s": 31.0, "end_s": 34.0, "source_start_s": 100.0, "source_end_s": 103.0 } ], "evidence": "The masked person visibly grips a window-cleaning tool in the supported samples; unsupported or occluded gaps are omitted.", "mask_relation": "subject matches the supplied selected-person mask", "source_subject_refs": "person_1" }, { "id": "action_2", "mentions": [ { "text": "makes several horizontal passes", "ranges": [ [ 36, 67 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "type": "light_verb", "subject_refs": [ "person_1" ], "object_refs": [ "window_glass_1" ], "confidence": "high", "temporal_status": "localized_sparse", "intervals": [ { "start_s": 5.0, "end_s": 10.0, "source_start_s": 74.0, "source_end_s": 79.0 } ], "evidence": "At clip-relative 5–9 seconds, the masked person holds the squeegee horizontally at successive positions across the upper window.", "mask_relation": "subject matches the supplied selected-person mask", "source_subject_refs": "person_1" }, { "id": "action_3", "mentions": [ { "text": "cleaning", "ranges": [ [ 145, 153 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "type": "simple", "subject_refs": [ "person_1" ], "object_refs": [ "window_glass_1" ], "confidence": "high", "temporal_status": "localized_sparse", "intervals": [ { "start_s": 13.0, "end_s": 20.0, "source_start_s": 82.0, "source_end_s": 89.0 }, { "start_s": 22.0, "end_s": 27.0, "source_start_s": 91.0, "source_end_s": 96.0 }, { "start_s": 31.0, "end_s": 34.0, "source_start_s": 100.0, "source_end_s": 103.0 } ], "evidence": "The masked person's hand or cleaning head moves from high to lower positions on the glass, with especially clear repeated downward passes at 22–26 seconds; the final supported sequence reaches the clip boundary.", "mask_relation": "subject matches the supplied selected-person mask", "source_subject_refs": "person_1" } ], "notes": [ "The supplied masklet identifies only the person referenced by both “he” and “He.”", "No source masks are supplied for the squeegee or window glass.", "“the window” and “the glass” are grouped as two references to the same visible window pane.", "“Continuing” and “switches” frame embedded actions and are excluded from the action spans.", "The action-naming phrases “several horizontal passes” and “downward vertical strokes” are not treated as concrete entity groups." ] }, "model_output_raw": { "caption": "Continuing to hold the squeegee, he makes several horizontal passes from left to right across the top section of the window. He then switches to cleaning the glass with downward vertical strokes.", "entity_groups": [ { "id": "person_1", "label": "person", "spans": [ { "start": 33, "end": 35, "text": "he" }, { "start": 125, "end": 127, "text": "He" } ], "source_mask_link": "selected_person_masklet" }, { "id": "squeegee_1", "label": "squeegee", "spans": [ { "start": 19, "end": 31, "text": "the squeegee" } ], "source_mask_link": null }, { "id": "window_glass_1", "label": "window glass", "spans": [ { "start": 113, "end": 123, "text": "the window" }, { "start": 154, "end": 163, "text": "the glass" } ], "source_mask_link": null } ], "actions": [ { "id": "action_1", "mentions": [ { "text": "hold", "occurrence": 1, "ranges": [ { "start": 14, "end": 18 } ], "particle": { "text": null, "occurrence": 1 } } ], "type": "simple", "subject_entity_group_id": "person_1", "object_entity_group_ids": [ "squeegee_1" ], "confidence": "high", "temporal_status": "localized_sparse", "intervals": [ { "start_s": 0.0, "end_s": 2.0, "source_start_s": 69.0, "source_end_s": 71.0 }, { "start_s": 5.0, "end_s": 11.0, "source_start_s": 74.0, "source_end_s": 80.0 }, { "start_s": 22.0, "end_s": 27.0, "source_start_s": 91.0, "source_end_s": 96.0 }, { "start_s": 31.0, "end_s": 34.0, "source_start_s": 100.0, "source_end_s": 103.0 } ], "evidence": "The masked person visibly grips a window-cleaning tool in the supported samples; unsupported or occluded gaps are omitted.", "mask_relation": "subject matches the supplied selected-person mask" }, { "id": "action_2", "mentions": [ { "text": "makes several horizontal passes", "occurrence": 1, "ranges": [ { "start": 36, "end": 67 } ], "particle": { "text": null, "occurrence": 1 } } ], "type": "light_verb", "subject_entity_group_id": "person_1", "object_entity_group_ids": [ "window_glass_1" ], "confidence": "high", "temporal_status": "localized_sparse", "intervals": [ { "start_s": 5.0, "end_s": 10.0, "source_start_s": 74.0, "source_end_s": 79.0 } ], "evidence": "At clip-relative 5–9 seconds, the masked person holds the squeegee horizontally at successive positions across the upper window.", "mask_relation": "subject matches the supplied selected-person mask" }, { "id": "action_3", "mentions": [ { "text": "cleaning", "occurrence": 1, "ranges": [ { "start": 145, "end": 153 } ], "particle": { "text": null, "occurrence": 1 } } ], "type": "simple", "subject_entity_group_id": "person_1", "object_entity_group_ids": [ "window_glass_1" ], "confidence": "high", "temporal_status": "localized_sparse", "intervals": [ { "start_s": 13.0, "end_s": 20.0, "source_start_s": 82.0, "source_end_s": 89.0 }, { "start_s": 22.0, "end_s": 27.0, "source_start_s": 91.0, "source_end_s": 96.0 }, { "start_s": 31.0, "end_s": 34.0, "source_start_s": 100.0, "source_end_s": 103.0 } ], "evidence": "The masked person's hand or cleaning head moves from high to lower positions on the glass, with especially clear repeated downward passes at 22–26 seconds; the final supported sequence reaches the clip boundary.", "mask_relation": "subject matches the supplied selected-person mask" } ], "notes": [ "The supplied masklet identifies only the person referenced by both “he” and “He.”", "No source masks are supplied for the squeegee or window glass.", "“the window” and “the glass” are grouped as two references to the same visible window pane.", "“Continuing” and “switches” frame embedded actions and are excluded from the action spans.", "The action-naming phrases “several horizontal passes” and “downward vertical strokes” are not treated as concrete entity groups." ] }, "normalization_repairs": 0 }, { "source_video_id": "v_9A0BwCxAKAg", "caption_count_category": 14, "clip": { "caption_index": 7, "caption": "The person continues to clean the window with the squeegee, using long, overlapping vertical strokes to clear the soapy water from the left and middle sections. After several strokes, he bends down.", "source_interval_sec": [ 103.0, 138.0 ], "loom_clip_id": "v_9A0BwCxAKAg_3087_4136", "clip_frame_interval_4fps": [ 3087, 4136 ], "sample_count_1fps": 35, "sampled_frames": [ { "frame": "00000", "timestamp_sec": 103.0, "offset_sec": 0, "rgb_path": "assets/cat_14/v_9A0BwCxAKAg_3087_4136/0000s_00000_rgb.jpg", "mask_path": "assets/cat_14/v_9A0BwCxAKAg_3087_4136/0000s_00000_mask.png", "mask_status": "available" }, { "frame": "00028", "timestamp_sec": 104.0, "offset_sec": 1, "rgb_path": "assets/cat_14/v_9A0BwCxAKAg_3087_4136/0001s_00028_rgb.jpg", "mask_path": "assets/cat_14/v_9A0BwCxAKAg_3087_4136/0001s_00028_mask.png", "mask_status": "available" }, { "frame": "00060", "timestamp_sec": 105.0, "offset_sec": 2, "rgb_path": "assets/cat_14/v_9A0BwCxAKAg_3087_4136/0002s_00060_rgb.jpg", "mask_path": "assets/cat_14/v_9A0BwCxAKAg_3087_4136/0002s_00060_mask.png", "mask_status": "available" }, { "frame": "00088", "timestamp_sec": 106.0, "offset_sec": 3, "rgb_path": "assets/cat_14/v_9A0BwCxAKAg_3087_4136/0003s_00088_rgb.jpg", "mask_path": "assets/cat_14/v_9A0BwCxAKAg_3087_4136/0003s_00088_mask.png", "mask_status": "available" }, { "frame": "00120", "timestamp_sec": 107.0, "offset_sec": 4, "rgb_path": "assets/cat_14/v_9A0BwCxAKAg_3087_4136/0004s_00120_rgb.jpg", "mask_path": "assets/cat_14/v_9A0BwCxAKAg_3087_4136/0004s_00120_mask.png", "mask_status": "available" }, { "frame": "00148", "timestamp_sec": 108.0, "offset_sec": 5, "rgb_path": "assets/cat_14/v_9A0BwCxAKAg_3087_4136/0005s_00148_rgb.jpg", "mask_path": "assets/cat_14/v_9A0BwCxAKAg_3087_4136/0005s_00148_mask.png", "mask_status": "available" }, { "frame": "00180", "timestamp_sec": 109.0, "offset_sec": 6, "rgb_path": "assets/cat_14/v_9A0BwCxAKAg_3087_4136/0006s_00180_rgb.jpg", "mask_path": "assets/cat_14/v_9A0BwCxAKAg_3087_4136/0006s_00180_mask.png", "mask_status": "available" }, { "frame": "00208", "timestamp_sec": 110.0, "offset_sec": 7, "rgb_path": "assets/cat_14/v_9A0BwCxAKAg_3087_4136/0007s_00208_rgb.jpg", "mask_path": "assets/cat_14/v_9A0BwCxAKAg_3087_4136/0007s_00208_mask.png", "mask_status": "available" }, { "frame": "00240", "timestamp_sec": 111.0, "offset_sec": 8, "rgb_path": "assets/cat_14/v_9A0BwCxAKAg_3087_4136/0008s_00240_rgb.jpg", "mask_path": "assets/cat_14/v_9A0BwCxAKAg_3087_4136/0008s_00240_mask.png", "mask_status": "available" }, { "frame": "00268", "timestamp_sec": 112.0, "offset_sec": 9, "rgb_path": "assets/cat_14/v_9A0BwCxAKAg_3087_4136/0009s_00268_rgb.jpg", "mask_path": "assets/cat_14/v_9A0BwCxAKAg_3087_4136/0009s_00268_mask.png", "mask_status": "available" }, { "frame": "00300", "timestamp_sec": 113.0, "offset_sec": 10, "rgb_path": "assets/cat_14/v_9A0BwCxAKAg_3087_4136/0010s_00300_rgb.jpg", "mask_path": "assets/cat_14/v_9A0BwCxAKAg_3087_4136/0010s_00300_mask.png", "mask_status": "available" }, { "frame": "00328", "timestamp_sec": 114.0, "offset_sec": 11, "rgb_path": "assets/cat_14/v_9A0BwCxAKAg_3087_4136/0011s_00328_rgb.jpg", "mask_path": "assets/cat_14/v_9A0BwCxAKAg_3087_4136/0011s_00328_mask.png", "mask_status": "available" }, { "frame": "00360", "timestamp_sec": 115.0, "offset_sec": 12, "rgb_path": "assets/cat_14/v_9A0BwCxAKAg_3087_4136/0012s_00360_rgb.jpg", "mask_path": "assets/cat_14/v_9A0BwCxAKAg_3087_4136/0012s_00360_mask.png", "mask_status": "available" }, { "frame": "00388", "timestamp_sec": 116.0, "offset_sec": 13, "rgb_path": "assets/cat_14/v_9A0BwCxAKAg_3087_4136/0013s_00388_rgb.jpg", "mask_path": "assets/cat_14/v_9A0BwCxAKAg_3087_4136/0013s_00388_mask.png", "mask_status": "available" }, { "frame": "00420", "timestamp_sec": 117.0, "offset_sec": 14, "rgb_path": "assets/cat_14/v_9A0BwCxAKAg_3087_4136/0014s_00420_rgb.jpg", "mask_path": "assets/cat_14/v_9A0BwCxAKAg_3087_4136/0014s_00420_mask.png", "mask_status": "available" }, { "frame": "00448", "timestamp_sec": 118.0, "offset_sec": 15, "rgb_path": "assets/cat_14/v_9A0BwCxAKAg_3087_4136/0015s_00448_rgb.jpg", "mask_path": "assets/cat_14/v_9A0BwCxAKAg_3087_4136/0015s_00448_mask.png", "mask_status": "available" }, { "frame": "00480", "timestamp_sec": 119.0, "offset_sec": 16, "rgb_path": "assets/cat_14/v_9A0BwCxAKAg_3087_4136/0016s_00480_rgb.jpg", "mask_path": "assets/cat_14/v_9A0BwCxAKAg_3087_4136/0016s_00480_mask.png", "mask_status": "available" }, { "frame": "00508", "timestamp_sec": 120.0, "offset_sec": 17, "rgb_path": "assets/cat_14/v_9A0BwCxAKAg_3087_4136/0017s_00508_rgb.jpg", "mask_path": "assets/cat_14/v_9A0BwCxAKAg_3087_4136/0017s_00508_mask.png", "mask_status": "available" }, { "frame": "00540", "timestamp_sec": 121.0, "offset_sec": 18, "rgb_path": "assets/cat_14/v_9A0BwCxAKAg_3087_4136/0018s_00540_rgb.jpg", "mask_path": "assets/cat_14/v_9A0BwCxAKAg_3087_4136/0018s_00540_mask.png", "mask_status": "available" }, { "frame": "00568", "timestamp_sec": 122.0, "offset_sec": 19, "rgb_path": "assets/cat_14/v_9A0BwCxAKAg_3087_4136/0019s_00568_rgb.jpg", "mask_path": "assets/cat_14/v_9A0BwCxAKAg_3087_4136/0019s_00568_mask.png", "mask_status": "available" }, { "frame": "00600", "timestamp_sec": 123.0, "offset_sec": 20, "rgb_path": "assets/cat_14/v_9A0BwCxAKAg_3087_4136/0020s_00600_rgb.jpg", "mask_path": "assets/cat_14/v_9A0BwCxAKAg_3087_4136/0020s_00600_mask.png", "mask_status": "available" }, { "frame": "00628", "timestamp_sec": 124.0, "offset_sec": 21, "rgb_path": "assets/cat_14/v_9A0BwCxAKAg_3087_4136/0021s_00628_rgb.jpg", "mask_path": "assets/cat_14/v_9A0BwCxAKAg_3087_4136/0021s_00628_mask.png", "mask_status": "available" }, { "frame": "00660", "timestamp_sec": 125.0, "offset_sec": 22, "rgb_path": "assets/cat_14/v_9A0BwCxAKAg_3087_4136/0022s_00660_rgb.jpg", "mask_path": "assets/cat_14/v_9A0BwCxAKAg_3087_4136/0022s_00660_mask.png", "mask_status": "available" }, { "frame": "00688", "timestamp_sec": 126.0, "offset_sec": 23, "rgb_path": "assets/cat_14/v_9A0BwCxAKAg_3087_4136/0023s_00688_rgb.jpg", "mask_path": "assets/cat_14/v_9A0BwCxAKAg_3087_4136/0023s_00688_mask.png", "mask_status": "available" }, { "frame": "00720", "timestamp_sec": 127.0, "offset_sec": 24, "rgb_path": "assets/cat_14/v_9A0BwCxAKAg_3087_4136/0024s_00720_rgb.jpg", "mask_path": "assets/cat_14/v_9A0BwCxAKAg_3087_4136/0024s_00720_mask.png", "mask_status": "available" }, { "frame": "00748", "timestamp_sec": 128.0, "offset_sec": 25, "rgb_path": "assets/cat_14/v_9A0BwCxAKAg_3087_4136/0025s_00748_rgb.jpg", "mask_path": "assets/cat_14/v_9A0BwCxAKAg_3087_4136/0025s_00748_mask.png", "mask_status": "available" }, { "frame": "00780", "timestamp_sec": 129.0, "offset_sec": 26, "rgb_path": "assets/cat_14/v_9A0BwCxAKAg_3087_4136/0026s_00780_rgb.jpg", "mask_path": "assets/cat_14/v_9A0BwCxAKAg_3087_4136/0026s_00780_mask.png", "mask_status": "available" }, { "frame": "00808", "timestamp_sec": 130.0, "offset_sec": 27, "rgb_path": "assets/cat_14/v_9A0BwCxAKAg_3087_4136/0027s_00808_rgb.jpg", "mask_path": "assets/cat_14/v_9A0BwCxAKAg_3087_4136/0027s_00808_mask.png", "mask_status": "available" }, { "frame": "00840", "timestamp_sec": 131.0, "offset_sec": 28, "rgb_path": "assets/cat_14/v_9A0BwCxAKAg_3087_4136/0028s_00840_rgb.jpg", "mask_path": "assets/cat_14/v_9A0BwCxAKAg_3087_4136/0028s_00840_mask.png", "mask_status": "available" }, { "frame": "00868", "timestamp_sec": 132.0, "offset_sec": 29, "rgb_path": "assets/cat_14/v_9A0BwCxAKAg_3087_4136/0029s_00868_rgb.jpg", "mask_path": "assets/cat_14/v_9A0BwCxAKAg_3087_4136/0029s_00868_mask.png", "mask_status": "available" }, { "frame": "00900", "timestamp_sec": 133.0, "offset_sec": 30, "rgb_path": "assets/cat_14/v_9A0BwCxAKAg_3087_4136/0030s_00900_rgb.jpg", "mask_path": "assets/cat_14/v_9A0BwCxAKAg_3087_4136/0030s_00900_mask.png", "mask_status": "available" }, { "frame": "00928", "timestamp_sec": 134.0, "offset_sec": 31, "rgb_path": "assets/cat_14/v_9A0BwCxAKAg_3087_4136/0031s_00928_rgb.jpg", "mask_path": "assets/cat_14/v_9A0BwCxAKAg_3087_4136/0031s_00928_mask.png", "mask_status": "available" }, { "frame": "00960", "timestamp_sec": 135.0, "offset_sec": 32, "rgb_path": "assets/cat_14/v_9A0BwCxAKAg_3087_4136/0032s_00960_rgb.jpg", "mask_path": "assets/cat_14/v_9A0BwCxAKAg_3087_4136/0032s_00960_mask.png", "mask_status": "available" }, { "frame": "00988", "timestamp_sec": 136.0, "offset_sec": 33, "rgb_path": "assets/cat_14/v_9A0BwCxAKAg_3087_4136/0033s_00988_rgb.jpg", "mask_path": "assets/cat_14/v_9A0BwCxAKAg_3087_4136/0033s_00988_mask.png", "mask_status": "available" }, { "frame": "01020", "timestamp_sec": 137.0, "offset_sec": 34, "rgb_path": "assets/cat_14/v_9A0BwCxAKAg_3087_4136/0034s_01020_rgb.jpg", "mask_path": "assets/cat_14/v_9A0BwCxAKAg_3087_4136/0034s_01020_mask.png", "mask_status": "available" } ], "obj_id": [ 0 ], "anno_id": [ 1269 ], "fps_inferred_from_clip_interval": 29.9714 }, "usage": { "input_tokens": 42455, "input_tokens_details": { "cache_write_tokens": 42452, "cached_tokens": 0 }, "output_tokens": 4511, "output_tokens_details": { "reasoning_tokens": 3364 }, "total_tokens": 46966 }, "model": "gpt-5.6-sol", "effort": "medium", "custom_id": "cat14_cap07_v_9A0BwCxAKAg_3087_4136", "annotation": { "caption": "The person continues to clean the window with the squeegee, using long, overlapping vertical strokes to clear the soapy water from the left and middle sections. After several strokes, he bends down.", "entity_groups": [ { "id": "person_1", "label": "person", "spans": [ { "text": "The person", "range": [ 0, 10 ] }, { "text": "he", "range": [ 184, 186 ] } ], "source_mask_link": "selected_person" }, { "id": "window_1", "label": "window", "spans": [ { "text": "the window", "range": [ 30, 40 ] } ], "source_mask_link": null }, { "id": "squeegee_1", "label": "squeegee", "spans": [ { "text": "the squeegee", "range": [ 46, 58 ] } ], "source_mask_link": null }, { "id": "water_1", "label": "soapy water", "spans": [ { "text": "the soapy water", "range": [ 110, 125 ] } ], "source_mask_link": null } ], "actions": [ { "id": "action_1", "mentions": [ { "text": "clean", "ranges": [ [ 24, 29 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } }, { "text": "clear", "ranges": [ [ 104, 109 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "type": "simple", "subject_refs": [ "person_1" ], "object_refs": [ "window_1", "water_1" ], "confidence": "high", "temporal_status": "localized", "intervals": [ { "start_s": 0.0, "end_s": 29.0, "source_start_s": 103.0, "source_end_s": 132.0 } ], "evidence": "Samples 0–28 show the masked person repeatedly moving a squeegee across the window and progressively working over the pane.", "mask_relation": "subject matches the supplied selected-person mask", "source_subject_refs": "person_1" }, { "id": "action_2", "mentions": [ { "text": "using long, overlapping vertical strokes", "ranges": [ [ 60, 100 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "type": "light_verb", "subject_refs": [ "person_1" ], "object_refs": [ "window_1" ], "confidence": "high", "temporal_status": "localized", "intervals": [ { "start_s": 0.0, "end_s": 29.0, "source_start_s": 103.0, "source_end_s": 132.0 } ], "evidence": "Across samples 0–28, the masked person performs repeated overlapping sweeps over the window with the squeegee.", "mask_relation": "subject matches the supplied selected-person mask", "source_subject_refs": "person_1" }, { "id": "action_3", "mentions": [ { "text": "bends", "ranges": [ [ 187, 192 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "type": "simple", "subject_refs": [ "person_1" ], "object_refs": [], "confidence": "high", "temporal_status": "localized", "intervals": [ { "start_s": 29.0, "end_s": 31.0, "source_start_s": 132.0, "source_end_s": 134.0 } ], "evidence": "Samples 29–30 show the masked person's head and torso lowering after the squeegee strokes; by sample 31 the person is rising or upright again.", "mask_relation": "subject matches the supplied selected-person mask", "source_subject_refs": "person_1" } ], "notes": [ "The supplied masklet consistently identifies the sole person, linking both “The person” and “he” to person_1.", "No separate source masks were supplied for the window, squeegee, or soapy water.", "The directional word “down” is excluded from the “bends” action span.", "The phrases referring to strokes are action descriptions rather than segmentable entity instances; the left and middle sections are spatial regions and are not entity groups." ] }, "model_output_raw": { "caption": "The person continues to clean the window with the squeegee, using long, overlapping vertical strokes to clear the soapy water from the left and middle sections. After several strokes, he bends down.", "entity_groups": [ { "id": "person_1", "entity": "person", "spans": [ { "text": "The person", "range": [ 0, 10 ] }, { "text": "he", "range": [ 184, 186 ] } ], "source_mask": "supplied_selected_person_masklet" }, { "id": "window_1", "entity": "window", "spans": [ { "text": "the window", "range": [ 30, 40 ] } ], "source_mask": null }, { "id": "squeegee_1", "entity": "squeegee", "spans": [ { "text": "the squeegee", "range": [ 46, 58 ] } ], "source_mask": null }, { "id": "water_1", "entity": "soapy water", "spans": [ { "text": "the soapy water", "range": [ 110, 125 ] } ], "source_mask": null } ], "actions": [ { "id": "action_1", "mentions": [ { "text": "clean", "ranges": [ [ 24, 29 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } }, { "text": "clear", "ranges": [ [ 104, 109 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "type": "simple", "subject": "person_1", "objects": [ "window_1", "water_1" ], "instrument": "squeegee_1", "confidence": "high", "mask_relation": "subject matches the supplied selected-person mask", "temporal_status": "localized", "intervals": [ { "clip_relative": [ 0.0, 29.0 ], "source_video": [ 103.0, 132.0 ] } ], "evidence": "Samples 0–28 show the masked person repeatedly moving a squeegee across the window and progressively working over the pane." }, { "id": "action_2", "mentions": [ { "text": "using long, overlapping vertical strokes", "ranges": [ [ 60, 100 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "type": "light_verb", "subject": "person_1", "objects": [ "window_1" ], "instrument": "squeegee_1", "confidence": "high", "mask_relation": "subject matches the supplied selected-person mask", "temporal_status": "localized", "intervals": [ { "clip_relative": [ 0.0, 29.0 ], "source_video": [ 103.0, 132.0 ] } ], "evidence": "Across samples 0–28, the masked person performs repeated overlapping sweeps over the window with the squeegee." }, { "id": "action_3", "mentions": [ { "text": "bends", "ranges": [ [ 187, 192 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "type": "simple", "subject": "person_1", "objects": [], "confidence": "high", "mask_relation": "subject matches the supplied selected-person mask", "temporal_status": "localized", "intervals": [ { "clip_relative": [ 29.0, 31.0 ], "source_video": [ 132.0, 134.0 ] } ], "evidence": "Samples 29–30 show the masked person's head and torso lowering after the squeegee strokes; by sample 31 the person is rising or upright again." } ], "notes": [ "The supplied masklet consistently identifies the sole person, linking both “The person” and “he” to person_1.", "No separate source masks were supplied for the window, squeegee, or soapy water.", "The directional word “down” is excluded from the “bends” action span.", "The phrases referring to strokes are action descriptions rather than segmentable entity instances; the left and middle sections are spatial regions and are not entity groups." ] }, "normalization_repairs": 0 }, { "source_video_id": "v_9A0BwCxAKAg", "caption_count_category": 14, "clip": { "caption_index": 8, "caption": "He straightens up, picks up the yellow sponge squeegee, and re-wets the top portion of the window. He then switches back to the metal squeegee to clean the top part of the glass again.", "source_interval_sec": [ 138.0, 163.5 ], "loom_clip_id": "v_9A0BwCxAKAg_4136_4900", "clip_frame_interval_4fps": [ 4136, 4900 ], "sample_count_1fps": 26, "sampled_frames": [ { "frame": "00000", "timestamp_sec": 138.0, "offset_sec": 0, "rgb_path": "assets/cat_14/v_9A0BwCxAKAg_4136_4900/0000s_00000_rgb.jpg", "mask_path": "assets/cat_14/v_9A0BwCxAKAg_4136_4900/0000s_00000_mask.png", "mask_status": "available" }, { "frame": "00028", "timestamp_sec": 139.0, "offset_sec": 1, "rgb_path": "assets/cat_14/v_9A0BwCxAKAg_4136_4900/0001s_00028_rgb.jpg", "mask_path": "assets/cat_14/v_9A0BwCxAKAg_4136_4900/0001s_00028_mask.png", "mask_status": "available" }, { "frame": "00060", "timestamp_sec": 140.0, "offset_sec": 2, "rgb_path": "assets/cat_14/v_9A0BwCxAKAg_4136_4900/0002s_00060_rgb.jpg", "mask_path": "assets/cat_14/v_9A0BwCxAKAg_4136_4900/0002s_00060_mask.png", "mask_status": "available" }, { "frame": "00088", "timestamp_sec": 141.0, "offset_sec": 3, "rgb_path": "assets/cat_14/v_9A0BwCxAKAg_4136_4900/0003s_00088_rgb.jpg", "mask_path": "assets/cat_14/v_9A0BwCxAKAg_4136_4900/0003s_00088_mask.png", "mask_status": "available" }, { "frame": "00120", "timestamp_sec": 142.0, "offset_sec": 4, "rgb_path": "assets/cat_14/v_9A0BwCxAKAg_4136_4900/0004s_00120_rgb.jpg", "mask_path": "assets/cat_14/v_9A0BwCxAKAg_4136_4900/0004s_00120_mask.png", "mask_status": "available" }, { "frame": "00148", "timestamp_sec": 143.0, "offset_sec": 5, "rgb_path": "assets/cat_14/v_9A0BwCxAKAg_4136_4900/0005s_00148_rgb.jpg", "mask_path": "assets/cat_14/v_9A0BwCxAKAg_4136_4900/0005s_00148_mask.png", "mask_status": "available" }, { "frame": "00180", "timestamp_sec": 144.0, "offset_sec": 6, "rgb_path": "assets/cat_14/v_9A0BwCxAKAg_4136_4900/0006s_00180_rgb.jpg", "mask_path": "assets/cat_14/v_9A0BwCxAKAg_4136_4900/0006s_00180_mask.png", "mask_status": "available" }, { "frame": "00208", "timestamp_sec": 145.0, "offset_sec": 7, "rgb_path": "assets/cat_14/v_9A0BwCxAKAg_4136_4900/0007s_00208_rgb.jpg", "mask_path": "assets/cat_14/v_9A0BwCxAKAg_4136_4900/0007s_00208_mask.png", "mask_status": "available" }, { "frame": "00240", "timestamp_sec": 146.0, "offset_sec": 8, "rgb_path": "assets/cat_14/v_9A0BwCxAKAg_4136_4900/0008s_00240_rgb.jpg", "mask_path": "assets/cat_14/v_9A0BwCxAKAg_4136_4900/0008s_00240_mask.png", "mask_status": "available" }, { "frame": "00268", "timestamp_sec": 147.0, "offset_sec": 9, "rgb_path": "assets/cat_14/v_9A0BwCxAKAg_4136_4900/0009s_00268_rgb.jpg", "mask_path": "assets/cat_14/v_9A0BwCxAKAg_4136_4900/0009s_00268_mask.png", "mask_status": "available" }, { "frame": "00300", "timestamp_sec": 148.0, "offset_sec": 10, "rgb_path": "assets/cat_14/v_9A0BwCxAKAg_4136_4900/0010s_00300_rgb.jpg", "mask_path": "assets/cat_14/v_9A0BwCxAKAg_4136_4900/0010s_00300_mask.png", "mask_status": "available" }, { "frame": "00328", "timestamp_sec": 149.0, "offset_sec": 11, "rgb_path": "assets/cat_14/v_9A0BwCxAKAg_4136_4900/0011s_00328_rgb.jpg", "mask_path": "assets/cat_14/v_9A0BwCxAKAg_4136_4900/0011s_00328_mask.png", "mask_status": "available" }, { "frame": "00360", "timestamp_sec": 150.0, "offset_sec": 12, "rgb_path": "assets/cat_14/v_9A0BwCxAKAg_4136_4900/0012s_00360_rgb.jpg", "mask_path": "assets/cat_14/v_9A0BwCxAKAg_4136_4900/0012s_00360_mask.png", "mask_status": "available" }, { "frame": "00388", "timestamp_sec": 151.0, "offset_sec": 13, "rgb_path": "assets/cat_14/v_9A0BwCxAKAg_4136_4900/0013s_00388_rgb.jpg", "mask_path": "assets/cat_14/v_9A0BwCxAKAg_4136_4900/0013s_00388_mask.png", "mask_status": "available" }, { "frame": "00420", "timestamp_sec": 152.0, "offset_sec": 14, "rgb_path": "assets/cat_14/v_9A0BwCxAKAg_4136_4900/0014s_00420_rgb.jpg", "mask_path": "assets/cat_14/v_9A0BwCxAKAg_4136_4900/0014s_00420_mask.png", "mask_status": "available" }, { "frame": "00448", "timestamp_sec": 153.0, "offset_sec": 15, "rgb_path": "assets/cat_14/v_9A0BwCxAKAg_4136_4900/0015s_00448_rgb.jpg", "mask_path": "assets/cat_14/v_9A0BwCxAKAg_4136_4900/0015s_00448_mask.png", "mask_status": "available" }, { "frame": "00480", "timestamp_sec": 154.0, "offset_sec": 16, "rgb_path": "assets/cat_14/v_9A0BwCxAKAg_4136_4900/0016s_00480_rgb.jpg", "mask_path": "assets/cat_14/v_9A0BwCxAKAg_4136_4900/0016s_00480_mask.png", "mask_status": "available" }, { "frame": "00508", "timestamp_sec": 155.0, "offset_sec": 17, "rgb_path": "assets/cat_14/v_9A0BwCxAKAg_4136_4900/0017s_00508_rgb.jpg", "mask_path": "assets/cat_14/v_9A0BwCxAKAg_4136_4900/0017s_00508_mask.png", "mask_status": "available" }, { "frame": "00540", "timestamp_sec": 156.0, "offset_sec": 18, "rgb_path": "assets/cat_14/v_9A0BwCxAKAg_4136_4900/0018s_00540_rgb.jpg", "mask_path": "assets/cat_14/v_9A0BwCxAKAg_4136_4900/0018s_00540_mask.png", "mask_status": "available" }, { "frame": "00568", "timestamp_sec": 157.0, "offset_sec": 19, "rgb_path": "assets/cat_14/v_9A0BwCxAKAg_4136_4900/0019s_00568_rgb.jpg", "mask_path": "assets/cat_14/v_9A0BwCxAKAg_4136_4900/0019s_00568_mask.png", "mask_status": "available" }, { "frame": "00600", "timestamp_sec": 158.0, "offset_sec": 20, "rgb_path": "assets/cat_14/v_9A0BwCxAKAg_4136_4900/0020s_00600_rgb.jpg", "mask_path": "assets/cat_14/v_9A0BwCxAKAg_4136_4900/0020s_00600_mask.png", "mask_status": "available" }, { "frame": "00628", "timestamp_sec": 159.0, "offset_sec": 21, "rgb_path": "assets/cat_14/v_9A0BwCxAKAg_4136_4900/0021s_00628_rgb.jpg", "mask_path": "assets/cat_14/v_9A0BwCxAKAg_4136_4900/0021s_00628_mask.png", "mask_status": "available" }, { "frame": "00660", "timestamp_sec": 160.0, "offset_sec": 22, "rgb_path": "assets/cat_14/v_9A0BwCxAKAg_4136_4900/0022s_00660_rgb.jpg", "mask_path": "assets/cat_14/v_9A0BwCxAKAg_4136_4900/0022s_00660_mask.png", "mask_status": "available" }, { "frame": "00688", "timestamp_sec": 161.0, "offset_sec": 23, "rgb_path": "assets/cat_14/v_9A0BwCxAKAg_4136_4900/0023s_00688_rgb.jpg", "mask_path": "assets/cat_14/v_9A0BwCxAKAg_4136_4900/0023s_00688_mask.png", "mask_status": "available" }, { "frame": "00720", "timestamp_sec": 162.0, "offset_sec": 24, "rgb_path": "assets/cat_14/v_9A0BwCxAKAg_4136_4900/0024s_00720_rgb.jpg", "mask_path": "assets/cat_14/v_9A0BwCxAKAg_4136_4900/0024s_00720_mask.png", "mask_status": "available" }, { "frame": "00748", "timestamp_sec": 163.0, "offset_sec": 25, "rgb_path": "assets/cat_14/v_9A0BwCxAKAg_4136_4900/0025s_00748_rgb.jpg", "mask_path": "assets/cat_14/v_9A0BwCxAKAg_4136_4900/0025s_00748_mask.png", "mask_status": "available" } ], "obj_id": [ 0 ], "anno_id": [ 1270 ], "fps_inferred_from_clip_interval": 29.9608 }, "usage": { "input_tokens": 35485, "input_tokens_details": { "cache_write_tokens": 35482, "cached_tokens": 0 }, "output_tokens": 4339, "output_tokens_details": { "reasoning_tokens": 2841 }, "total_tokens": 39824 }, "model": "gpt-5.6-sol", "effort": "medium", "custom_id": "cat14_cap08_v_9A0BwCxAKAg_4136_4900", "annotation": { "caption": "He straightens up, picks up the yellow sponge squeegee, and re-wets the top portion of the window. He then switches back to the metal squeegee to clean the top part of the glass again.", "entity_groups": [ { "id": "person_1", "label": "person", "spans": [ { "text": "He", "range": [ 0, 2 ] }, { "text": "He", "range": [ 99, 101 ] } ], "source_mask_link": "selected_person" }, { "id": "sponge_squeegee_1", "label": "sponge squeegee", "spans": [ { "text": "the yellow sponge squeegee", "range": [ 28, 54 ] } ], "source_mask_link": null }, { "id": "window_1", "label": "window/glass", "spans": [ { "text": "the window", "range": [ 87, 97 ] }, { "text": "the glass", "range": [ 168, 177 ] } ], "source_mask_link": null }, { "id": "metal_squeegee_1", "label": "metal squeegee", "spans": [ { "text": "the metal squeegee", "range": [ 124, 142 ] } ], "source_mask_link": null } ], "actions": [ { "id": "action_1", "mentions": [ { "text": "straightens up", "ranges": [ [ 3, 17 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "type": "phrasal", "subject_refs": [ "person_1" ], "object_refs": [], "confidence": "high", "temporal_status": "uncertain", "intervals": [], "evidence": "The selected person is visible at the clip start, but the sparse frames do not clearly capture a distinct straightening transition.", "mask_relation": "subject_matches_supplied_selected_person_mask", "source_subject_refs": [ "person_1" ] }, { "id": "action_2", "mentions": [ { "text": "picks up", "ranges": [ [ 19, 27 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "type": "phrasal", "subject_refs": [ "person_1" ], "object_refs": [ "sponge_squeegee_1" ], "confidence": "high", "temporal_status": "localized", "intervals": [ { "start_s": 1.0, "end_s": 3.0, "source_start_s": 139.0, "source_end_s": 141.0 } ], "evidence": "The sponge squeegee is absent from the person's raised hands at 1.0 s and is held against the window at 2.0 s.", "mask_relation": "subject_matches_supplied_selected_person_mask", "source_subject_refs": [ "person_1" ] }, { "id": "action_3", "mentions": [ { "text": "re-wets", "ranges": [ [ 60, 67 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "type": "simple", "subject_refs": [ "person_1" ], "object_refs": [ "window_1" ], "confidence": "high", "temporal_status": "localized", "intervals": [ { "start_s": 2.0, "end_s": 6.0, "source_start_s": 140.0, "source_end_s": 144.0 } ], "evidence": "From 2.0–5.0 s, the masked person moves the yellow sponge squeegee across the window, including its upper area.", "mask_relation": "subject_matches_supplied_selected_person_mask", "source_subject_refs": [ "person_1" ] }, { "id": "action_4", "mentions": [ { "text": "switches back", "ranges": [ [ 107, 120 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "type": "phrasal", "subject_refs": [ "person_1" ], "object_refs": [ "metal_squeegee_1" ], "confidence": "high", "temporal_status": "localized", "intervals": [ { "start_s": 12.0, "end_s": 14.0, "source_start_s": 150.0, "source_end_s": 152.0 } ], "evidence": "The person bends down at 12.0 s and appears at 13.0 s holding the narrow metal squeegee at the top of the window.", "mask_relation": "subject_matches_supplied_selected_person_mask", "source_subject_refs": [ "person_1" ] }, { "id": "action_5", "mentions": [ { "text": "clean", "ranges": [ [ 146, 151 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "type": "simple", "subject_refs": [ "person_1" ], "object_refs": [ "window_1" ], "confidence": "high", "temporal_status": "localized", "intervals": [ { "start_s": 13.0, "end_s": 24.0, "source_start_s": 151.0, "source_end_s": 162.0 } ], "evidence": "Across samples 13.0–23.0 s, the masked person repeatedly draws the metal squeegee along the upper window glass.", "mask_relation": "subject_matches_supplied_selected_person_mask", "source_subject_refs": [ "person_1" ] } ], "notes": [ "The supplied masklet consistently identifies the sole person referred to by both occurrences of “He.”", "No source masks were supplied for either squeegee or for the window/glass.", "“the top portion” and “the top part” denote spatial regions of the window rather than separate concrete entity instances, so they are not separate entity groups.", "Temporal intervals are sparse support intervals based on the sampled clean RGB frames, not exact continuous-video boundaries." ] }, "model_output_raw": { "caption": "He straightens up, picks up the yellow sponge squeegee, and re-wets the top portion of the window. He then switches back to the metal squeegee to clean the top part of the glass again.", "entity_groups": [ { "id": "person_1", "entity_type": "person", "spans": [ { "text": "He", "range": [ 0, 2 ] }, { "text": "He", "range": [ 99, 101 ] } ], "source_mask": "selected_person_masklet" }, { "id": "sponge_squeegee_1", "entity_type": "sponge squeegee", "spans": [ { "text": "the yellow sponge squeegee", "range": [ 28, 54 ] } ], "source_mask": null }, { "id": "window_1", "entity_type": "window/glass", "spans": [ { "text": "the window", "range": [ 87, 97 ] }, { "text": "the glass", "range": [ 168, 177 ] } ], "source_mask": null }, { "id": "metal_squeegee_1", "entity_type": "metal squeegee", "spans": [ { "text": "the metal squeegee", "range": [ 124, 142 ] } ], "source_mask": null } ], "actions": [ { "id": "action_1", "mentions": [ { "text": "straightens up", "ranges": [ [ 3, 17 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "type": "phrasal", "subject_refs": [ "person_1" ], "object_refs": [], "confidence": "high", "temporal_status": "uncertain", "intervals": [], "evidence": "The selected person is visible at the clip start, but the sparse frames do not clearly capture a distinct straightening transition.", "mask_relation": "subject_matches_supplied_selected_person_mask" }, { "id": "action_2", "mentions": [ { "text": "picks up", "ranges": [ [ 19, 27 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "type": "phrasal", "subject_refs": [ "person_1" ], "object_refs": [ "sponge_squeegee_1" ], "confidence": "high", "temporal_status": "localized", "intervals": [ { "start_s": 1.0, "end_s": 3.0 } ], "evidence": "The sponge squeegee is absent from the person's raised hands at 1.0 s and is held against the window at 2.0 s.", "mask_relation": "subject_matches_supplied_selected_person_mask" }, { "id": "action_3", "mentions": [ { "text": "re-wets", "ranges": [ [ 60, 67 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "type": "simple", "subject_refs": [ "person_1" ], "object_refs": [ "window_1" ], "confidence": "high", "temporal_status": "localized", "intervals": [ { "start_s": 2.0, "end_s": 6.0 } ], "evidence": "From 2.0–5.0 s, the masked person moves the yellow sponge squeegee across the window, including its upper area.", "mask_relation": "subject_matches_supplied_selected_person_mask" }, { "id": "action_4", "mentions": [ { "text": "switches back", "ranges": [ [ 107, 120 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "type": "phrasal", "subject_refs": [ "person_1" ], "object_refs": [ "metal_squeegee_1" ], "confidence": "high", "temporal_status": "localized", "intervals": [ { "start_s": 12.0, "end_s": 14.0 } ], "evidence": "The person bends down at 12.0 s and appears at 13.0 s holding the narrow metal squeegee at the top of the window.", "mask_relation": "subject_matches_supplied_selected_person_mask" }, { "id": "action_5", "mentions": [ { "text": "clean", "ranges": [ [ 146, 151 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "type": "simple", "subject_refs": [ "person_1" ], "object_refs": [ "window_1" ], "confidence": "high", "temporal_status": "localized", "intervals": [ { "start_s": 13.0, "end_s": 24.0 } ], "evidence": "Across samples 13.0–23.0 s, the masked person repeatedly draws the metal squeegee along the upper window glass.", "mask_relation": "subject_matches_supplied_selected_person_mask" } ], "notes": [ "The supplied masklet consistently identifies the sole person referred to by both occurrences of “He.”", "No source masks were supplied for either squeegee or for the window/glass.", "“the top portion” and “the top part” denote spatial regions of the window rather than separate concrete entity instances, so they are not separate entity groups.", "Temporal intervals are sparse support intervals based on the sampled clean RGB frames, not exact continuous-video boundaries." ] }, "normalization_repairs": 0 }, { "source_video_id": "v_9A0BwCxAKAg", "caption_count_category": 14, "clip": { "caption_index": 9, "caption": "The person turns and gestures before using the squeegee to clean the lower right corner of the window. He then puts the squeegee down, bends over, and moves out of the frame to the left.", "source_interval_sec": [ 163.5, 192.96 ], "loom_clip_id": "v_9A0BwCxAKAg_4900_5783", "clip_frame_interval_4fps": [ 4900, 5783 ], "sample_count_1fps": 30, "sampled_frames": [ { "frame": "00000", "timestamp_sec": 163.5, "offset_sec": 0, "rgb_path": "assets/cat_14/v_9A0BwCxAKAg_4900_5783/0000s_00000_rgb.jpg", "mask_path": "assets/cat_14/v_9A0BwCxAKAg_4900_5783/0000s_00000_mask.png", "mask_status": "available" }, { "frame": "00028", "timestamp_sec": 164.5, "offset_sec": 1, "rgb_path": "assets/cat_14/v_9A0BwCxAKAg_4900_5783/0001s_00028_rgb.jpg", "mask_path": "assets/cat_14/v_9A0BwCxAKAg_4900_5783/0001s_00028_mask.png", "mask_status": "available" }, { "frame": "00060", "timestamp_sec": 165.5, "offset_sec": 2, "rgb_path": "assets/cat_14/v_9A0BwCxAKAg_4900_5783/0002s_00060_rgb.jpg", "mask_path": "assets/cat_14/v_9A0BwCxAKAg_4900_5783/0002s_00060_mask.png", "mask_status": "available" }, { "frame": "00088", "timestamp_sec": 166.5, "offset_sec": 3, "rgb_path": "assets/cat_14/v_9A0BwCxAKAg_4900_5783/0003s_00088_rgb.jpg", "mask_path": "assets/cat_14/v_9A0BwCxAKAg_4900_5783/0003s_00088_mask.png", "mask_status": "available" }, { "frame": "00120", "timestamp_sec": 167.5, "offset_sec": 4, "rgb_path": "assets/cat_14/v_9A0BwCxAKAg_4900_5783/0004s_00120_rgb.jpg", "mask_path": "assets/cat_14/v_9A0BwCxAKAg_4900_5783/0004s_00120_mask.png", "mask_status": "available" }, { "frame": "00148", "timestamp_sec": 168.5, "offset_sec": 5, "rgb_path": "assets/cat_14/v_9A0BwCxAKAg_4900_5783/0005s_00148_rgb.jpg", "mask_path": "assets/cat_14/v_9A0BwCxAKAg_4900_5783/0005s_00148_mask.png", "mask_status": "available" }, { "frame": "00180", "timestamp_sec": 169.5, "offset_sec": 6, "rgb_path": "assets/cat_14/v_9A0BwCxAKAg_4900_5783/0006s_00180_rgb.jpg", "mask_path": "assets/cat_14/v_9A0BwCxAKAg_4900_5783/0006s_00180_mask.png", "mask_status": "available" }, { "frame": "00208", "timestamp_sec": 170.5, "offset_sec": 7, "rgb_path": "assets/cat_14/v_9A0BwCxAKAg_4900_5783/0007s_00208_rgb.jpg", "mask_path": "assets/cat_14/v_9A0BwCxAKAg_4900_5783/0007s_00208_mask.png", "mask_status": "available" }, { "frame": "00240", "timestamp_sec": 171.5, "offset_sec": 8, "rgb_path": "assets/cat_14/v_9A0BwCxAKAg_4900_5783/0008s_00240_rgb.jpg", "mask_path": "assets/cat_14/v_9A0BwCxAKAg_4900_5783/0008s_00240_mask.png", "mask_status": "available" }, { "frame": "00268", "timestamp_sec": 172.5, "offset_sec": 9, "rgb_path": "assets/cat_14/v_9A0BwCxAKAg_4900_5783/0009s_00268_rgb.jpg", "mask_path": "assets/cat_14/v_9A0BwCxAKAg_4900_5783/0009s_00268_mask.png", "mask_status": "available" }, { "frame": "00300", "timestamp_sec": 173.5, "offset_sec": 10, "rgb_path": "assets/cat_14/v_9A0BwCxAKAg_4900_5783/0010s_00300_rgb.jpg", "mask_path": "assets/cat_14/v_9A0BwCxAKAg_4900_5783/0010s_00300_mask.png", "mask_status": "available" }, { "frame": "00328", "timestamp_sec": 174.5, "offset_sec": 11, "rgb_path": "assets/cat_14/v_9A0BwCxAKAg_4900_5783/0011s_00328_rgb.jpg", "mask_path": "assets/cat_14/v_9A0BwCxAKAg_4900_5783/0011s_00328_mask.png", "mask_status": "available" }, { "frame": "00360", "timestamp_sec": 175.5, "offset_sec": 12, "rgb_path": "assets/cat_14/v_9A0BwCxAKAg_4900_5783/0012s_00360_rgb.jpg", "mask_path": "assets/cat_14/v_9A0BwCxAKAg_4900_5783/0012s_00360_mask.png", "mask_status": "available" }, { "frame": "00388", "timestamp_sec": 176.5, "offset_sec": 13, "rgb_path": "assets/cat_14/v_9A0BwCxAKAg_4900_5783/0013s_00388_rgb.jpg", "mask_path": "assets/cat_14/v_9A0BwCxAKAg_4900_5783/0013s_00388_mask.png", "mask_status": "available" }, { "frame": "00420", "timestamp_sec": 177.5, "offset_sec": 14, "rgb_path": "assets/cat_14/v_9A0BwCxAKAg_4900_5783/0014s_00420_rgb.jpg", "mask_path": "assets/cat_14/v_9A0BwCxAKAg_4900_5783/0014s_00420_mask.png", "mask_status": "available" }, { "frame": "00448", "timestamp_sec": 178.5, "offset_sec": 15, "rgb_path": "assets/cat_14/v_9A0BwCxAKAg_4900_5783/0015s_00448_rgb.jpg", "mask_path": "assets/cat_14/v_9A0BwCxAKAg_4900_5783/0015s_00448_mask.png", "mask_status": "available" }, { "frame": "00480", "timestamp_sec": 179.5, "offset_sec": 16, "rgb_path": "assets/cat_14/v_9A0BwCxAKAg_4900_5783/0016s_00480_rgb.jpg", "mask_path": "assets/cat_14/v_9A0BwCxAKAg_4900_5783/0016s_00480_mask.png", "mask_status": "available" }, { "frame": "00508", "timestamp_sec": 180.5, "offset_sec": 17, "rgb_path": "assets/cat_14/v_9A0BwCxAKAg_4900_5783/0017s_00508_rgb.jpg", "mask_path": "assets/cat_14/v_9A0BwCxAKAg_4900_5783/0017s_00508_mask.png", "mask_status": "available" }, { "frame": "00540", "timestamp_sec": 181.5, "offset_sec": 18, "rgb_path": "assets/cat_14/v_9A0BwCxAKAg_4900_5783/0018s_00540_rgb.jpg", "mask_path": "assets/cat_14/v_9A0BwCxAKAg_4900_5783/0018s_00540_mask.png", "mask_status": "available" }, { "frame": "00568", "timestamp_sec": 182.5, "offset_sec": 19, "rgb_path": "assets/cat_14/v_9A0BwCxAKAg_4900_5783/0019s_00568_rgb.jpg", "mask_path": "assets/cat_14/v_9A0BwCxAKAg_4900_5783/0019s_00568_mask.png", "mask_status": "available" }, { "frame": "00600", "timestamp_sec": 183.5, "offset_sec": 20, "rgb_path": "assets/cat_14/v_9A0BwCxAKAg_4900_5783/0020s_00600_rgb.jpg", "mask_path": "assets/cat_14/v_9A0BwCxAKAg_4900_5783/0020s_00600_mask.png", "mask_status": "available" }, { "frame": "00628", "timestamp_sec": 184.5, "offset_sec": 21, "rgb_path": "assets/cat_14/v_9A0BwCxAKAg_4900_5783/0021s_00628_rgb.jpg", "mask_path": "assets/cat_14/v_9A0BwCxAKAg_4900_5783/0021s_00628_mask.png", "mask_status": "available" }, { "frame": "00660", "timestamp_sec": 185.5, "offset_sec": 22, "rgb_path": "assets/cat_14/v_9A0BwCxAKAg_4900_5783/0022s_00660_rgb.jpg", "mask_path": "assets/cat_14/v_9A0BwCxAKAg_4900_5783/0022s_00660_mask.png", "mask_status": "available" }, { "frame": "00688", "timestamp_sec": 186.5, "offset_sec": 23, "rgb_path": "assets/cat_14/v_9A0BwCxAKAg_4900_5783/0023s_00688_rgb.jpg", "mask_path": "assets/cat_14/v_9A0BwCxAKAg_4900_5783/0023s_00688_mask.png", "mask_status": "available" }, { "frame": "00720", "timestamp_sec": 187.5, "offset_sec": 24, "rgb_path": "assets/cat_14/v_9A0BwCxAKAg_4900_5783/0024s_00720_rgb.jpg", "mask_path": "assets/cat_14/v_9A0BwCxAKAg_4900_5783/0024s_00720_mask.png", "mask_status": "available" }, { "frame": "00748", "timestamp_sec": 188.5, "offset_sec": 25, "rgb_path": "assets/cat_14/v_9A0BwCxAKAg_4900_5783/0025s_00748_rgb.jpg", "mask_path": "assets/cat_14/v_9A0BwCxAKAg_4900_5783/0025s_00748_mask.png", "mask_status": "available" }, { "frame": "00780", "timestamp_sec": 189.5, "offset_sec": 26, "rgb_path": "assets/cat_14/v_9A0BwCxAKAg_4900_5783/0026s_00780_rgb.jpg", "mask_path": "assets/cat_14/v_9A0BwCxAKAg_4900_5783/0026s_00780_mask.png", "mask_status": "available" }, { "frame": "00808", "timestamp_sec": 190.5, "offset_sec": 27, "rgb_path": "assets/cat_14/v_9A0BwCxAKAg_4900_5783/0027s_00808_rgb.jpg", "mask_path": "assets/cat_14/v_9A0BwCxAKAg_4900_5783/0027s_00808_mask.png", "mask_status": "available" }, { "frame": "00840", "timestamp_sec": 191.5, "offset_sec": 28, "rgb_path": "assets/cat_14/v_9A0BwCxAKAg_4900_5783/0028s_00840_rgb.jpg", "mask_path": "assets/cat_14/v_9A0BwCxAKAg_4900_5783/0028s_00840_mask.png", "mask_status": "available" }, { "frame": "00868", "timestamp_sec": 192.5, "offset_sec": 29, "rgb_path": "assets/cat_14/v_9A0BwCxAKAg_4900_5783/0029s_00868_rgb.jpg", "mask_path": "assets/cat_14/v_9A0BwCxAKAg_4900_5783/0029s_00868_mask.png", "mask_status": "available" } ], "obj_id": [ 0 ], "anno_id": [ 1272 ], "fps_inferred_from_clip_interval": 29.9728 }, "usage": { "input_tokens": 38583, "input_tokens_details": { "cache_write_tokens": 38580, "cached_tokens": 0 }, "output_tokens": 4420, "output_tokens_details": { "reasoning_tokens": 2588 }, "total_tokens": 43003 }, "model": "gpt-5.6-sol", "effort": "medium", "custom_id": "cat14_cap09_v_9A0BwCxAKAg_4900_5783", "annotation": { "caption": "The person turns and gestures before using the squeegee to clean the lower right corner of the window. He then puts the squeegee down, bends over, and moves out of the frame to the left.", "entity_groups": [ { "id": "entity_person_1", "label": "person", "spans": [ { "text": "The person", "range": [ 0, 10 ] }, { "text": "He", "range": [ 103, 105 ] } ], "source_mask_link": "selected_person" }, { "id": "entity_squeegee_1", "label": "squeegee", "spans": [ { "text": "the squeegee", "range": [ 43, 55 ] }, { "text": "the squeegee", "range": [ 116, 128 ] } ], "source_mask_link": null }, { "id": "entity_window_1", "label": "window", "spans": [ { "text": "the window", "range": [ 91, 101 ] } ], "source_mask_link": null } ], "actions": [ { "id": "action_1", "mentions": [ { "text": "turns", "ranges": [ [ 11, 16 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "type": "simple", "subject_refs": [ "entity_person_1" ], "object_refs": [], "confidence": "high", "temporal_status": "supported", "intervals": [ { "start_s": 0.0, "end_s": 3.0, "source_start_s": 163.5, "source_end_s": 166.5 } ], "evidence": "Across the clean frames at 0.0–2.0 seconds, the masked person changes from a left-side profile/orientation to a centered, back-facing orientation.", "mask_relation": "subject matches the supplied selected-person mask", "source_subject_refs": "entity_person_1" }, { "id": "action_2", "mentions": [ { "text": "gestures", "ranges": [ [ 21, 29 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "type": "simple", "subject_refs": [ "entity_person_1" ], "object_refs": [], "confidence": "medium", "temporal_status": "supported", "intervals": [ { "start_s": 3.0, "end_s": 4.0, "source_start_s": 166.5, "source_end_s": 167.5 } ], "evidence": "At 3.0 seconds, the masked person raises a hand in front of the window immediately before sustained tool use begins.", "mask_relation": "subject matches the supplied selected-person mask", "source_subject_refs": "entity_person_1" }, { "id": "action_3", "mentions": [ { "text": "using", "ranges": [ [ 37, 42 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "type": "simple", "subject_refs": [ "entity_person_1" ], "object_refs": [ "entity_squeegee_1" ], "confidence": "high", "temporal_status": "supported", "intervals": [ { "start_s": 4.0, "end_s": 24.0, "source_start_s": 167.5, "source_end_s": 187.5 } ], "evidence": "From 4.0 through 23.0 seconds, the masked person visibly manipulates a handheld window-cleaning tool against the window.", "mask_relation": "subject matches the supplied selected-person mask", "source_subject_refs": "entity_person_1" }, { "id": "action_4", "mentions": [ { "text": "clean", "ranges": [ [ 59, 64 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "type": "simple", "subject_refs": [ "entity_person_1" ], "object_refs": [ "entity_window_1" ], "confidence": "high", "temporal_status": "supported", "intervals": [ { "start_s": 4.0, "end_s": 24.0, "source_start_s": 167.5, "source_end_s": 187.5 } ], "evidence": "The clean frames from 4.0 through 23.0 seconds show repeated wiping and scraping motions by the masked person across the window, including its lower area.", "mask_relation": "subject matches the supplied selected-person mask", "source_subject_refs": "entity_person_1" }, { "id": "action_5", "mentions": [ { "text": "puts", "ranges": [ [ 111, 115 ] ], "occurrence": 1, "particle": { "text": "down", "occurrence": 1, "range": [ 129, 133 ] } } ], "type": "phrasal", "subject_refs": [ "entity_person_1" ], "object_refs": [ "entity_squeegee_1" ], "confidence": "medium", "temporal_status": "supported", "intervals": [ { "start_s": 24.0, "end_s": 26.0, "source_start_s": 187.5, "source_end_s": 189.5 } ], "evidence": "At 24.0 seconds the masked person holds the cleaning tool low; by 25.0 seconds it is no longer visible as the person leans toward the area below the frame.", "mask_relation": "subject matches the supplied selected-person mask", "source_subject_refs": "entity_person_1" }, { "id": "action_6", "mentions": [ { "text": "bends over", "ranges": [ [ 135, 145 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "type": "phrasal", "subject_refs": [ "entity_person_1" ], "object_refs": [], "confidence": "high", "temporal_status": "supported", "intervals": [ { "start_s": 25.0, "end_s": 29.0, "source_start_s": 188.5, "source_end_s": 192.5 } ], "evidence": "The masked person's head and torso progressively lower from 25.0 through 28.0 seconds, producing a clearly bent-over posture.", "mask_relation": "subject matches the supplied selected-person mask", "source_subject_refs": "entity_person_1" }, { "id": "action_7", "mentions": [ { "text": "moves", "ranges": [ [ 151, 156 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "type": "simple", "subject_refs": [ "entity_person_1" ], "object_refs": [], "confidence": "high", "temporal_status": "supported_to_clip_boundary", "intervals": [ { "start_s": 29.0, "end_s": 29.460000000000008, "source_start_s": 192.5, "source_end_s": 192.96 } ], "evidence": "At the final sampled frame, 29.0 seconds, the masked person is moving toward and partially beyond the left edge; support continues to the clip boundary at 29.46 seconds.", "mask_relation": "subject matches the supplied selected-person mask", "source_subject_refs": "entity_person_1" } ], "notes": [ "The supplied masklet consistently identifies the person mentioned by both “The person” and “He.”", "No source masks were supplied for the squeegee or window, so their source-mask links are null.", "“the lower right corner” is excluded as a spatial concept, while “the window” is retained as a concrete entity.", "“the frame” denotes the video/image frame and “the left” denotes a spatial region; neither is annotated as an entity.", "All temporal intervals are sparse clip-relative support intervals derived from the sampled clean RGB frames. The original source-video interval is [163.500000, 192.960000)." ] }, "model_output_raw": { "caption": "The person turns and gestures before using the squeegee to clean the lower right corner of the window. He then puts the squeegee down, bends over, and moves out of the frame to the left.", "entity_groups": [ { "id": "entity_person_1", "label": "person", "spans": [ { "text": "The person", "range": [ 0, 10 ] }, { "text": "He", "range": [ 103, 105 ] } ], "source_mask": "selected_person_masklet" }, { "id": "entity_squeegee_1", "label": "squeegee", "spans": [ { "text": "the squeegee", "range": [ 43, 55 ] }, { "text": "the squeegee", "range": [ 116, 128 ] } ], "source_mask": null }, { "id": "entity_window_1", "label": "window", "spans": [ { "text": "the window", "range": [ 91, 101 ] } ], "source_mask": null } ], "actions": [ { "id": "action_1", "mentions": [ { "text": "turns", "occurrence": 1, "ranges": [ [ 11, 16 ] ], "particle": { "text": null, "occurrence": 1 } } ], "type": "simple", "subject": "entity_person_1", "objects": [], "confidence": "high", "temporal_status": "supported", "intervals": [ { "start_s": 0.0, "end_s": 3.0 } ], "mask_relation": "subject matches the supplied selected-person mask", "evidence": "Across the clean frames at 0.0–2.0 seconds, the masked person changes from a left-side profile/orientation to a centered, back-facing orientation." }, { "id": "action_2", "mentions": [ { "text": "gestures", "occurrence": 1, "ranges": [ [ 21, 29 ] ], "particle": { "text": null, "occurrence": 1 } } ], "type": "simple", "subject": "entity_person_1", "objects": [], "confidence": "medium", "temporal_status": "supported", "intervals": [ { "start_s": 3.0, "end_s": 4.0 } ], "mask_relation": "subject matches the supplied selected-person mask", "evidence": "At 3.0 seconds, the masked person raises a hand in front of the window immediately before sustained tool use begins." }, { "id": "action_3", "mentions": [ { "text": "using", "occurrence": 1, "ranges": [ [ 37, 42 ] ], "particle": { "text": null, "occurrence": 1 } } ], "type": "simple", "subject": "entity_person_1", "objects": [ "entity_squeegee_1" ], "confidence": "high", "temporal_status": "supported", "intervals": [ { "start_s": 4.0, "end_s": 24.0 } ], "mask_relation": "subject matches the supplied selected-person mask", "evidence": "From 4.0 through 23.0 seconds, the masked person visibly manipulates a handheld window-cleaning tool against the window." }, { "id": "action_4", "mentions": [ { "text": "clean", "occurrence": 1, "ranges": [ [ 59, 64 ] ], "particle": { "text": null, "occurrence": 1 } } ], "type": "simple", "subject": "entity_person_1", "objects": [ "entity_window_1" ], "confidence": "high", "temporal_status": "supported", "intervals": [ { "start_s": 4.0, "end_s": 24.0 } ], "mask_relation": "subject matches the supplied selected-person mask", "evidence": "The clean frames from 4.0 through 23.0 seconds show repeated wiping and scraping motions by the masked person across the window, including its lower area." }, { "id": "action_5", "mentions": [ { "text": "puts", "occurrence": 1, "ranges": [ [ 111, 115 ] ], "particle": { "text": "down", "occurrence": 1, "range": [ 129, 133 ] } } ], "type": "phrasal", "subject": "entity_person_1", "objects": [ "entity_squeegee_1" ], "confidence": "medium", "temporal_status": "supported", "intervals": [ { "start_s": 24.0, "end_s": 26.0 } ], "mask_relation": "subject matches the supplied selected-person mask", "evidence": "At 24.0 seconds the masked person holds the cleaning tool low; by 25.0 seconds it is no longer visible as the person leans toward the area below the frame." }, { "id": "action_6", "mentions": [ { "text": "bends over", "occurrence": 1, "ranges": [ [ 135, 145 ] ], "particle": { "text": null, "occurrence": 1 } } ], "type": "phrasal", "subject": "entity_person_1", "objects": [], "confidence": "high", "temporal_status": "supported", "intervals": [ { "start_s": 25.0, "end_s": 29.0 } ], "mask_relation": "subject matches the supplied selected-person mask", "evidence": "The masked person's head and torso progressively lower from 25.0 through 28.0 seconds, producing a clearly bent-over posture." }, { "id": "action_7", "mentions": [ { "text": "moves", "occurrence": 1, "ranges": [ [ 151, 156 ] ], "particle": { "text": null, "occurrence": 1 } } ], "type": "simple", "subject": "entity_person_1", "objects": [], "confidence": "high", "temporal_status": "supported_to_clip_boundary", "intervals": [ { "start_s": 29.0, "end_s": 29.46 } ], "mask_relation": "subject matches the supplied selected-person mask", "evidence": "At the final sampled frame, 29.0 seconds, the masked person is moving toward and partially beyond the left edge; support continues to the clip boundary at 29.46 seconds." } ], "notes": [ "The supplied masklet consistently identifies the person mentioned by both “The person” and “He.”", "No source masks were supplied for the squeegee or window, so their source-mask links are null.", "“the lower right corner” is excluded as a spatial concept, while “the window” is retained as a concrete entity.", "“the frame” denotes the video/image frame and “the left” denotes a spatial region; neither is annotated as an entity.", "All temporal intervals are sparse clip-relative support intervals derived from the sampled clean RGB frames. The original source-video interval is [163.500000, 192.960000)." ] }, "normalization_repairs": 0 }, { "source_video_id": "v_9A0BwCxAKAg", "caption_count_category": 14, "clip": { "caption_index": 10, "caption": "A person with dark hair, wearing a black short-sleeved shirt, enters from the left and approaches a window frame. They bend over and use their right hand to work on the bottom portion of the frame. They remain in a bent position, focusing on the lower section.", "source_interval_sec": [ 193.99, 200.49 ], "loom_clip_id": "v_9A0BwCxAKAg_5814_6009", "clip_frame_interval_4fps": [ 5814, 6009 ], "sample_count_1fps": 7, "sampled_frames": [ { "frame": "00000", "timestamp_sec": 193.99, "offset_sec": 0, "rgb_path": "assets/cat_14/v_9A0BwCxAKAg_5814_6009/0000s_00000_rgb.jpg", "mask_path": "assets/cat_14/v_9A0BwCxAKAg_5814_6009/0000s_00000_mask.png", "mask_status": "available" }, { "frame": "00028", "timestamp_sec": 194.99, "offset_sec": 1, "rgb_path": "assets/cat_14/v_9A0BwCxAKAg_5814_6009/0001s_00028_rgb.jpg", "mask_path": "assets/cat_14/v_9A0BwCxAKAg_5814_6009/0001s_00028_mask.png", "mask_status": "available" }, { "frame": "00060", "timestamp_sec": 195.99, "offset_sec": 2, "rgb_path": "assets/cat_14/v_9A0BwCxAKAg_5814_6009/0002s_00060_rgb.jpg", "mask_path": "assets/cat_14/v_9A0BwCxAKAg_5814_6009/0002s_00060_mask.png", "mask_status": "available" }, { "frame": "00088", "timestamp_sec": 196.99, "offset_sec": 3, "rgb_path": "assets/cat_14/v_9A0BwCxAKAg_5814_6009/0003s_00088_rgb.jpg", "mask_path": "assets/cat_14/v_9A0BwCxAKAg_5814_6009/0003s_00088_mask.png", "mask_status": "available" }, { "frame": "00120", "timestamp_sec": 197.99, "offset_sec": 4, "rgb_path": "assets/cat_14/v_9A0BwCxAKAg_5814_6009/0004s_00120_rgb.jpg", "mask_path": "assets/cat_14/v_9A0BwCxAKAg_5814_6009/0004s_00120_mask.png", "mask_status": "available" }, { "frame": "00148", "timestamp_sec": 198.99, "offset_sec": 5, "rgb_path": "assets/cat_14/v_9A0BwCxAKAg_5814_6009/0005s_00148_rgb.jpg", "mask_path": "assets/cat_14/v_9A0BwCxAKAg_5814_6009/0005s_00148_mask.png", "mask_status": "available" }, { "frame": "00180", "timestamp_sec": 199.99, "offset_sec": 6, "rgb_path": "assets/cat_14/v_9A0BwCxAKAg_5814_6009/0006s_00180_rgb.jpg", "mask_path": "assets/cat_14/v_9A0BwCxAKAg_5814_6009/0006s_00180_mask.png", "mask_status": "available" } ], "obj_id": [ 0 ], "anno_id": [ 1273 ], "fps_inferred_from_clip_interval": 30.0 }, "usage": { "input_tokens": 20771, "input_tokens_details": { "cache_write_tokens": 20768, "cached_tokens": 0 }, "output_tokens": 5345, "output_tokens_details": { "reasoning_tokens": 3457 }, "total_tokens": 26116 }, "model": "gpt-5.6-sol", "effort": "medium", "custom_id": "cat14_cap10_v_9A0BwCxAKAg_5814_6009", "annotation": { "caption": "A person with dark hair, wearing a black short-sleeved shirt, enters from the left and approaches a window frame. They bend over and use their right hand to work on the bottom portion of the frame. They remain in a bent position, focusing on the lower section.", "entity_groups": [ { "id": "person_1", "label": "person", "spans": [ { "text": "A person", "range": [ 0, 8 ] }, { "text": "They", "range": [ 114, 118 ] }, { "text": "their", "range": [ 137, 142 ] }, { "text": "They", "range": [ 198, 202 ] } ], "source_mask_link": "selected_person" }, { "id": "hair_1", "label": "hair", "spans": [ { "text": "dark hair", "range": [ 14, 23 ] } ], "source_mask_link": null }, { "id": "shirt_1", "label": "shirt", "spans": [ { "text": "a black short-sleeved shirt", "range": [ 33, 60 ] } ], "source_mask_link": null }, { "id": "window_frame_1", "label": "window frame", "spans": [ { "text": "a window frame", "range": [ 98, 112 ] }, { "text": "the frame", "range": [ 187, 196 ] } ], "source_mask_link": null }, { "id": "hand_1", "label": "hand", "spans": [ { "text": "their right hand", "range": [ 137, 153 ] } ], "source_mask_link": null } ], "actions": [ { "id": "action_1", "mentions": [ { "text": "enters", "ranges": [ [ 62, 68 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "type": "simple", "subject_refs": [ "person_1" ], "object_refs": [], "confidence": "high", "temporal_status": "supported", "intervals": [ { "start_s": 1.0, "end_s": 2.0, "source_start_s": 194.99, "source_end_s": 195.99 } ], "evidence": "The selected person is absent at 0.0 seconds and appears entering from the left at 1.0 seconds.", "mask_relation": "The action subject matches the supplied selected-person mask.", "source_subject_refs": [ "person_1" ] }, { "id": "action_2", "mentions": [ { "text": "approaches", "ranges": [ [ 87, 97 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "type": "simple", "subject_refs": [ "person_1" ], "object_refs": [ "window_frame_1" ], "confidence": "high", "temporal_status": "supported", "intervals": [ { "start_s": 1.0, "end_s": 4.0, "source_start_s": 194.99, "source_end_s": 197.99 } ], "evidence": "From 1.0 through 3.0 seconds, the masked person moves into position directly beside the window frame.", "mask_relation": "The action subject matches the supplied selected-person mask.", "source_subject_refs": [ "person_1" ] }, { "id": "action_3", "mentions": [ { "text": "bend over", "ranges": [ [ 119, 128 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "type": "phrasal", "subject_refs": [ "person_1" ], "object_refs": [], "confidence": "high", "temporal_status": "supported", "intervals": [ { "start_s": 4.0, "end_s": 6.0, "source_start_s": 197.99, "source_end_s": 199.99 } ], "evidence": "At 4.0 and 5.0 seconds, the masked person's torso and head descend into a bent posture.", "mask_relation": "The action subject matches the supplied selected-person mask.", "source_subject_refs": [ "person_1" ] }, { "id": "action_4", "mentions": [ { "text": "use", "ranges": [ [ 133, 136 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "type": "simple", "subject_refs": [ "person_1" ], "object_refs": [ "hand_1" ], "confidence": "medium", "temporal_status": "supported", "intervals": [ { "start_s": 1.0, "end_s": 4.0, "source_start_s": 194.99, "source_end_s": 197.99 } ], "evidence": "At 1.0–3.0 seconds, the masked person extends a hand toward the lower part of the frame; the low resolution limits hand-detail certainty.", "mask_relation": "The action subject matches the supplied selected-person mask.", "source_subject_refs": [ "person_1" ] }, { "id": "action_5", "mentions": [ { "text": "work", "ranges": [ [ 157, 161 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "type": "prepositional", "subject_refs": [ "person_1" ], "object_refs": [ "window_frame_1" ], "confidence": "medium", "temporal_status": "supported", "intervals": [ { "start_s": 1.0, "end_s": 4.0, "source_start_s": 194.99, "source_end_s": 197.99 } ], "evidence": "At 1.0–3.0 seconds, the masked person makes hand contact or manipulation motions near the bottom of the window frame.", "mask_relation": "The action subject matches the supplied selected-person mask.", "source_subject_refs": [ "person_1" ] }, { "id": "action_6", "mentions": [ { "text": "remain", "ranges": [ [ 203, 209 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "type": "posture", "subject_refs": [ "person_1" ], "object_refs": [], "confidence": "high", "temporal_status": "supported", "intervals": [ { "start_s": 5.0, "end_s": 6.5, "source_start_s": 198.99, "source_end_s": 200.49 } ], "evidence": "At 5.0 and 6.0 seconds, the masked person continues holding a bent-over posture through the clip boundary.", "mask_relation": "The action subject matches the supplied selected-person mask.", "source_subject_refs": [ "person_1" ] } ], "notes": [ "The supplied masklet identifies the single person from 1.0 seconds onward; its mask is empty at 0.0 seconds because the person is not yet visible.", "No separate masks were supplied for the hair, shirt, hand, or window frame.", "\"the bottom portion\" and \"the lower section\" denote spatial regions of the frame and were not treated as independent concrete entity instances.", "\"wearing\" describes a clothing state rather than an eligible action, and \"focusing\" describes a mental state; neither was annotated as an action." ] }, "model_output_raw": { "caption": "A person with dark hair, wearing a black short-sleeved shirt, enters from the left and approaches a window frame. They bend over and use their right hand to work on the bottom portion of the frame. They remain in a bent position, focusing on the lower section.", "entity_groups": [ { "id": "person_1", "entity_type": "person", "spans": [ { "text": "A person", "range": [ 0, 8 ] }, { "text": "They", "range": [ 114, 118 ] }, { "text": "their", "range": [ 137, 142 ] }, { "text": "They", "range": [ 198, 202 ] } ], "source_mask": "selected_person_masklet" }, { "id": "hair_1", "entity_type": "hair", "spans": [ { "text": "dark hair", "range": [ 14, 23 ] } ], "source_mask": null }, { "id": "shirt_1", "entity_type": "shirt", "spans": [ { "text": "a black short-sleeved shirt", "range": [ 33, 60 ] } ], "source_mask": null }, { "id": "window_frame_1", "entity_type": "window frame", "spans": [ { "text": "a window frame", "range": [ 98, 112 ] }, { "text": "the frame", "range": [ 187, 196 ] } ], "source_mask": null }, { "id": "hand_1", "entity_type": "hand", "spans": [ { "text": "their right hand", "range": [ 137, 153 ] } ], "source_mask": null } ], "actions": [ { "id": "action_1", "mentions": [ { "text": "enters", "occurrence": 1, "ranges": [ [ 62, 68 ] ], "particle": { "text": null, "occurrence": 1 } } ], "type": "simple", "subject_refs": [ "person_1" ], "object_refs": [], "confidence": "high", "temporal_status": "supported", "intervals": [ { "start_s": 1.0, "end_s": 2.0, "source_start_s": 194.99, "source_end_s": 195.99 } ], "evidence": "The selected person is absent at 0.0 seconds and appears entering from the left at 1.0 seconds.", "mask_relation": "The action subject matches the supplied selected-person mask." }, { "id": "action_2", "mentions": [ { "text": "approaches", "occurrence": 1, "ranges": [ [ 87, 97 ] ], "particle": { "text": null, "occurrence": 1 } } ], "type": "simple", "subject_refs": [ "person_1" ], "object_refs": [ "window_frame_1" ], "confidence": "high", "temporal_status": "supported", "intervals": [ { "start_s": 1.0, "end_s": 4.0, "source_start_s": 194.99, "source_end_s": 197.99 } ], "evidence": "From 1.0 through 3.0 seconds, the masked person moves into position directly beside the window frame.", "mask_relation": "The action subject matches the supplied selected-person mask." }, { "id": "action_3", "mentions": [ { "text": "bend over", "occurrence": 1, "ranges": [ [ 119, 128 ] ], "particle": { "text": null, "occurrence": 1 } } ], "type": "phrasal", "subject_refs": [ "person_1" ], "object_refs": [], "confidence": "high", "temporal_status": "supported", "intervals": [ { "start_s": 4.0, "end_s": 6.0, "source_start_s": 197.99, "source_end_s": 199.99 } ], "evidence": "At 4.0 and 5.0 seconds, the masked person's torso and head descend into a bent posture.", "mask_relation": "The action subject matches the supplied selected-person mask." }, { "id": "action_4", "mentions": [ { "text": "use", "occurrence": 1, "ranges": [ [ 133, 136 ] ], "particle": { "text": null, "occurrence": 1 } } ], "type": "simple", "subject_refs": [ "person_1" ], "object_refs": [ "hand_1" ], "confidence": "medium", "temporal_status": "supported", "intervals": [ { "start_s": 1.0, "end_s": 4.0, "source_start_s": 194.99, "source_end_s": 197.99 } ], "evidence": "At 1.0–3.0 seconds, the masked person extends a hand toward the lower part of the frame; the low resolution limits hand-detail certainty.", "mask_relation": "The action subject matches the supplied selected-person mask." }, { "id": "action_5", "mentions": [ { "text": "work", "occurrence": 1, "ranges": [ [ 157, 161 ] ], "particle": { "text": null, "occurrence": 1 } } ], "type": "prepositional", "subject_refs": [ "person_1" ], "object_refs": [ "window_frame_1" ], "confidence": "medium", "temporal_status": "supported", "intervals": [ { "start_s": 1.0, "end_s": 4.0, "source_start_s": 194.99, "source_end_s": 197.99 } ], "evidence": "At 1.0–3.0 seconds, the masked person makes hand contact or manipulation motions near the bottom of the window frame.", "mask_relation": "The action subject matches the supplied selected-person mask." }, { "id": "action_6", "mentions": [ { "text": "remain", "occurrence": 1, "ranges": [ [ 203, 209 ] ], "particle": { "text": null, "occurrence": 1 } } ], "type": "posture", "subject_refs": [ "person_1" ], "object_refs": [], "confidence": "high", "temporal_status": "supported", "intervals": [ { "start_s": 5.0, "end_s": 6.5, "source_start_s": 198.99, "source_end_s": 200.49 } ], "evidence": "At 5.0 and 6.0 seconds, the masked person continues holding a bent-over posture through the clip boundary.", "mask_relation": "The action subject matches the supplied selected-person mask." } ], "notes": [ "The supplied masklet identifies the single person from 1.0 seconds onward; its mask is empty at 0.0 seconds because the person is not yet visible.", "No separate masks were supplied for the hair, shirt, hand, or window frame.", "\"the bottom portion\" and \"the lower section\" denote spatial regions of the frame and were not treated as independent concrete entity instances.", "\"wearing\" describes a clothing state rather than an eligible action, and \"focusing\" describes a mental state; neither was annotated as an action." ] }, "normalization_repairs": 0 }, { "source_video_id": "v_9A0BwCxAKAg", "caption_count_category": 14, "clip": { "caption_index": 11, "caption": "The person straightens up and picks up a grey cloth with both hands. They unfold the cloth and examine it briefly. They then raise their right hand, holding the cloth, and begin wiping the upper right corner of the window frame.", "source_interval_sec": [ 200.49, 207.99 ], "loom_clip_id": "v_9A0BwCxAKAg_6009_6233", "clip_frame_interval_4fps": [ 6009, 6233 ], "sample_count_1fps": 8, "sampled_frames": [ { "frame": "00000", "timestamp_sec": 200.49, "offset_sec": 0, "rgb_path": "assets/cat_14/v_9A0BwCxAKAg_6009_6233/0000s_00000_rgb.jpg", "mask_path": "assets/cat_14/v_9A0BwCxAKAg_6009_6233/0000s_00000_mask.png", "mask_status": "available" }, { "frame": "00028", "timestamp_sec": 201.49, "offset_sec": 1, "rgb_path": "assets/cat_14/v_9A0BwCxAKAg_6009_6233/0001s_00028_rgb.jpg", "mask_path": "assets/cat_14/v_9A0BwCxAKAg_6009_6233/0001s_00028_mask.png", "mask_status": "available" }, { "frame": "00060", "timestamp_sec": 202.49, "offset_sec": 2, "rgb_path": "assets/cat_14/v_9A0BwCxAKAg_6009_6233/0002s_00060_rgb.jpg", "mask_path": "assets/cat_14/v_9A0BwCxAKAg_6009_6233/0002s_00060_mask.png", "mask_status": "available" }, { "frame": "00088", "timestamp_sec": 203.49, "offset_sec": 3, "rgb_path": "assets/cat_14/v_9A0BwCxAKAg_6009_6233/0003s_00088_rgb.jpg", "mask_path": "assets/cat_14/v_9A0BwCxAKAg_6009_6233/0003s_00088_mask.png", "mask_status": "available" }, { "frame": "00120", "timestamp_sec": 204.49, "offset_sec": 4, "rgb_path": "assets/cat_14/v_9A0BwCxAKAg_6009_6233/0004s_00120_rgb.jpg", "mask_path": "assets/cat_14/v_9A0BwCxAKAg_6009_6233/0004s_00120_mask.png", "mask_status": "available" }, { "frame": "00148", "timestamp_sec": 205.49, "offset_sec": 5, "rgb_path": "assets/cat_14/v_9A0BwCxAKAg_6009_6233/0005s_00148_rgb.jpg", "mask_path": "assets/cat_14/v_9A0BwCxAKAg_6009_6233/0005s_00148_mask.png", "mask_status": "available" }, { "frame": "00180", "timestamp_sec": 206.49, "offset_sec": 6, "rgb_path": "assets/cat_14/v_9A0BwCxAKAg_6009_6233/0006s_00180_rgb.jpg", "mask_path": "assets/cat_14/v_9A0BwCxAKAg_6009_6233/0006s_00180_mask.png", "mask_status": "available" }, { "frame": "00208", "timestamp_sec": 207.49, "offset_sec": 7, "rgb_path": "assets/cat_14/v_9A0BwCxAKAg_6009_6233/0007s_00208_rgb.jpg", "mask_path": "assets/cat_14/v_9A0BwCxAKAg_6009_6233/0007s_00208_mask.png", "mask_status": "available" } ], "obj_id": [ 0 ], "anno_id": [ 1274 ], "fps_inferred_from_clip_interval": 29.8667 }, "usage": { "input_tokens": 21537, "input_tokens_details": { "cache_write_tokens": 21534, "cached_tokens": 0 }, "output_tokens": 5140, "output_tokens_details": { "reasoning_tokens": 2975 }, "total_tokens": 26677 }, "model": "gpt-5.6-sol", "effort": "medium", "custom_id": "cat14_cap11_v_9A0BwCxAKAg_6009_6233", "annotation": { "caption": "The person straightens up and picks up a grey cloth with both hands. They unfold the cloth and examine it briefly. They then raise their right hand, holding the cloth, and begin wiping the upper right corner of the window frame.", "entity_groups": [ { "id": "person_1", "label": "person", "spans": [ { "text": "The person", "range": [ 0, 10 ] }, { "text": "They", "range": [ 69, 73 ] }, { "text": "They", "range": [ 115, 119 ] }, { "text": "their", "range": [ 131, 136 ] } ], "source_mask_link": "selected_person" }, { "id": "cloth_1", "label": "cloth", "spans": [ { "text": "a grey cloth", "range": [ 39, 51 ] }, { "text": "the cloth", "range": [ 81, 90 ] }, { "text": "it", "range": [ 103, 105 ] }, { "text": "the cloth", "range": [ 157, 166 ] } ], "source_mask_link": null }, { "id": "right_hand_1", "label": "hand", "spans": [ { "text": "both hands", "range": [ 57, 67 ] }, { "text": "their right hand", "range": [ 131, 147 ] } ], "source_mask_link": null }, { "id": "other_hand_1", "label": "hand", "spans": [ { "text": "both hands", "range": [ 57, 67 ] } ], "source_mask_link": null }, { "id": "window_frame_1", "label": "window_frame", "spans": [ { "text": "the window frame", "range": [ 211, 227 ] } ], "source_mask_link": null } ], "actions": [ { "id": "action_1", "mentions": [ { "text": "straightens up", "ranges": [ [ 11, 25 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "type": "phrasal", "subject_refs": [ "person_1" ], "object_refs": [], "confidence": "high", "temporal_status": "localized_sparse_support", "intervals": [ { "start_s": 0.0, "end_s": 1.0, "source_start_s": 200.49, "source_end_s": 201.49 } ], "evidence": "The masked person changes from a bent-forward posture at 0.0 seconds to a more upright posture by 1.0 seconds.", "mask_relation": "The action subject matches the supplied selected-person mask.", "source_subject_refs": [ "person_1" ] }, { "id": "action_2", "mentions": [ { "text": "picks up", "ranges": [ [ 30, 38 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "type": "phrasal", "subject_refs": [ "person_1" ], "object_refs": [ "cloth_1" ], "confidence": "high", "temporal_status": "localized_sparse_support", "intervals": [ { "start_s": 1.0, "end_s": 2.0, "source_start_s": 201.49, "source_end_s": 202.49 } ], "evidence": "The masked person reaches at 1.0 seconds and is visibly holding the cloth by 2.0 seconds.", "mask_relation": "The action subject matches the supplied selected-person mask.", "source_subject_refs": [ "person_1" ] }, { "id": "action_3", "mentions": [ { "text": "unfold", "ranges": [ [ 74, 80 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "type": "simple", "subject_refs": [ "person_1" ], "object_refs": [ "cloth_1" ], "confidence": "high", "temporal_status": "localized_sparse_support", "intervals": [ { "start_s": 2.0, "end_s": 5.0, "source_start_s": 202.49, "source_end_s": 205.49 } ], "evidence": "From 2.0 through 4.0 seconds, the masked person manipulates and spreads the cloth between both hands.", "mask_relation": "The action subject matches the supplied selected-person mask.", "source_subject_refs": [ "person_1" ] }, { "id": "action_4", "mentions": [ { "text": "examine", "ranges": [ [ 95, 102 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "type": "simple", "subject_refs": [ "person_1" ], "object_refs": [ "cloth_1" ], "confidence": "high", "temporal_status": "localized_sparse_support", "intervals": [ { "start_s": 4.0, "end_s": 6.0, "source_start_s": 204.49, "source_end_s": 206.49 } ], "evidence": "At 4.0 and 5.0 seconds, the masked person holds the cloth in front of the body and looks toward it.", "mask_relation": "The action subject matches the supplied selected-person mask.", "source_subject_refs": [ "person_1" ] }, { "id": "action_5", "mentions": [ { "text": "raise", "ranges": [ [ 125, 130 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "type": "simple", "subject_refs": [ "person_1" ], "object_refs": [ "right_hand_1" ], "confidence": "high", "temporal_status": "localized_sparse_support", "intervals": [ { "start_s": 5.0, "end_s": 7.0, "source_start_s": 205.49, "source_end_s": 207.49 } ], "evidence": "The masked person's right arm moves from a low position at 5.0 seconds to above shoulder level at 6.0 seconds.", "mask_relation": "The action subject matches the supplied selected-person mask.", "source_subject_refs": [ "person_1" ] }, { "id": "action_6", "mentions": [ { "text": "holding", "ranges": [ [ 149, 156 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "type": "simple", "subject_refs": [ "person_1" ], "object_refs": [ "cloth_1" ], "confidence": "high", "temporal_status": "localized_sparse_support", "intervals": [ { "start_s": 2.0, "end_s": 7.5, "source_start_s": 202.49, "source_end_s": 207.99 } ], "evidence": "The cloth remains visibly held by the masked person from 2.0 seconds through the final sampled frame.", "mask_relation": "The action subject matches the supplied selected-person mask.", "source_subject_refs": [ "person_1" ] }, { "id": "action_7", "mentions": [ { "text": "wiping", "ranges": [ [ 178, 184 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "type": "simple", "subject_refs": [ "person_1" ], "object_refs": [ "window_frame_1" ], "confidence": "high", "temporal_status": "localized_sparse_support", "intervals": [ { "start_s": 6.0, "end_s": 7.5, "source_start_s": 206.49, "source_end_s": 207.99 } ], "evidence": "At 6.0 seconds the masked person places the cloth at the upper-right frame area, and at 7.0 seconds the hand and cloth have moved downward along it.", "mask_relation": "The action subject matches the supplied selected-person mask.", "source_subject_refs": [ "person_1" ] } ], "notes": [ "The supplied masklet consistently identifies the captioned person across all sampled frames.", "No separate source masks were supplied for the cloth, hands, or window frame; their source-mask links are therefore null.", "The two hands are split into instance groups because the later phrase specifically identifies the right hand; the shared phrase \"both hands\" anchors both groups.", "The spatial phrase \"the upper right corner\" is not an entity group because corners and spatial regions are excluded by the text-annotation rules.", "The aspectual verb \"begin\" is excluded from the wiping action span." ] }, "model_output_raw": { "caption": "The person straightens up and picks up a grey cloth with both hands. They unfold the cloth and examine it briefly. They then raise their right hand, holding the cloth, and begin wiping the upper right corner of the window frame.", "entity_groups": [ { "id": "person_1", "entity_type": "person", "spans": [ { "text": "The person", "range": [ 0, 10 ] }, { "text": "They", "range": [ 69, 73 ] }, { "text": "They", "range": [ 115, 119 ] }, { "text": "their", "range": [ 131, 136 ] } ], "source_mask": "selected_person_masklet" }, { "id": "cloth_1", "entity_type": "cloth", "spans": [ { "text": "a grey cloth", "range": [ 39, 51 ] }, { "text": "the cloth", "range": [ 81, 90 ] }, { "text": "it", "range": [ 103, 105 ] }, { "text": "the cloth", "range": [ 157, 166 ] } ], "source_mask": null }, { "id": "right_hand_1", "entity_type": "hand", "spans": [ { "text": "both hands", "range": [ 57, 67 ] }, { "text": "their right hand", "range": [ 131, 147 ] } ], "source_mask": null }, { "id": "other_hand_1", "entity_type": "hand", "spans": [ { "text": "both hands", "range": [ 57, 67 ] } ], "source_mask": null }, { "id": "window_frame_1", "entity_type": "window_frame", "spans": [ { "text": "the window frame", "range": [ 211, 227 ] } ], "source_mask": null } ], "actions": [ { "id": "action_1", "mentions": [ { "text": "straightens up", "occurrence": 1, "ranges": [ [ 11, 25 ] ], "particle": { "text": null, "occurrence": 1 } } ], "type": "phrasal", "subject_refs": [ "person_1" ], "object_refs": [], "confidence": "high", "temporal_status": "localized_sparse_support", "intervals": [ { "clip_relative_s": [ 0.0, 1.0 ], "source_video_s": [ 200.49, 201.49 ] } ], "evidence": "The masked person changes from a bent-forward posture at 0.0 seconds to a more upright posture by 1.0 seconds.", "mask_relation": "The action subject matches the supplied selected-person mask." }, { "id": "action_2", "mentions": [ { "text": "picks up", "occurrence": 1, "ranges": [ [ 30, 38 ] ], "particle": { "text": null, "occurrence": 1 } } ], "type": "phrasal", "subject_refs": [ "person_1" ], "object_refs": [ "cloth_1" ], "instrument_refs": [ "right_hand_1", "other_hand_1" ], "confidence": "high", "temporal_status": "localized_sparse_support", "intervals": [ { "clip_relative_s": [ 1.0, 2.0 ], "source_video_s": [ 201.49, 202.49 ] } ], "evidence": "The masked person reaches at 1.0 seconds and is visibly holding the cloth by 2.0 seconds.", "mask_relation": "The action subject matches the supplied selected-person mask." }, { "id": "action_3", "mentions": [ { "text": "unfold", "occurrence": 1, "ranges": [ [ 74, 80 ] ], "particle": { "text": null, "occurrence": 1 } } ], "type": "simple", "subject_refs": [ "person_1" ], "object_refs": [ "cloth_1" ], "confidence": "high", "temporal_status": "localized_sparse_support", "intervals": [ { "clip_relative_s": [ 2.0, 5.0 ], "source_video_s": [ 202.49, 205.49 ] } ], "evidence": "From 2.0 through 4.0 seconds, the masked person manipulates and spreads the cloth between both hands.", "mask_relation": "The action subject matches the supplied selected-person mask." }, { "id": "action_4", "mentions": [ { "text": "examine", "occurrence": 1, "ranges": [ [ 95, 102 ] ], "particle": { "text": null, "occurrence": 1 } } ], "type": "simple", "subject_refs": [ "person_1" ], "object_refs": [ "cloth_1" ], "confidence": "high", "temporal_status": "localized_sparse_support", "intervals": [ { "clip_relative_s": [ 4.0, 6.0 ], "source_video_s": [ 204.49, 206.49 ] } ], "evidence": "At 4.0 and 5.0 seconds, the masked person holds the cloth in front of the body and looks toward it.", "mask_relation": "The action subject matches the supplied selected-person mask." }, { "id": "action_5", "mentions": [ { "text": "raise", "occurrence": 1, "ranges": [ [ 125, 130 ] ], "particle": { "text": null, "occurrence": 1 } } ], "type": "simple", "subject_refs": [ "person_1" ], "object_refs": [ "right_hand_1" ], "confidence": "high", "temporal_status": "localized_sparse_support", "intervals": [ { "clip_relative_s": [ 5.0, 7.0 ], "source_video_s": [ 205.49, 207.49 ] } ], "evidence": "The masked person's right arm moves from a low position at 5.0 seconds to above shoulder level at 6.0 seconds.", "mask_relation": "The action subject matches the supplied selected-person mask." }, { "id": "action_6", "mentions": [ { "text": "holding", "occurrence": 1, "ranges": [ [ 149, 156 ] ], "particle": { "text": null, "occurrence": 1 } } ], "type": "simple", "subject_refs": [ "person_1" ], "object_refs": [ "cloth_1" ], "confidence": "high", "temporal_status": "localized_sparse_support", "intervals": [ { "clip_relative_s": [ 2.0, 7.5 ], "source_video_s": [ 202.49, 207.99 ] } ], "evidence": "The cloth remains visibly held by the masked person from 2.0 seconds through the final sampled frame.", "mask_relation": "The action subject matches the supplied selected-person mask." }, { "id": "action_7", "mentions": [ { "text": "wiping", "occurrence": 1, "ranges": [ [ 178, 184 ] ], "particle": { "text": null, "occurrence": 1 } } ], "type": "simple", "subject_refs": [ "person_1" ], "object_refs": [ "window_frame_1" ], "confidence": "high", "temporal_status": "localized_sparse_support", "intervals": [ { "clip_relative_s": [ 6.0, 7.5 ], "source_video_s": [ 206.49, 207.99 ] } ], "evidence": "At 6.0 seconds the masked person places the cloth at the upper-right frame area, and at 7.0 seconds the hand and cloth have moved downward along it.", "mask_relation": "The action subject matches the supplied selected-person mask." } ], "notes": [ "The supplied masklet consistently identifies the captioned person across all sampled frames.", "No separate source masks were supplied for the cloth, hands, or window frame; their source-mask links are therefore null.", "The two hands are split into instance groups because the later phrase specifically identifies the right hand; the shared phrase \"both hands\" anchors both groups.", "The spatial phrase \"the upper right corner\" is not an entity group because corners and spatial regions are excluded by the text-annotation rules.", "The aspectual verb \"begin\" is excluded from the wiping action span." ] }, "normalization_repairs": 0 }, { "source_video_id": "v_9A0BwCxAKAg", "caption_count_category": 14, "clip": { "caption_index": 12, "caption": "The person uses the cloth to wipe the top part of the window frame, moving from right to left. They hold the cloth with both hands for part of the motion. They then focus on wiping the top left corner of the frame.", "source_interval_sec": [ 207.99, 214.49 ], "loom_clip_id": "v_9A0BwCxAKAg_6233_6428", "clip_frame_interval_4fps": [ 6233, 6428 ], "sample_count_1fps": 7, "sampled_frames": [ { "frame": "00000", "timestamp_sec": 207.99, "offset_sec": 0, "rgb_path": "assets/cat_14/v_9A0BwCxAKAg_6233_6428/0000s_00000_rgb.jpg", "mask_path": "assets/cat_14/v_9A0BwCxAKAg_6233_6428/0000s_00000_mask.png", "mask_status": "available" }, { "frame": "00028", "timestamp_sec": 208.99, "offset_sec": 1, "rgb_path": "assets/cat_14/v_9A0BwCxAKAg_6233_6428/0001s_00028_rgb.jpg", "mask_path": "assets/cat_14/v_9A0BwCxAKAg_6233_6428/0001s_00028_mask.png", "mask_status": "available" }, { "frame": "00060", "timestamp_sec": 209.99, "offset_sec": 2, "rgb_path": "assets/cat_14/v_9A0BwCxAKAg_6233_6428/0002s_00060_rgb.jpg", "mask_path": "assets/cat_14/v_9A0BwCxAKAg_6233_6428/0002s_00060_mask.png", "mask_status": "available" }, { "frame": "00088", "timestamp_sec": 210.99, "offset_sec": 3, "rgb_path": "assets/cat_14/v_9A0BwCxAKAg_6233_6428/0003s_00088_rgb.jpg", "mask_path": "assets/cat_14/v_9A0BwCxAKAg_6233_6428/0003s_00088_mask.png", "mask_status": "available" }, { "frame": "00120", "timestamp_sec": 211.99, "offset_sec": 4, "rgb_path": "assets/cat_14/v_9A0BwCxAKAg_6233_6428/0004s_00120_rgb.jpg", "mask_path": "assets/cat_14/v_9A0BwCxAKAg_6233_6428/0004s_00120_mask.png", "mask_status": "available" }, { "frame": "00148", "timestamp_sec": 212.99, "offset_sec": 5, "rgb_path": "assets/cat_14/v_9A0BwCxAKAg_6233_6428/0005s_00148_rgb.jpg", "mask_path": "assets/cat_14/v_9A0BwCxAKAg_6233_6428/0005s_00148_mask.png", "mask_status": "available" }, { "frame": "00180", "timestamp_sec": 213.99, "offset_sec": 6, "rgb_path": "assets/cat_14/v_9A0BwCxAKAg_6233_6428/0006s_00180_rgb.jpg", "mask_path": "assets/cat_14/v_9A0BwCxAKAg_6233_6428/0006s_00180_mask.png", "mask_status": "available" } ], "obj_id": [ 0 ], "anno_id": [ 1275 ], "fps_inferred_from_clip_interval": 30.0 }, "usage": { "input_tokens": 20762, "input_tokens_details": { "cache_write_tokens": 20759, "cached_tokens": 0 }, "output_tokens": 4474, "output_tokens_details": { "reasoning_tokens": 2998 }, "total_tokens": 25236 }, "model": "gpt-5.6-sol", "effort": "medium", "custom_id": "cat14_cap12_v_9A0BwCxAKAg_6233_6428", "annotation": { "caption": "The person uses the cloth to wipe the top part of the window frame, moving from right to left. They hold the cloth with both hands for part of the motion. They then focus on wiping the top left corner of the frame.", "entity_groups": [ { "id": "person_1", "label": "person", "spans": [ { "text": "The person", "range": [ 0, 10 ] }, { "text": "They", "range": [ 95, 99 ] }, { "text": "They", "range": [ 155, 159 ] } ], "source_mask_link": "selected_person" }, { "id": "cloth_1", "label": "cloth", "spans": [ { "text": "the cloth", "range": [ 16, 25 ] }, { "text": "the cloth", "range": [ 105, 114 ] } ], "source_mask_link": null }, { "id": "window_frame_1", "label": "window frame", "spans": [ { "text": "the window frame", "range": [ 50, 66 ] }, { "text": "the frame", "range": [ 204, 213 ] } ], "source_mask_link": null }, { "id": "hands_1", "label": "hands", "spans": [ { "text": "both hands", "range": [ 120, 130 ] } ], "source_mask_link": null } ], "actions": [ { "id": "action_1", "mentions": [ { "text": "uses", "ranges": [ [ 11, 15 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "type": "simple", "subject_refs": [ "person_1" ], "object_refs": [ "cloth_1" ], "confidence": "high", "temporal_status": "localized", "intervals": [ { "start_s": 2.0, "end_s": 6.0, "source_start_s": 209.99, "source_end_s": 213.99 } ], "evidence": "At samples 2.000–5.000 seconds, the masked person visibly manipulates the cloth against the window area.", "mask_relation": "The action subject matches the supplied selected-person mask.", "source_subject_refs": "person_1" }, { "id": "action_2", "mentions": [ { "text": "wipe", "ranges": [ [ 29, 33 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } }, { "text": "wiping", "ranges": [ [ 174, 180 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "type": "simple", "subject_refs": [ "person_1" ], "object_refs": [ "window_frame_1" ], "confidence": "high", "temporal_status": "localized", "intervals": [ { "start_s": 2.0, "end_s": 6.0, "source_start_s": 209.99, "source_end_s": 213.99 } ], "evidence": "Samples 2.000–5.000 seconds show the masked person pressing and moving the cloth across the upper portion of the window frame; the later left-side wiping is a continuation of the same wiping event.", "mask_relation": "The action subject matches the supplied selected-person mask.", "source_subject_refs": "person_1" }, { "id": "action_3", "mentions": [ { "text": "moving", "ranges": [ [ 68, 74 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "type": "simple", "subject_refs": [ "person_1" ], "object_refs": [], "confidence": "medium", "temporal_status": "localized", "intervals": [ { "start_s": 3.0, "end_s": 6.0, "source_start_s": 210.99, "source_end_s": 213.99 } ], "evidence": "Across samples 3.000–5.000 seconds, the masked person's raised hands and cloth progress from the right side toward the left along the upper frame.", "mask_relation": "The action subject matches the supplied selected-person mask.", "source_subject_refs": "person_1" }, { "id": "action_4", "mentions": [ { "text": "hold", "ranges": [ [ 100, 104 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "type": "simple", "subject_refs": [ "person_1" ], "object_refs": [ "cloth_1" ], "confidence": "high", "temporal_status": "localized", "intervals": [ { "start_s": 2.0, "end_s": 6.5, "source_start_s": 209.99, "source_end_s": 214.49 } ], "evidence": "Samples 2.000–6.000 seconds show both hands of the masked person grasping the cloth, including the final sampled frame.", "mask_relation": "The action subject matches the supplied selected-person mask.", "source_subject_refs": "person_1" } ], "notes": [ "The supplied masklet consistently identifies the person described by “The person” and both occurrences of “They.”", "No separate masks were supplied for the cloth, window frame, or hands, so those entity groups have null source-mask links.", "The spatial expressions “the top part,” “right,” “left,” and “the top left corner” were not treated as separate concrete entities.", "“focus” was excluded because it describes attention rather than a visually meaningful physical action." ] }, "model_output_raw": { "caption": "The person uses the cloth to wipe the top part of the window frame, moving from right to left. They hold the cloth with both hands for part of the motion. They then focus on wiping the top left corner of the frame.", "entity_groups": [ { "id": "person_1", "entity_type": "person", "mentions": [ { "text": "The person", "range": [ 0, 10 ] }, { "text": "They", "range": [ 95, 99 ] }, { "text": "They", "range": [ 155, 159 ] } ], "source_mask": "selected_person_masklet" }, { "id": "cloth_1", "entity_type": "cloth", "mentions": [ { "text": "the cloth", "range": [ 16, 25 ] }, { "text": "the cloth", "range": [ 105, 114 ] } ], "source_mask": null }, { "id": "window_frame_1", "entity_type": "window frame", "mentions": [ { "text": "the window frame", "range": [ 50, 66 ] }, { "text": "the frame", "range": [ 204, 213 ] } ], "source_mask": null }, { "id": "hands_1", "entity_type": "hands", "mentions": [ { "text": "both hands", "range": [ 120, 130 ] } ], "source_mask": null } ], "actions": [ { "id": "action_1", "mentions": [ { "text": "uses", "ranges": [ [ 11, 15 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "type": "simple", "subject": "person_1", "objects": [ "cloth_1" ], "confidence": "high", "mask_relation": "The action subject matches the supplied selected-person mask.", "temporal_status": "localized", "intervals": [ { "start_s": 2.0, "end_s": 6.0, "source_start_s": 209.99, "source_end_s": 213.99 } ], "evidence": "At samples 2.000–5.000 seconds, the masked person visibly manipulates the cloth against the window area." }, { "id": "action_2", "mentions": [ { "text": "wipe", "ranges": [ [ 29, 33 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } }, { "text": "wiping", "ranges": [ [ 174, 180 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "type": "simple", "subject": "person_1", "objects": [ "window_frame_1" ], "instruments": [ "cloth_1" ], "confidence": "high", "mask_relation": "The action subject matches the supplied selected-person mask.", "temporal_status": "localized", "intervals": [ { "start_s": 2.0, "end_s": 6.0, "source_start_s": 209.99, "source_end_s": 213.99 } ], "evidence": "Samples 2.000–5.000 seconds show the masked person pressing and moving the cloth across the upper portion of the window frame; the later left-side wiping is a continuation of the same wiping event." }, { "id": "action_3", "mentions": [ { "text": "moving", "ranges": [ [ 68, 74 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "type": "simple", "subject": "person_1", "objects": [], "confidence": "medium", "mask_relation": "The action subject matches the supplied selected-person mask.", "temporal_status": "localized", "intervals": [ { "start_s": 3.0, "end_s": 6.0, "source_start_s": 210.99, "source_end_s": 213.99 } ], "evidence": "Across samples 3.000–5.000 seconds, the masked person's raised hands and cloth progress from the right side toward the left along the upper frame." }, { "id": "action_4", "mentions": [ { "text": "hold", "ranges": [ [ 100, 104 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "type": "simple", "subject": "person_1", "objects": [ "cloth_1" ], "confidence": "high", "mask_relation": "The action subject matches the supplied selected-person mask.", "temporal_status": "localized", "intervals": [ { "start_s": 2.0, "end_s": 6.5, "source_start_s": 209.99, "source_end_s": 214.49 } ], "evidence": "Samples 2.000–6.000 seconds show both hands of the masked person grasping the cloth, including the final sampled frame." } ], "notes": [ "The supplied masklet consistently identifies the person described by “The person” and both occurrences of “They.”", "No separate masks were supplied for the cloth, window frame, or hands, so those entity groups have null source-mask links.", "The spatial expressions “the top part,” “right,” “left,” and “the top left corner” were not treated as separate concrete entities.", "“focus” was excluded because it describes attention rather than a visually meaningful physical action." ] }, "normalization_repairs": 0 }, { "source_video_id": "v_9A0BwCxAKAg", "caption_count_category": 14, "clip": { "caption_index": 13, "caption": "After finishing with the top, the person lowers their arms and bends down. They use the cloth in their right hand to wipe the bottom portion of the window frame. Their posture remains bent over as they clean the lower section.", "source_interval_sec": [ 214.49, 221.99 ], "loom_clip_id": "v_9A0BwCxAKAg_6428_6653", "clip_frame_interval_4fps": [ 6428, 6653 ], "sample_count_1fps": 8, "sampled_frames": [ { "frame": "00000", "timestamp_sec": 214.49, "offset_sec": 0, "rgb_path": "assets/cat_14/v_9A0BwCxAKAg_6428_6653/0000s_00000_rgb.jpg", "mask_path": "assets/cat_14/v_9A0BwCxAKAg_6428_6653/0000s_00000_mask.png", "mask_status": "available" }, { "frame": "00028", "timestamp_sec": 215.49, "offset_sec": 1, "rgb_path": "assets/cat_14/v_9A0BwCxAKAg_6428_6653/0001s_00028_rgb.jpg", "mask_path": "assets/cat_14/v_9A0BwCxAKAg_6428_6653/0001s_00028_mask.png", "mask_status": "available" }, { "frame": "00060", "timestamp_sec": 216.49, "offset_sec": 2, "rgb_path": "assets/cat_14/v_9A0BwCxAKAg_6428_6653/0002s_00060_rgb.jpg", "mask_path": "assets/cat_14/v_9A0BwCxAKAg_6428_6653/0002s_00060_mask.png", "mask_status": "available" }, { "frame": "00088", "timestamp_sec": 217.49, "offset_sec": 3, "rgb_path": "assets/cat_14/v_9A0BwCxAKAg_6428_6653/0003s_00088_rgb.jpg", "mask_path": "assets/cat_14/v_9A0BwCxAKAg_6428_6653/0003s_00088_mask.png", "mask_status": "available" }, { "frame": "00120", "timestamp_sec": 218.49, "offset_sec": 4, "rgb_path": "assets/cat_14/v_9A0BwCxAKAg_6428_6653/0004s_00120_rgb.jpg", "mask_path": "assets/cat_14/v_9A0BwCxAKAg_6428_6653/0004s_00120_mask.png", "mask_status": "available" }, { "frame": "00148", "timestamp_sec": 219.49, "offset_sec": 5, "rgb_path": "assets/cat_14/v_9A0BwCxAKAg_6428_6653/0005s_00148_rgb.jpg", "mask_path": "assets/cat_14/v_9A0BwCxAKAg_6428_6653/0005s_00148_mask.png", "mask_status": "available" }, { "frame": "00180", "timestamp_sec": 220.49, "offset_sec": 6, "rgb_path": "assets/cat_14/v_9A0BwCxAKAg_6428_6653/0006s_00180_rgb.jpg", "mask_path": "assets/cat_14/v_9A0BwCxAKAg_6428_6653/0006s_00180_mask.png", "mask_status": "available" }, { "frame": "00208", "timestamp_sec": 221.49, "offset_sec": 7, "rgb_path": "assets/cat_14/v_9A0BwCxAKAg_6428_6653/0007s_00208_rgb.jpg", "mask_path": "assets/cat_14/v_9A0BwCxAKAg_6428_6653/0007s_00208_mask.png", "mask_status": "available" } ], "obj_id": [ 0 ], "anno_id": [ 1276 ], "fps_inferred_from_clip_interval": 30.0 }, "usage": { "input_tokens": 21535, "input_tokens_details": { "cache_write_tokens": 21532, "cached_tokens": 0 }, "output_tokens": 5348, "output_tokens_details": { "reasoning_tokens": 3774 }, "total_tokens": 26883 }, "model": "gpt-5.6-sol", "effort": "medium", "custom_id": "cat14_cap13_v_9A0BwCxAKAg_6428_6653", "annotation": { "caption": "After finishing with the top, the person lowers their arms and bends down. They use the cloth in their right hand to wipe the bottom portion of the window frame. Their posture remains bent over as they clean the lower section.", "entity_groups": [ { "id": "person", "label": "person", "spans": [ { "text": "the person", "range": [ 30, 40 ] }, { "text": "their", "range": [ 48, 53 ] }, { "text": "They", "range": [ 75, 79 ] }, { "text": "their", "range": [ 97, 102 ] }, { "text": "Their", "range": [ 162, 167 ] }, { "text": "they", "range": [ 197, 201 ] } ], "source_mask_link": "selected_person" }, { "id": "arms", "label": "arms", "spans": [ { "text": "their arms", "range": [ 48, 58 ] } ], "source_mask_link": null }, { "id": "cloth", "label": "cloth", "spans": [ { "text": "the cloth", "range": [ 84, 93 ] } ], "source_mask_link": null }, { "id": "right_hand", "label": "right hand", "spans": [ { "text": "their right hand", "range": [ 97, 113 ] } ], "source_mask_link": null }, { "id": "lower_frame_section", "label": "lower section of window frame", "spans": [ { "text": "the bottom portion", "range": [ 122, 140 ] }, { "text": "the lower section", "range": [ 208, 225 ] } ], "source_mask_link": null }, { "id": "window_frame", "label": "window frame", "spans": [ { "text": "the window frame", "range": [ 144, 160 ] } ], "source_mask_link": null } ], "actions": [ { "id": "lower_arms", "mentions": [ { "text": "lowers", "ranges": [ [ 41, 47 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "type": "simple", "subject_refs": [ "person" ], "object_refs": [ "arms" ], "confidence": "high", "temporal_status": "localized", "intervals": [ { "start_s": 0.0, "end_s": 2.0, "source_start_s": 214.49, "source_end_s": 216.49 } ], "evidence": "At 0.000 s the selected person has both arms raised near the upper frame; by 1.000 s the arms have moved downward.", "mask_relation": "The action subject matches the supplied selected-person mask.", "source_subject_refs": [ "person" ] }, { "id": "bend_down_and_remain_bent", "mentions": [ { "text": "bends down", "ranges": [ [ 63, 73 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } }, { "text": "bent over", "ranges": [ [ 184, 193 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "type": "posture", "subject_refs": [ "person" ], "object_refs": [], "confidence": "high", "temporal_status": "localized", "intervals": [ { "start_s": 1.0, "end_s": 7.5, "source_start_s": 215.49, "source_end_s": 221.99 } ], "evidence": "The selected person transitions downward between 1.000 and 2.000 s and remains forward-bent while working near the lower frame through the final sample.", "mask_relation": "The action subject matches the supplied selected-person mask.", "source_subject_refs": [ "person" ] }, { "id": "use_cloth", "mentions": [ { "text": "use", "ranges": [ [ 80, 83 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "type": "simple", "subject_refs": [ "person" ], "object_refs": [ "cloth" ], "confidence": "medium", "temporal_status": "localized", "intervals": [ { "start_s": 3.0, "end_s": 5.0, "source_start_s": 217.49, "source_end_s": 219.49 } ], "evidence": "At 3.000 and 4.000 s the selected person's hand is positioned at the lower frame with a small cleaning item visible near it.", "mask_relation": "The action subject matches the supplied selected-person mask.", "source_subject_refs": [ "person" ] }, { "id": "wipe_lower_frame", "mentions": [ { "text": "wipe", "ranges": [ [ 117, 121 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } }, { "text": "clean", "ranges": [ [ 202, 207 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "type": "simple", "subject_refs": [ "person" ], "object_refs": [ "lower_frame_section", "window_frame" ], "confidence": "high", "temporal_status": "localized", "intervals": [ { "start_s": 3.0, "end_s": 5.0, "source_start_s": 217.49, "source_end_s": 219.49 } ], "evidence": "The selected person reaches and works along the bottom edge of the window frame at 3.000 and 4.000 s, consistent with wiping the lower section.", "mask_relation": "The action subject matches the supplied selected-person mask.", "source_subject_refs": [ "person" ] } ], "notes": [ "The supplied mask consistently identifies the captioned person; all annotated action subjects are linked to that mask.", "No separate masks were supplied for the arms, cloth, right hand, lower frame section, or window frame, so their source-mask links are null.", "\"finishing\" was omitted because it aspectually indicates completion of an unspecified prior activity rather than naming a distinct visible action.", "\"remains\" was excluded as an aspectual/stative framing verb; the held posture is represented by \"bent over\".", "\"the top\" was treated as a spatial region rather than a separate concrete entity." ] }, "model_output_raw": { "caption": "After finishing with the top, the person lowers their arms and bends down. They use the cloth in their right hand to wipe the bottom portion of the window frame. Their posture remains bent over as they clean the lower section.", "entity_groups": [ { "id": "person", "label": "person", "spans": [ { "text": "the person", "range": [ 30, 40 ] }, { "text": "their", "range": [ 48, 53 ] }, { "text": "They", "range": [ 75, 79 ] }, { "text": "their", "range": [ 97, 102 ] }, { "text": "Their", "range": [ 162, 167 ] }, { "text": "they", "range": [ 197, 201 ] } ], "source_mask": "selected_person_mask" }, { "id": "arms", "label": "arms", "spans": [ { "text": "their arms", "range": [ 48, 58 ] } ], "source_mask": null }, { "id": "cloth", "label": "cloth", "spans": [ { "text": "the cloth", "range": [ 84, 93 ] } ], "source_mask": null }, { "id": "right_hand", "label": "right hand", "spans": [ { "text": "their right hand", "range": [ 97, 113 ] } ], "source_mask": null }, { "id": "lower_frame_section", "label": "lower section of window frame", "spans": [ { "text": "the bottom portion", "range": [ 122, 140 ] }, { "text": "the lower section", "range": [ 208, 225 ] } ], "source_mask": null }, { "id": "window_frame", "label": "window frame", "spans": [ { "text": "the window frame", "range": [ 144, 160 ] } ], "source_mask": null } ], "actions": [ { "id": "lower_arms", "mentions": [ { "text": "lowers", "ranges": [ [ 41, 47 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "type": "simple", "subject_refs": [ "person" ], "object_refs": [ "arms" ], "confidence": "high", "temporal_status": "localized", "intervals": [ { "start_s": 0.0, "end_s": 2.0 } ], "evidence": "At 0.000 s the selected person has both arms raised near the upper frame; by 1.000 s the arms have moved downward.", "mask_relation": "The action subject matches the supplied selected-person mask." }, { "id": "bend_down_and_remain_bent", "mentions": [ { "text": "bends down", "ranges": [ [ 63, 73 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } }, { "text": "bent over", "ranges": [ [ 184, 193 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "type": "posture", "subject_refs": [ "person" ], "object_refs": [], "confidence": "high", "temporal_status": "localized", "intervals": [ { "start_s": 1.0, "end_s": 7.5 } ], "evidence": "The selected person transitions downward between 1.000 and 2.000 s and remains forward-bent while working near the lower frame through the final sample.", "mask_relation": "The action subject matches the supplied selected-person mask." }, { "id": "use_cloth", "mentions": [ { "text": "use", "ranges": [ [ 80, 83 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "type": "simple", "subject_refs": [ "person" ], "object_refs": [ "cloth" ], "confidence": "medium", "temporal_status": "localized", "intervals": [ { "start_s": 3.0, "end_s": 5.0 } ], "evidence": "At 3.000 and 4.000 s the selected person's hand is positioned at the lower frame with a small cleaning item visible near it.", "mask_relation": "The action subject matches the supplied selected-person mask." }, { "id": "wipe_lower_frame", "mentions": [ { "text": "wipe", "ranges": [ [ 117, 121 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } }, { "text": "clean", "ranges": [ [ 202, 207 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "type": "simple", "subject_refs": [ "person" ], "object_refs": [ "lower_frame_section", "window_frame" ], "confidence": "high", "temporal_status": "localized", "intervals": [ { "start_s": 3.0, "end_s": 5.0 } ], "evidence": "The selected person reaches and works along the bottom edge of the window frame at 3.000 and 4.000 s, consistent with wiping the lower section.", "mask_relation": "The action subject matches the supplied selected-person mask." } ], "notes": [ "The supplied mask consistently identifies the captioned person; all annotated action subjects are linked to that mask.", "No separate masks were supplied for the arms, cloth, right hand, lower frame section, or window frame, so their source-mask links are null.", "\"finishing\" was omitted because it aspectually indicates completion of an unspecified prior activity rather than naming a distinct visible action.", "\"remains\" was excluded as an aspectual/stative framing verb; the held posture is represented by \"bent over\".", "\"the top\" was treated as a spatial region rather than a separate concrete entity." ] }, "normalization_repairs": 0 }, { "source_video_id": "v_9A0BwCxAKAg", "caption_count_category": 14, "clip": { "caption_index": 14, "caption": "The person stands up straight and observes the window frame. They briefly turn their head to the left and appear to speak. They then turn back to face the window, continuing to look at it.", "source_interval_sec": [ 221.99, 230.83 ], "loom_clip_id": "v_9A0BwCxAKAg_6653_6918", "clip_frame_interval_4fps": [ 6653, 6918 ], "sample_count_1fps": 9, "sampled_frames": [ { "frame": "00000", "timestamp_sec": 221.99, "offset_sec": 0, "rgb_path": "assets/cat_14/v_9A0BwCxAKAg_6653_6918/0000s_00000_rgb.jpg", "mask_path": "assets/cat_14/v_9A0BwCxAKAg_6653_6918/0000s_00000_mask.png", "mask_status": "available" }, { "frame": "00028", "timestamp_sec": 222.99, "offset_sec": 1, "rgb_path": "assets/cat_14/v_9A0BwCxAKAg_6653_6918/0001s_00028_rgb.jpg", "mask_path": "assets/cat_14/v_9A0BwCxAKAg_6653_6918/0001s_00028_mask.png", "mask_status": "available" }, { "frame": "00060", "timestamp_sec": 223.99, "offset_sec": 2, "rgb_path": "assets/cat_14/v_9A0BwCxAKAg_6653_6918/0002s_00060_rgb.jpg", "mask_path": "assets/cat_14/v_9A0BwCxAKAg_6653_6918/0002s_00060_mask.png", "mask_status": "available" }, { "frame": "00088", "timestamp_sec": 224.99, "offset_sec": 3, "rgb_path": "assets/cat_14/v_9A0BwCxAKAg_6653_6918/0003s_00088_rgb.jpg", "mask_path": "assets/cat_14/v_9A0BwCxAKAg_6653_6918/0003s_00088_mask.png", "mask_status": "available" }, { "frame": "00120", "timestamp_sec": 225.99, "offset_sec": 4, "rgb_path": "assets/cat_14/v_9A0BwCxAKAg_6653_6918/0004s_00120_rgb.jpg", "mask_path": "assets/cat_14/v_9A0BwCxAKAg_6653_6918/0004s_00120_mask.png", "mask_status": "available" }, { "frame": "00148", "timestamp_sec": 226.99, "offset_sec": 5, "rgb_path": "assets/cat_14/v_9A0BwCxAKAg_6653_6918/0005s_00148_rgb.jpg", "mask_path": "assets/cat_14/v_9A0BwCxAKAg_6653_6918/0005s_00148_mask.png", "mask_status": "available" }, { "frame": "00180", "timestamp_sec": 227.99, "offset_sec": 6, "rgb_path": "assets/cat_14/v_9A0BwCxAKAg_6653_6918/0006s_00180_rgb.jpg", "mask_path": "assets/cat_14/v_9A0BwCxAKAg_6653_6918/0006s_00180_mask.png", "mask_status": "available" }, { "frame": "00208", "timestamp_sec": 228.99, "offset_sec": 7, "rgb_path": "assets/cat_14/v_9A0BwCxAKAg_6653_6918/0007s_00208_rgb.jpg", "mask_path": "assets/cat_14/v_9A0BwCxAKAg_6653_6918/0007s_00208_mask.png", "mask_status": "available" }, { "frame": "00240", "timestamp_sec": 229.99, "offset_sec": 8, "rgb_path": "assets/cat_14/v_9A0BwCxAKAg_6653_6918/0008s_00240_rgb.jpg", "mask_path": "assets/cat_14/v_9A0BwCxAKAg_6653_6918/0008s_00240_mask.png", "mask_status": "available" } ], "obj_id": [ 0 ], "anno_id": [ 1277 ], "fps_inferred_from_clip_interval": 29.9774 }, "usage": { "input_tokens": 22304, "input_tokens_details": { "cache_write_tokens": 22301, "cached_tokens": 0 }, "output_tokens": 5309, "output_tokens_details": { "reasoning_tokens": 3241 }, "total_tokens": 27613 }, "model": "gpt-5.6-sol", "effort": "medium", "custom_id": "cat14_cap14_v_9A0BwCxAKAg_6653_6918", "annotation": { "caption": "The person stands up straight and observes the window frame. They briefly turn their head to the left and appear to speak. They then turn back to face the window, continuing to look at it.", "entity_groups": [ { "id": "person_1", "label": "person", "spans": [ { "text": "The person", "range": [ 0, 10 ] }, { "text": "They", "range": [ 61, 65 ] }, { "text": "their", "range": [ 79, 84 ] }, { "text": "They", "range": [ 123, 127 ] } ], "source_mask_link": "selected_person" }, { "id": "window_frame_1", "label": "window frame", "spans": [ { "text": "the window frame", "range": [ 43, 59 ] } ], "source_mask_link": null }, { "id": "head_1", "label": "head", "spans": [ { "text": "their head", "range": [ 79, 89 ] } ], "source_mask_link": null }, { "id": "window_1", "label": "window", "spans": [ { "text": "the window", "range": [ 151, 161 ] }, { "text": "it", "range": [ 185, 187 ] } ], "source_mask_link": null } ], "actions": [ { "id": "action_1", "mentions": [ { "text": "stands up", "ranges": [ [ 11, 20 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "type": "phrasal", "subject_refs": [ "person_1" ], "object_refs": [], "confidence": "medium", "temporal_status": "localized", "intervals": [ { "start_s": 0.0, "end_s": 2.0, "source_start_s": 221.99, "source_end_s": 223.99 } ], "evidence": "Across samples 0 and 1, the masked person's head and upper body rise into a more upright position.", "mask_relation": "subject_matches_supplied_selected_person_mask", "source_subject_refs": [ "person_1" ] }, { "id": "action_2", "mentions": [ { "text": "observes", "ranges": [ [ 34, 42 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } }, { "text": "look", "ranges": [ [ 177, 181 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "type": "other", "subject_refs": [ "person_1" ], "object_refs": [ "window_frame_1", "window_1" ], "confidence": "medium", "temporal_status": "localized", "intervals": [ { "start_s": 0.0, "end_s": 4.0, "source_start_s": 221.99, "source_end_s": 225.99 }, { "start_s": 6.0, "end_s": 8.840000000000003, "source_start_s": 227.99, "source_end_s": 230.83 } ], "evidence": "The masked person is oriented toward the window area in the early samples and again from sample 6 through the final sample. The two caption mentions describe the same continuing visual-attention event.", "mask_relation": "subject_matches_supplied_selected_person_mask", "source_subject_refs": [ "person_1" ] }, { "id": "action_3", "mentions": [ { "text": "turn", "ranges": [ [ 74, 78 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "type": "simple", "subject_refs": [ "person_1" ], "object_refs": [ "head_1" ], "confidence": "high", "temporal_status": "localized", "intervals": [ { "start_s": 4.0, "end_s": 6.0, "source_start_s": 225.99, "source_end_s": 227.99 } ], "evidence": "The masked person's head changes from a side-facing profile at sample 4 to a more front/left-facing orientation at sample 5.", "mask_relation": "subject_matches_supplied_selected_person_mask", "source_subject_refs": [ "person_1" ] }, { "id": "action_4", "mentions": [ { "text": "speak", "ranges": [ [ 116, 121 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "type": "simple", "subject_refs": [ "person_1" ], "object_refs": [], "confidence": "medium", "temporal_status": "localized", "intervals": [ { "start_s": 5.0, "end_s": 6.0, "source_start_s": 226.99, "source_end_s": 227.99 } ], "evidence": "At sample 5, the masked person faces more toward the viewer with the mouth visibly open; the sparse frame supports the caption's qualified speaking description.", "mask_relation": "subject_matches_supplied_selected_person_mask", "source_subject_refs": [ "person_1" ] }, { "id": "action_5", "mentions": [ { "text": "turn back", "ranges": [ [ 133, 142 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "type": "phrasal", "subject_refs": [ "person_1" ], "object_refs": [], "confidence": "high", "temporal_status": "localized", "intervals": [ { "start_s": 5.0, "end_s": 7.0, "source_start_s": 226.99, "source_end_s": 228.99 } ], "evidence": "Between samples 5 and 6, the masked person reverses the preceding head orientation and turns back toward the window.", "mask_relation": "subject_matches_supplied_selected_person_mask", "source_subject_refs": [ "person_1" ] }, { "id": "action_6", "mentions": [ { "text": "face", "ranges": [ [ 146, 150 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "type": "posture", "subject_refs": [ "person_1" ], "object_refs": [ "window_1" ], "confidence": "high", "temporal_status": "localized", "intervals": [ { "start_s": 6.0, "end_s": 8.840000000000003, "source_start_s": 227.99, "source_end_s": 230.83 } ], "evidence": "In samples 6 through 8, the masked person's head and body remain oriented toward the window; support reaches the clip's right boundary.", "mask_relation": "subject_matches_supplied_selected_person_mask", "source_subject_refs": [ "person_1" ] } ], "notes": [ "The supplied masklet consistently identifies the single person, so all annotated action subjects link to person_1.", "No separate source masks are supplied for the window frame, head, or window; their source-mask links remain null.", "The window frame and window are retained as separate entity groups because the first noun phrase names the frame component, while the later noun phrase and pronoun name the window.", "Temporal intervals are sparse sampled-frame support intervals rather than exact continuous-video onset and offset estimates." ] }, "model_output_raw": { "caption": "The person stands up straight and observes the window frame. They briefly turn their head to the left and appear to speak. They then turn back to face the window, continuing to look at it.", "entity_groups": [ { "id": "person_1", "entity_type": "person", "spans": [ { "text": "The person", "range": [ 0, 10 ] }, { "text": "They", "range": [ 61, 65 ] }, { "text": "their", "range": [ 79, 84 ] }, { "text": "They", "range": [ 123, 127 ] } ], "source_mask": "selected_person_masklet" }, { "id": "window_frame_1", "entity_type": "window frame", "spans": [ { "text": "the window frame", "range": [ 43, 59 ] } ], "source_mask": null }, { "id": "head_1", "entity_type": "head", "spans": [ { "text": "their head", "range": [ 79, 89 ] } ], "source_mask": null }, { "id": "window_1", "entity_type": "window", "spans": [ { "text": "the window", "range": [ 151, 161 ] }, { "text": "it", "range": [ 185, 187 ] } ], "source_mask": null } ], "actions": [ { "id": "action_1", "mentions": [ { "text": "stands up", "ranges": [ [ 11, 20 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "type": "phrasal", "subject_refs": [ "person_1" ], "object_refs": [], "confidence": "medium", "mask_relation": "subject_matches_supplied_selected_person_mask", "temporal_status": "localized", "intervals": [ { "start_s": 0.0, "end_s": 2.0, "source_start_s": 221.99, "source_end_s": 223.99, "supporting_sample_timestamps_s": [ 0.0, 1.0 ] } ], "evidence": "Across samples 0 and 1, the masked person's head and upper body rise into a more upright position." }, { "id": "action_2", "mentions": [ { "text": "observes", "ranges": [ [ 34, 42 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } }, { "text": "look", "ranges": [ [ 177, 181 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "type": "other", "subject_refs": [ "person_1" ], "object_refs": [ "window_frame_1", "window_1" ], "confidence": "medium", "mask_relation": "subject_matches_supplied_selected_person_mask", "temporal_status": "localized", "intervals": [ { "start_s": 0.0, "end_s": 4.0, "source_start_s": 221.99, "source_end_s": 225.99, "supporting_sample_timestamps_s": [ 0.0, 1.0, 2.0, 3.0 ] }, { "start_s": 6.0, "end_s": 8.84, "source_start_s": 227.99, "source_end_s": 230.83, "supporting_sample_timestamps_s": [ 6.0, 7.0, 8.0 ] } ], "evidence": "The masked person is oriented toward the window area in the early samples and again from sample 6 through the final sample. The two caption mentions describe the same continuing visual-attention event." }, { "id": "action_3", "mentions": [ { "text": "turn", "ranges": [ [ 74, 78 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "type": "simple", "subject_refs": [ "person_1" ], "object_refs": [ "head_1" ], "confidence": "high", "mask_relation": "subject_matches_supplied_selected_person_mask", "temporal_status": "localized", "intervals": [ { "start_s": 4.0, "end_s": 6.0, "source_start_s": 225.99, "source_end_s": 227.99, "supporting_sample_timestamps_s": [ 4.0, 5.0 ] } ], "evidence": "The masked person's head changes from a side-facing profile at sample 4 to a more front/left-facing orientation at sample 5." }, { "id": "action_4", "mentions": [ { "text": "speak", "ranges": [ [ 116, 121 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "type": "simple", "subject_refs": [ "person_1" ], "object_refs": [], "confidence": "medium", "mask_relation": "subject_matches_supplied_selected_person_mask", "temporal_status": "localized", "intervals": [ { "start_s": 5.0, "end_s": 6.0, "source_start_s": 226.99, "source_end_s": 227.99, "supporting_sample_timestamps_s": [ 5.0 ] } ], "evidence": "At sample 5, the masked person faces more toward the viewer with the mouth visibly open; the sparse frame supports the caption's qualified speaking description." }, { "id": "action_5", "mentions": [ { "text": "turn back", "ranges": [ [ 133, 142 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "type": "phrasal", "subject_refs": [ "person_1" ], "object_refs": [], "confidence": "high", "mask_relation": "subject_matches_supplied_selected_person_mask", "temporal_status": "localized", "intervals": [ { "start_s": 5.0, "end_s": 7.0, "source_start_s": 226.99, "source_end_s": 228.99, "supporting_sample_timestamps_s": [ 5.0, 6.0 ] } ], "evidence": "Between samples 5 and 6, the masked person reverses the preceding head orientation and turns back toward the window." }, { "id": "action_6", "mentions": [ { "text": "face", "ranges": [ [ 146, 150 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "type": "posture", "subject_refs": [ "person_1" ], "object_refs": [ "window_1" ], "confidence": "high", "mask_relation": "subject_matches_supplied_selected_person_mask", "temporal_status": "localized", "intervals": [ { "start_s": 6.0, "end_s": 8.84, "source_start_s": 227.99, "source_end_s": 230.83, "supporting_sample_timestamps_s": [ 6.0, 7.0, 8.0 ] } ], "evidence": "In samples 6 through 8, the masked person's head and body remain oriented toward the window; support reaches the clip's right boundary." } ], "notes": [ "The supplied masklet consistently identifies the single person, so all annotated action subjects link to person_1.", "No separate source masks are supplied for the window frame, head, or window; their source-mask links remain null.", "The window frame and window are retained as separate entity groups because the first noun phrase names the frame component, while the later noun phrase and pronoun name the window.", "Temporal intervals are sparse sampled-frame support intervals rather than exact continuous-video onset and offset estimates." ] }, "normalization_repairs": 0 }, { "source_video_id": "v_6gZuc4umTPk", "caption_count_category": 15, "clip": { "caption_index": 1, "caption": "A person with long brown hair, wearing a patterned top over a black long-sleeved shirt, sits and speaks. They initially bring their hands toward their face, then place their left hand on another person's shoulder. They proceed to engage in a hand-clapping game before leaning forward towards a surface.", "source_interval_sec": [ 0.0, 14.2 ], "loom_clip_id": "v_6gZuc4umTPk_1_426", "clip_frame_interval_4fps": [ 1, 426 ], "sample_count_1fps": 15, "sampled_frames": [ { "frame": "00000", "timestamp_sec": 0.0, "offset_sec": 0, "rgb_path": "assets/cat_15/v_6gZuc4umTPk_1_426/0000s_00000_rgb.jpg", "mask_path": "assets/cat_15/v_6gZuc4umTPk_1_426/0000s_00000_mask.png", "mask_status": "available" }, { "frame": "00028", "timestamp_sec": 1.0, "offset_sec": 1, "rgb_path": "assets/cat_15/v_6gZuc4umTPk_1_426/0001s_00028_rgb.jpg", "mask_path": "assets/cat_15/v_6gZuc4umTPk_1_426/0001s_00028_mask.png", "mask_status": "available" }, { "frame": "00060", "timestamp_sec": 2.0, "offset_sec": 2, "rgb_path": "assets/cat_15/v_6gZuc4umTPk_1_426/0002s_00060_rgb.jpg", "mask_path": "assets/cat_15/v_6gZuc4umTPk_1_426/0002s_00060_mask.png", "mask_status": "available" }, { "frame": "00088", "timestamp_sec": 3.0, "offset_sec": 3, "rgb_path": "assets/cat_15/v_6gZuc4umTPk_1_426/0003s_00088_rgb.jpg", "mask_path": "assets/cat_15/v_6gZuc4umTPk_1_426/0003s_00088_mask.png", "mask_status": "available" }, { "frame": "00120", "timestamp_sec": 4.0, "offset_sec": 4, "rgb_path": "assets/cat_15/v_6gZuc4umTPk_1_426/0004s_00120_rgb.jpg", "mask_path": "assets/cat_15/v_6gZuc4umTPk_1_426/0004s_00120_mask.png", "mask_status": "available" }, { "frame": "00148", "timestamp_sec": 5.0, "offset_sec": 5, "rgb_path": "assets/cat_15/v_6gZuc4umTPk_1_426/0005s_00148_rgb.jpg", "mask_path": "assets/cat_15/v_6gZuc4umTPk_1_426/0005s_00148_mask.png", "mask_status": "available" }, { "frame": "00180", "timestamp_sec": 6.0, "offset_sec": 6, "rgb_path": "assets/cat_15/v_6gZuc4umTPk_1_426/0006s_00180_rgb.jpg", "mask_path": "assets/cat_15/v_6gZuc4umTPk_1_426/0006s_00180_mask.png", "mask_status": "available" }, { "frame": "00208", "timestamp_sec": 7.0, "offset_sec": 7, "rgb_path": "assets/cat_15/v_6gZuc4umTPk_1_426/0007s_00208_rgb.jpg", "mask_path": "assets/cat_15/v_6gZuc4umTPk_1_426/0007s_00208_mask.png", "mask_status": "available" }, { "frame": "00240", "timestamp_sec": 8.0, "offset_sec": 8, "rgb_path": "assets/cat_15/v_6gZuc4umTPk_1_426/0008s_00240_rgb.jpg", "mask_path": "assets/cat_15/v_6gZuc4umTPk_1_426/0008s_00240_mask.png", "mask_status": "available" }, { "frame": "00268", "timestamp_sec": 9.0, "offset_sec": 9, "rgb_path": "assets/cat_15/v_6gZuc4umTPk_1_426/0009s_00268_rgb.jpg", "mask_path": "assets/cat_15/v_6gZuc4umTPk_1_426/0009s_00268_mask.png", "mask_status": "available" }, { "frame": "00300", "timestamp_sec": 10.0, "offset_sec": 10, "rgb_path": "assets/cat_15/v_6gZuc4umTPk_1_426/0010s_00300_rgb.jpg", "mask_path": "assets/cat_15/v_6gZuc4umTPk_1_426/0010s_00300_mask.png", "mask_status": "available" }, { "frame": "00328", "timestamp_sec": 11.0, "offset_sec": 11, "rgb_path": "assets/cat_15/v_6gZuc4umTPk_1_426/0011s_00328_rgb.jpg", "mask_path": "assets/cat_15/v_6gZuc4umTPk_1_426/0011s_00328_mask.png", "mask_status": "available" }, { "frame": "00360", "timestamp_sec": 12.0, "offset_sec": 12, "rgb_path": "assets/cat_15/v_6gZuc4umTPk_1_426/0012s_00360_rgb.jpg", "mask_path": "assets/cat_15/v_6gZuc4umTPk_1_426/0012s_00360_mask.png", "mask_status": "available" }, { "frame": "00388", "timestamp_sec": 13.0, "offset_sec": 13, "rgb_path": "assets/cat_15/v_6gZuc4umTPk_1_426/0013s_00388_rgb.jpg", "mask_path": "assets/cat_15/v_6gZuc4umTPk_1_426/0013s_00388_mask.png", "mask_status": "available" }, { "frame": "00420", "timestamp_sec": 14.0, "offset_sec": 14, "rgb_path": "assets/cat_15/v_6gZuc4umTPk_1_426/0014s_00420_rgb.jpg", "mask_path": "assets/cat_15/v_6gZuc4umTPk_1_426/0014s_00420_mask.png", "mask_status": "available" } ], "obj_id": [ 0 ], "anno_id": [ 934 ], "fps_inferred_from_clip_interval": 29.9296 }, "usage": { "input_tokens": 22655, "input_tokens_details": { "cache_write_tokens": 22652, "cached_tokens": 0 }, "output_tokens": 5160, "output_tokens_details": { "reasoning_tokens": 2854 }, "total_tokens": 27815 }, "model": "gpt-5.6-sol", "effort": "medium", "custom_id": "cat15_cap01_v_6gZuc4umTPk_1_426", "annotation": { "caption": "A person with long brown hair, wearing a patterned top over a black long-sleeved shirt, sits and speaks. They initially bring their hands toward their face, then place their left hand on another person's shoulder. They proceed to engage in a hand-clapping game before leaning forward towards a surface.", "entity_groups": [ { "id": "person_1", "label": "person", "spans": [ { "text": "A person", "range": [ 0, 8 ] }, { "text": "They", "range": [ 105, 109 ] }, { "text": "their", "range": [ 126, 131 ] }, { "text": "their", "range": [ 145, 150 ] }, { "text": "their", "range": [ 168, 173 ] }, { "text": "They", "range": [ 214, 218 ] } ], "source_mask_link": "selected_person" }, { "id": "hair_1", "label": "hair", "spans": [ { "text": "long brown hair", "range": [ 14, 29 ] } ], "source_mask_link": null }, { "id": "top_1", "label": "top", "spans": [ { "text": "a patterned top", "range": [ 39, 54 ] } ], "source_mask_link": null }, { "id": "shirt_1", "label": "shirt", "spans": [ { "text": "a black long-sleeved shirt", "range": [ 60, 86 ] } ], "source_mask_link": null }, { "id": "hands_1", "label": "hands", "spans": [ { "text": "their hands", "range": [ 126, 137 ] } ], "source_mask_link": null }, { "id": "face_1", "label": "face", "spans": [ { "text": "their face", "range": [ 145, 155 ] } ], "source_mask_link": null }, { "id": "left_hand_1", "label": "left hand", "spans": [ { "text": "their left hand", "range": [ 168, 183 ] } ], "source_mask_link": null }, { "id": "person_2", "label": "person", "spans": [ { "text": "another person's", "range": [ 187, 203 ] } ], "source_mask_link": null }, { "id": "shoulder_1", "label": "shoulder", "spans": [ { "text": "another person's shoulder", "range": [ 187, 212 ] } ], "source_mask_link": null }, { "id": "surface_1", "label": "surface", "spans": [ { "text": "a surface", "range": [ 292, 301 ] } ], "source_mask_link": null } ], "actions": [ { "id": "action_1", "mentions": [ { "text": "wearing", "ranges": [ [ 31, 38 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "type": "simple", "subject_refs": [ "person_1" ], "object_refs": [ "top_1", "shirt_1" ], "confidence": "high", "temporal_status": "localized", "intervals": [ { "start_s": 0.0, "end_s": 14.2, "source_start_s": 0.0, "source_end_s": 14.2 } ], "evidence": "The mask-linked person visibly wears the patterned top and black long-sleeved shirt throughout the sampled clip.", "mask_relation": "subject_matches_supplied_selected_person_mask", "source_subject_refs": [ "person_1" ] }, { "id": "action_2", "mentions": [ { "text": "sits", "ranges": [ [ 88, 92 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "type": "posture", "subject_refs": [ "person_1" ], "object_refs": [], "confidence": "high", "temporal_status": "localized", "intervals": [ { "start_s": 0.0, "end_s": 14.2, "source_start_s": 0.0, "source_end_s": 14.2 } ], "evidence": "The mask-linked person maintains a seated posture across the sampled frames, including while reaching forward near the end.", "mask_relation": "subject_matches_supplied_selected_person_mask", "source_subject_refs": [ "person_1" ] }, { "id": "action_3", "mentions": [ { "text": "speaks", "ranges": [ [ 97, 103 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "type": "simple", "subject_refs": [ "person_1" ], "object_refs": [], "confidence": "medium", "temporal_status": "localized", "intervals": [ { "start_s": 0.0, "end_s": 1.0, "source_start_s": 0.0, "source_end_s": 1.0 }, { "start_s": 2.0, "end_s": 4.0, "source_start_s": 2.0, "source_end_s": 4.0 }, { "start_s": 9.0, "end_s": 14.0, "source_start_s": 9.0, "source_end_s": 14.0 } ], "evidence": "The mask-linked person's changing open and shaped mouth is visually consistent with speaking, though speech cannot be confirmed from still frames alone.", "mask_relation": "subject_matches_supplied_selected_person_mask", "source_subject_refs": [ "person_1" ] }, { "id": "action_4", "mentions": [ { "text": "bring", "ranges": [ [ 120, 125 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "type": "simple", "subject_refs": [ "person_1" ], "object_refs": [ "hands_1", "face_1" ], "confidence": "high", "temporal_status": "localized", "intervals": [ { "start_s": 0.0, "end_s": 1.0, "source_start_s": 0.0, "source_end_s": 1.0 } ], "evidence": "At 0.000 seconds, the mask-linked person has both hands raised beside and toward her face; they are lowered by 1.000 seconds.", "mask_relation": "subject_matches_supplied_selected_person_mask", "source_subject_refs": [ "person_1" ] }, { "id": "action_5", "mentions": [ { "text": "place", "ranges": [ [ 162, 167 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "type": "simple", "subject_refs": [ "person_1" ], "object_refs": [ "left_hand_1", "shoulder_1", "person_2" ], "confidence": "high", "temporal_status": "localized", "intervals": [ { "start_s": 3.0, "end_s": 4.0, "source_start_s": 3.0, "source_end_s": 4.0 } ], "evidence": "At 3.000 seconds, the mask-linked person's left hand visibly rests on the blonde person's shoulder.", "mask_relation": "subject_matches_supplied_selected_person_mask", "source_subject_refs": [ "person_1" ] }, { "id": "action_6", "mentions": [ { "text": "engage", "ranges": [ [ 230, 236 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "type": "prepositional", "subject_refs": [ "person_1" ], "object_refs": [], "confidence": "high", "temporal_status": "localized", "intervals": [ { "start_s": 5.0, "end_s": 11.0, "source_start_s": 5.0, "source_end_s": 11.0 } ], "evidence": "From 5.000 through 10.000 seconds, the mask-linked person repeatedly brings her hands together with or manipulates the other person's hands, supporting the stated hand-clapping game.", "mask_relation": "subject_matches_supplied_selected_person_mask", "source_subject_refs": [ "person_1" ] }, { "id": "action_7", "mentions": [ { "text": "leaning", "ranges": [ [ 268, 275 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "type": "simple", "subject_refs": [ "person_1" ], "object_refs": [ "surface_1" ], "confidence": "high", "temporal_status": "localized", "intervals": [ { "start_s": 13.0, "end_s": 14.2, "source_start_s": 13.0, "source_end_s": 14.2 } ], "evidence": "At 13.000 and 14.000 seconds, the mask-linked person moves her torso forward and reaches toward the visible surface, with support continuing to the clip boundary.", "mask_relation": "subject_matches_supplied_selected_person_mask", "source_subject_refs": [ "person_1" ] } ], "notes": [ "The supplied mask consistently identifies the long-brown-haired person in the patterned top, so only that person's actions are annotated.", "The blonde person is visible but has no supplied instance mask; no actions are assigned to that person.", "The hair, garments, body parts, shoulder, and surface have no separate supplied masks, so their source-mask links are null.", "The phrase 'a hand-clapping game' names an activity rather than a tangible entity and therefore is not an entity group." ] }, "model_output_raw": { "caption": "A person with long brown hair, wearing a patterned top over a black long-sleeved shirt, sits and speaks. They initially bring their hands toward their face, then place their left hand on another person's shoulder. They proceed to engage in a hand-clapping game before leaning forward towards a surface.", "entity_groups": [ { "id": "person_1", "label": "person", "spans": [ { "text": "A person", "range": [ 0, 8 ] }, { "text": "They", "range": [ 105, 109 ] }, { "text": "their", "range": [ 126, 131 ] }, { "text": "their", "range": [ 145, 150 ] }, { "text": "their", "range": [ 169, 174 ] }, { "text": "They", "range": [ 215, 219 ] } ], "source_mask": "selected_person_masklet" }, { "id": "hair_1", "label": "hair", "spans": [ { "text": "long brown hair", "range": [ 14, 29 ] } ], "source_mask": null }, { "id": "top_1", "label": "top", "spans": [ { "text": "a patterned top", "range": [ 39, 54 ] } ], "source_mask": null }, { "id": "shirt_1", "label": "shirt", "spans": [ { "text": "a black long-sleeved shirt", "range": [ 60, 86 ] } ], "source_mask": null }, { "id": "hands_1", "label": "hands", "spans": [ { "text": "their hands", "range": [ 126, 137 ] } ], "source_mask": null }, { "id": "face_1", "label": "face", "spans": [ { "text": "their face", "range": [ 145, 155 ] } ], "source_mask": null }, { "id": "left_hand_1", "label": "left hand", "spans": [ { "text": "their left hand", "range": [ 169, 184 ] } ], "source_mask": null }, { "id": "person_2", "label": "person", "spans": [ { "text": "another person's", "range": [ 188, 204 ] } ], "source_mask": null }, { "id": "shoulder_1", "label": "shoulder", "spans": [ { "text": "another person's shoulder", "range": [ 188, 213 ] } ], "source_mask": null }, { "id": "surface_1", "label": "surface", "spans": [ { "text": "a surface", "range": [ 293, 302 ] } ], "source_mask": null } ], "actions": [ { "id": "action_1", "mentions": [ { "text": "wearing", "occurrence": 1, "ranges": [ [ 31, 38 ] ], "particle": { "text": null, "occurrence": 1 } } ], "type": "simple", "subject_refs": [ "person_1" ], "object_refs": [ "top_1", "shirt_1" ], "confidence": "high", "temporal_status": "localized", "intervals": [ { "start_s": 0.0, "end_s": 14.2 } ], "evidence": "The mask-linked person visibly wears the patterned top and black long-sleeved shirt throughout the sampled clip.", "mask_relation": "subject_matches_supplied_selected_person_mask" }, { "id": "action_2", "mentions": [ { "text": "sits", "occurrence": 1, "ranges": [ [ 88, 92 ] ], "particle": { "text": null, "occurrence": 1 } } ], "type": "posture", "subject_refs": [ "person_1" ], "object_refs": [], "confidence": "high", "temporal_status": "localized", "intervals": [ { "start_s": 0.0, "end_s": 14.2 } ], "evidence": "The mask-linked person maintains a seated posture across the sampled frames, including while reaching forward near the end.", "mask_relation": "subject_matches_supplied_selected_person_mask" }, { "id": "action_3", "mentions": [ { "text": "speaks", "occurrence": 1, "ranges": [ [ 97, 103 ] ], "particle": { "text": null, "occurrence": 1 } } ], "type": "simple", "subject_refs": [ "person_1" ], "object_refs": [], "confidence": "medium", "temporal_status": "localized", "intervals": [ { "start_s": 0.0, "end_s": 1.0 }, { "start_s": 2.0, "end_s": 4.0 }, { "start_s": 9.0, "end_s": 14.0 } ], "evidence": "The mask-linked person's changing open and shaped mouth is visually consistent with speaking, though speech cannot be confirmed from still frames alone.", "mask_relation": "subject_matches_supplied_selected_person_mask" }, { "id": "action_4", "mentions": [ { "text": "bring", "occurrence": 1, "ranges": [ [ 120, 125 ] ], "particle": { "text": null, "occurrence": 1 } } ], "type": "simple", "subject_refs": [ "person_1" ], "object_refs": [ "hands_1", "face_1" ], "confidence": "high", "temporal_status": "localized", "intervals": [ { "start_s": 0.0, "end_s": 1.0 } ], "evidence": "At 0.000 seconds, the mask-linked person has both hands raised beside and toward her face; they are lowered by 1.000 seconds.", "mask_relation": "subject_matches_supplied_selected_person_mask" }, { "id": "action_5", "mentions": [ { "text": "place", "occurrence": 1, "ranges": [ [ 163, 168 ] ], "particle": { "text": null, "occurrence": 1 } } ], "type": "simple", "subject_refs": [ "person_1" ], "object_refs": [ "left_hand_1", "shoulder_1", "person_2" ], "confidence": "high", "temporal_status": "localized", "intervals": [ { "start_s": 3.0, "end_s": 4.0 } ], "evidence": "At 3.000 seconds, the mask-linked person's left hand visibly rests on the blonde person's shoulder.", "mask_relation": "subject_matches_supplied_selected_person_mask" }, { "id": "action_6", "mentions": [ { "text": "engage", "occurrence": 1, "ranges": [ [ 231, 237 ] ], "particle": { "text": null, "occurrence": 1 } } ], "type": "prepositional", "subject_refs": [ "person_1" ], "object_refs": [], "confidence": "high", "temporal_status": "localized", "intervals": [ { "start_s": 5.0, "end_s": 11.0 } ], "evidence": "From 5.000 through 10.000 seconds, the mask-linked person repeatedly brings her hands together with or manipulates the other person's hands, supporting the stated hand-clapping game.", "mask_relation": "subject_matches_supplied_selected_person_mask" }, { "id": "action_7", "mentions": [ { "text": "leaning", "occurrence": 1, "ranges": [ [ 269, 276 ] ], "particle": { "text": null, "occurrence": 1 } } ], "type": "simple", "subject_refs": [ "person_1" ], "object_refs": [ "surface_1" ], "confidence": "high", "temporal_status": "localized", "intervals": [ { "start_s": 13.0, "end_s": 14.2 } ], "evidence": "At 13.000 and 14.000 seconds, the mask-linked person moves her torso forward and reaches toward the visible surface, with support continuing to the clip boundary.", "mask_relation": "subject_matches_supplied_selected_person_mask" } ], "notes": [ "The supplied mask consistently identifies the long-brown-haired person in the patterned top, so only that person's actions are annotated.", "The blonde person is visible but has no supplied instance mask; no actions are assigned to that person.", "The hair, garments, body parts, shoulder, and surface have no separate supplied masks, so their source-mask links are null.", "The phrase 'a hand-clapping game' names an activity rather than a tangible entity and therefore is not an entity group." ] }, "normalization_repairs": 9 }, { "source_video_id": "v_6gZuc4umTPk", "caption_count_category": 15, "clip": { "caption_index": 2, "caption": "A person with light brown hair, wearing a black and white patterned zip-up hoodie over a pink shirt, sits at a blue table. They pick up a bottle of red nail polish and proceed to open it.", "source_interval_sec": [ 15.83, 22.83 ], "loom_clip_id": "v_6gZuc4umTPk_475_685", "clip_frame_interval_4fps": [ 475, 685 ], "sample_count_1fps": 7, "sampled_frames": [ { "frame": "00000", "timestamp_sec": 15.83, "offset_sec": 0, "rgb_path": "assets/cat_15/v_6gZuc4umTPk_475_685/0000s_00000_rgb.jpg", "mask_path": "assets/cat_15/v_6gZuc4umTPk_475_685/0000s_00000_mask.png", "mask_status": "available" }, { "frame": "00032", "timestamp_sec": 16.83, "offset_sec": 1, "rgb_path": "assets/cat_15/v_6gZuc4umTPk_475_685/0001s_00032_rgb.jpg", "mask_path": "assets/cat_15/v_6gZuc4umTPk_475_685/0001s_00032_mask.png", "mask_status": "available" }, { "frame": "00060", "timestamp_sec": 17.83, "offset_sec": 2, "rgb_path": "assets/cat_15/v_6gZuc4umTPk_475_685/0002s_00060_rgb.jpg", "mask_path": "assets/cat_15/v_6gZuc4umTPk_475_685/0002s_00060_mask.png", "mask_status": "available" }, { "frame": "00092", "timestamp_sec": 18.83, "offset_sec": 3, "rgb_path": "assets/cat_15/v_6gZuc4umTPk_475_685/0003s_00092_rgb.jpg", "mask_path": "assets/cat_15/v_6gZuc4umTPk_475_685/0003s_00092_mask.png", "mask_status": "available" }, { "frame": "00120", "timestamp_sec": 19.83, "offset_sec": 4, "rgb_path": "assets/cat_15/v_6gZuc4umTPk_475_685/0004s_00120_rgb.jpg", "mask_path": "assets/cat_15/v_6gZuc4umTPk_475_685/0004s_00120_mask.png", "mask_status": "available" }, { "frame": "00152", "timestamp_sec": 20.83, "offset_sec": 5, "rgb_path": "assets/cat_15/v_6gZuc4umTPk_475_685/0005s_00152_rgb.jpg", "mask_path": "assets/cat_15/v_6gZuc4umTPk_475_685/0005s_00152_mask.png", "mask_status": "available" }, { "frame": "00180", "timestamp_sec": 21.83, "offset_sec": 6, "rgb_path": "assets/cat_15/v_6gZuc4umTPk_475_685/0006s_00180_rgb.jpg", "mask_path": "assets/cat_15/v_6gZuc4umTPk_475_685/0006s_00180_mask.png", "mask_status": "available" } ], "obj_id": [ 0 ], "anno_id": [ 940 ], "fps_inferred_from_clip_interval": 30.0 }, "usage": { "input_tokens": 18741, "input_tokens_details": { "cache_write_tokens": 18738, "cached_tokens": 0 }, "output_tokens": 4507, "output_tokens_details": { "reasoning_tokens": 3106 }, "total_tokens": 23248 }, "model": "gpt-5.6-sol", "effort": "medium", "custom_id": "cat15_cap02_v_6gZuc4umTPk_475_685", "annotation": { "caption": "A person with light brown hair, wearing a black and white patterned zip-up hoodie over a pink shirt, sits at a blue table. They pick up a bottle of red nail polish and proceed to open it.", "entity_groups": [ { "id": "person_1", "label": "person", "spans": [ { "text": "A person", "range": [ 0, 8 ] }, { "text": "They", "range": [ 123, 127 ] } ], "source_mask_link": "selected_person" }, { "id": "hair_1", "label": "hair", "spans": [ { "text": "light brown hair", "range": [ 14, 30 ] } ], "source_mask_link": null }, { "id": "hoodie_1", "label": "hoodie", "spans": [ { "text": "a black and white patterned zip-up hoodie", "range": [ 40, 81 ] } ], "source_mask_link": null }, { "id": "shirt_1", "label": "shirt", "spans": [ { "text": "a pink shirt", "range": [ 87, 99 ] } ], "source_mask_link": null }, { "id": "table_1", "label": "table", "spans": [ { "text": "a blue table", "range": [ 109, 121 ] } ], "source_mask_link": null }, { "id": "bottle_1", "label": "bottle", "spans": [ { "text": "a bottle of red nail polish", "range": [ 136, 163 ] }, { "text": "it", "range": [ 184, 186 ] } ], "source_mask_link": null }, { "id": "nail_polish_1", "label": "nail polish", "spans": [ { "text": "a bottle of red nail polish", "range": [ 136, 163 ] } ], "source_mask_link": null } ], "actions": [ { "id": "action_1", "mentions": [ { "text": "wearing", "ranges": [ [ 32, 39 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "type": "simple", "subject_refs": [ "person_1" ], "object_refs": [ "hoodie_1", "shirt_1" ], "confidence": "high", "temporal_status": "localized", "intervals": [ { "start_s": 1.0, "end_s": 6.999999999999998, "source_start_s": 16.83, "source_end_s": 22.83 } ], "evidence": "Samples at 1.0–6.0 seconds show the selected person continuously wearing the patterned hoodie over the pink shirt.", "mask_relation": "The action subject matches the supplied selected-person mask.", "source_subject_refs": [ "person_1" ] }, { "id": "action_2", "mentions": [ { "text": "sits", "ranges": [ [ 101, 105 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "type": "posture", "subject_refs": [ "person_1" ], "object_refs": [], "confidence": "high", "temporal_status": "localized", "intervals": [ { "start_s": 1.0, "end_s": 6.999999999999998, "source_start_s": 16.83, "source_end_s": 22.83 } ], "evidence": "Samples at 1.0–6.0 seconds show the selected person maintaining a seated posture behind the blue table.", "mask_relation": "The action subject matches the supplied selected-person mask.", "source_subject_refs": [ "person_1" ] }, { "id": "action_3", "mentions": [ { "text": "pick up", "ranges": [ [ 128, 135 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "type": "phrasal", "subject_refs": [ "person_1" ], "object_refs": [ "bottle_1" ], "confidence": "low", "temporal_status": "uncertain", "intervals": [], "evidence": "The selected person reaches near the bottle in the early samples, but no sampled clean frame clearly shows the bottle lifted from the table.", "mask_relation": "The action subject matches the supplied selected-person mask.", "source_subject_refs": [ "person_1" ] }, { "id": "action_4", "mentions": [ { "text": "open", "ranges": [ [ 179, 183 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "type": "simple", "subject_refs": [ "person_1" ], "object_refs": [ "bottle_1" ], "confidence": "medium", "temporal_status": "localized", "intervals": [ { "start_s": 2.0, "end_s": 4.0, "source_start_s": 17.83, "source_end_s": 19.83 } ], "evidence": "At 2.0 seconds the bottle is capped while the selected person's hands are nearby; by 3.0 seconds its neck is exposed and the masked hands remain engaged with it.", "mask_relation": "The action subject matches the supplied selected-person mask.", "source_subject_refs": [ "person_1" ] } ], "notes": [ "The RGB frame at 0.0 seconds is blank and its paired mask is empty, so it supplies no action support.", "The supplied masklet identifies only the central person corresponding to “A person” and “They.”", "No separate source masks were supplied for the hair, clothing, table, bottle, or nail polish.", "The bottle and nail-polish groups share the full container-content span under the container-plus-content grounding rule.", "“proceed” is excluded as a control/aspectual predicate introducing “open.”" ] }, "model_output_raw": { "caption": "A person with light brown hair, wearing a black and white patterned zip-up hoodie over a pink shirt, sits at a blue table. They pick up a bottle of red nail polish and proceed to open it.", "entity_groups": [ { "id": "person_1", "label": "person", "spans": [ { "text": "A person", "range": [ 0, 8 ] }, { "text": "They", "range": [ 123, 127 ] } ], "source_mask_link": "selected_person_masklet" }, { "id": "hair_1", "label": "hair", "spans": [ { "text": "light brown hair", "range": [ 14, 30 ] } ], "source_mask_link": null }, { "id": "hoodie_1", "label": "hoodie", "spans": [ { "text": "a black and white patterned zip-up hoodie", "range": [ 40, 81 ] } ], "source_mask_link": null }, { "id": "shirt_1", "label": "shirt", "spans": [ { "text": "a pink shirt", "range": [ 87, 99 ] } ], "source_mask_link": null }, { "id": "table_1", "label": "table", "spans": [ { "text": "a blue table", "range": [ 109, 121 ] } ], "source_mask_link": null }, { "id": "bottle_1", "label": "bottle", "spans": [ { "text": "a bottle of red nail polish", "range": [ 136, 163 ] }, { "text": "it", "range": [ 184, 186 ] } ], "source_mask_link": null }, { "id": "nail_polish_1", "label": "nail polish", "spans": [ { "text": "a bottle of red nail polish", "range": [ 136, 163 ] } ], "source_mask_link": null } ], "actions": [ { "mentions": [ { "text": "wearing", "ranges": [ [ 32, 39 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "type": "simple", "subject_refs": [ "person_1" ], "object_refs": [ "hoodie_1", "shirt_1" ], "confidence": "high", "temporal_status": "localized", "intervals": [ { "start_s": 1.0, "end_s": 7.0 } ], "evidence": "Samples at 1.0–6.0 seconds show the selected person continuously wearing the patterned hoodie over the pink shirt.", "mask_relation": "The action subject matches the supplied selected-person mask." }, { "mentions": [ { "text": "sits", "ranges": [ [ 101, 105 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "type": "posture", "subject_refs": [ "person_1" ], "object_refs": [], "confidence": "high", "temporal_status": "localized", "intervals": [ { "start_s": 1.0, "end_s": 7.0 } ], "evidence": "Samples at 1.0–6.0 seconds show the selected person maintaining a seated posture behind the blue table.", "mask_relation": "The action subject matches the supplied selected-person mask." }, { "mentions": [ { "text": "pick up", "ranges": [ [ 128, 135 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "type": "phrasal", "subject_refs": [ "person_1" ], "object_refs": [ "bottle_1" ], "confidence": "low", "temporal_status": "uncertain", "intervals": [], "evidence": "The selected person reaches near the bottle in the early samples, but no sampled clean frame clearly shows the bottle lifted from the table.", "mask_relation": "The action subject matches the supplied selected-person mask." }, { "mentions": [ { "text": "open", "ranges": [ [ 179, 183 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "type": "simple", "subject_refs": [ "person_1" ], "object_refs": [ "bottle_1" ], "confidence": "medium", "temporal_status": "localized", "intervals": [ { "start_s": 2.0, "end_s": 4.0 } ], "evidence": "At 2.0 seconds the bottle is capped while the selected person's hands are nearby; by 3.0 seconds its neck is exposed and the masked hands remain engaged with it.", "mask_relation": "The action subject matches the supplied selected-person mask." } ], "notes": [ "The RGB frame at 0.0 seconds is blank and its paired mask is empty, so it supplies no action support.", "The supplied masklet identifies only the central person corresponding to “A person” and “They.”", "No separate source masks were supplied for the hair, clothing, table, bottle, or nail polish.", "The bottle and nail-polish groups share the full container-content span under the container-plus-content grounding rule.", "“proceed” is excluded as a control/aspectual predicate introducing “open.”" ] }, "normalization_repairs": 0 }, { "source_video_id": "v_6gZuc4umTPk", "caption_count_category": 15, "clip": { "caption_index": 3, "caption": "The person leans forward, holding another person's hand steady with their left hand. They carefully apply the red nail polish to one of the fingers. After finishing, they sit up and place the brush back into the bottle.", "source_interval_sec": [ 22.83, 34.83 ], "loom_clip_id": "v_6gZuc4umTPk_685_1045", "clip_frame_interval_4fps": [ 685, 1045 ], "sample_count_1fps": 12, "sampled_frames": [ { "frame": "00000", "timestamp_sec": 22.83, "offset_sec": 0, "rgb_path": "assets/cat_15/v_6gZuc4umTPk_685_1045/0000s_00000_rgb.jpg", "mask_path": "assets/cat_15/v_6gZuc4umTPk_685_1045/0000s_00000_mask.png", "mask_status": "available" }, { "frame": "00028", "timestamp_sec": 23.83, "offset_sec": 1, "rgb_path": "assets/cat_15/v_6gZuc4umTPk_685_1045/0001s_00028_rgb.jpg", "mask_path": "assets/cat_15/v_6gZuc4umTPk_685_1045/0001s_00028_mask.png", "mask_status": "available" }, { "frame": "00060", "timestamp_sec": 24.83, "offset_sec": 2, "rgb_path": "assets/cat_15/v_6gZuc4umTPk_685_1045/0002s_00060_rgb.jpg", "mask_path": "assets/cat_15/v_6gZuc4umTPk_685_1045/0002s_00060_mask.png", "mask_status": "available" }, { "frame": "00088", "timestamp_sec": 25.83, "offset_sec": 3, "rgb_path": "assets/cat_15/v_6gZuc4umTPk_685_1045/0003s_00088_rgb.jpg", "mask_path": "assets/cat_15/v_6gZuc4umTPk_685_1045/0003s_00088_mask.png", "mask_status": "available" }, { "frame": "00120", "timestamp_sec": 26.83, "offset_sec": 4, "rgb_path": "assets/cat_15/v_6gZuc4umTPk_685_1045/0004s_00120_rgb.jpg", "mask_path": "assets/cat_15/v_6gZuc4umTPk_685_1045/0004s_00120_mask.png", "mask_status": "available" }, { "frame": "00148", "timestamp_sec": 27.83, "offset_sec": 5, "rgb_path": "assets/cat_15/v_6gZuc4umTPk_685_1045/0005s_00148_rgb.jpg", "mask_path": "assets/cat_15/v_6gZuc4umTPk_685_1045/0005s_00148_mask.png", "mask_status": "available" }, { "frame": "00180", "timestamp_sec": 28.83, "offset_sec": 6, "rgb_path": "assets/cat_15/v_6gZuc4umTPk_685_1045/0006s_00180_rgb.jpg", "mask_path": "assets/cat_15/v_6gZuc4umTPk_685_1045/0006s_00180_mask.png", "mask_status": "available" }, { "frame": "00208", "timestamp_sec": 29.83, "offset_sec": 7, "rgb_path": "assets/cat_15/v_6gZuc4umTPk_685_1045/0007s_00208_rgb.jpg", "mask_path": "assets/cat_15/v_6gZuc4umTPk_685_1045/0007s_00208_mask.png", "mask_status": "available" }, { "frame": "00240", "timestamp_sec": 30.83, "offset_sec": 8, "rgb_path": "assets/cat_15/v_6gZuc4umTPk_685_1045/0008s_00240_rgb.jpg", "mask_path": "assets/cat_15/v_6gZuc4umTPk_685_1045/0008s_00240_mask.png", "mask_status": "available" }, { "frame": "00268", "timestamp_sec": 31.83, "offset_sec": 9, "rgb_path": "assets/cat_15/v_6gZuc4umTPk_685_1045/0009s_00268_rgb.jpg", "mask_path": "assets/cat_15/v_6gZuc4umTPk_685_1045/0009s_00268_mask.png", "mask_status": "available" }, { "frame": "00300", "timestamp_sec": 32.83, "offset_sec": 10, "rgb_path": "assets/cat_15/v_6gZuc4umTPk_685_1045/0010s_00300_rgb.jpg", "mask_path": "assets/cat_15/v_6gZuc4umTPk_685_1045/0010s_00300_mask.png", "mask_status": "available" }, { "frame": "00328", "timestamp_sec": 33.83, "offset_sec": 11, "rgb_path": "assets/cat_15/v_6gZuc4umTPk_685_1045/0011s_00328_rgb.jpg", "mask_path": "assets/cat_15/v_6gZuc4umTPk_685_1045/0011s_00328_mask.png", "mask_status": "available" } ], "obj_id": [ 0 ], "anno_id": [ 945 ], "fps_inferred_from_clip_interval": 30.0 }, "usage": { "input_tokens": 21178, "input_tokens_details": { "cache_write_tokens": 21175, "cached_tokens": 0 }, "output_tokens": 4619, "output_tokens_details": { "reasoning_tokens": 2709 }, "total_tokens": 25797 }, "model": "gpt-5.6-sol", "effort": "medium", "custom_id": "cat15_cap03_v_6gZuc4umTPk_685_1045", "annotation": { "caption": "The person leans forward, holding another person's hand steady with their left hand. They carefully apply the red nail polish to one of the fingers. After finishing, they sit up and place the brush back into the bottle.", "entity_groups": [ { "id": "E1", "label": "selected person", "spans": [ { "text": "The person", "range": [ 0, 10 ] }, { "text": "their", "range": [ 68, 73 ] }, { "text": "They", "range": [ 85, 89 ] }, { "text": "they", "range": [ 166, 170 ] } ], "source_mask_link": "selected_person" }, { "id": "E2", "label": "another person", "spans": [ { "text": "another person's", "range": [ 34, 50 ] } ], "source_mask_link": null }, { "id": "E3", "label": "another person's hand", "spans": [ { "text": "another person's hand", "range": [ 34, 55 ] } ], "source_mask_link": null }, { "id": "E4", "label": "selected person's left hand", "spans": [ { "text": "their left hand", "range": [ 68, 83 ] } ], "source_mask_link": null }, { "id": "E5", "label": "red nail polish", "spans": [ { "text": "the red nail polish", "range": [ 106, 125 ] } ], "source_mask_link": null }, { "id": "E6", "label": "selected finger", "spans": [ { "text": "one", "range": [ 129, 132 ] } ], "source_mask_link": null }, { "id": "E7", "label": "fingers", "spans": [ { "text": "the fingers", "range": [ 136, 147 ] } ], "source_mask_link": null }, { "id": "E8", "label": "brush", "spans": [ { "text": "the brush", "range": [ 188, 197 ] } ], "source_mask_link": null }, { "id": "E9", "label": "bottle", "spans": [ { "text": "the bottle", "range": [ 208, 218 ] } ], "source_mask_link": null } ], "actions": [ { "id": "A1", "mentions": [ { "text": "leans", "ranges": [ [ 11, 16 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "type": "posture", "subject_refs": [ "E1" ], "object_refs": [], "confidence": "high", "temporal_status": "supported", "intervals": [ { "start_s": 0.0, "end_s": 8.0, "source_start_s": 22.83, "source_end_s": 30.83 } ], "evidence": "At clip-relative samples 0–7 seconds, the mask-linked person keeps her upper body bent forward over the other person's hand.", "mask_relation": "subject_matches_selected_person_mask", "source_subject_refs": [ "E1" ] }, { "id": "A2", "mentions": [ { "text": "holding", "ranges": [ [ 26, 33 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "type": "simple", "subject_refs": [ "E1" ], "object_refs": [ "E3" ], "confidence": "high", "temporal_status": "supported", "intervals": [ { "start_s": 0.0, "end_s": 9.0, "source_start_s": 22.83, "source_end_s": 31.83 } ], "evidence": "Across samples 0–8 seconds, the mask-linked person steadies the other person's extended hand while working on its fingers.", "mask_relation": "subject_matches_selected_person_mask", "source_subject_refs": [ "E1" ] }, { "id": "A3", "mentions": [ { "text": "apply", "ranges": [ [ 100, 105 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "type": "simple", "subject_refs": [ "E1" ], "object_refs": [ "E5" ], "confidence": "high", "temporal_status": "supported", "intervals": [ { "start_s": 0.0, "end_s": 8.0, "source_start_s": 22.83, "source_end_s": 30.83 } ], "evidence": "Samples 0–7 seconds show the mask-linked person repeatedly bringing the polish brush into contact with a fingernail; by sample 8 the brush is being lifted away.", "mask_relation": "subject_matches_selected_person_mask", "source_subject_refs": [ "E1" ] }, { "id": "A4", "mentions": [ { "text": "sit up", "ranges": [ [ 171, 177 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "type": "phrasal", "subject_refs": [ "E1" ], "object_refs": [], "confidence": "high", "temporal_status": "supported", "intervals": [ { "start_s": 8.0, "end_s": 10.0, "source_start_s": 30.83, "source_end_s": 32.83 } ], "evidence": "Between samples 8 and 9 seconds, the mask-linked person raises her torso from the earlier forward-leaning position.", "mask_relation": "subject_matches_selected_person_mask", "source_subject_refs": [ "E1" ] }, { "id": "A5", "mentions": [ { "text": "place", "ranges": [ [ 182, 187 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "type": "simple", "subject_refs": [ "E1" ], "object_refs": [ "E8" ], "confidence": "high", "temporal_status": "supported", "intervals": [ { "start_s": 9.0, "end_s": 12.0, "source_start_s": 31.83, "source_end_s": 34.83 } ], "evidence": "Samples 9–11 seconds show the brush and black cap being lowered to the open red bottle at the bottom of the frame; support continues through the clip boundary.", "mask_relation": "subject_matches_selected_person_mask", "source_subject_refs": [ "E1" ] } ], "notes": [ "The supplied mask track identifies the central person described by “The person,” “their,” “They,” and “they.”", "Only the selected person has a supplied mask. The other person, both hands, the fingers, nail polish, brush, and bottle have no independent source masks, so their source-mask links are null.", "The selected person's manipulating hand is partially missed or occluded in portions of the binary masks, but the clean-frame sequence links its movements to the mask-tracked central person.", "“finishing” is omitted as an aspectual completion predicate framing the nail-polish application rather than a separate physical action span.", "Directional or manner modifiers are excluded from action spans: “forward,” “steady,” “carefully,” and “back.”" ] }, "model_output_raw": { "caption": "The person leans forward, holding another person's hand steady with their left hand. They carefully apply the red nail polish to one of the fingers. After finishing, they sit up and place the brush back into the bottle.", "entity_groups": [ { "id": "E1", "entity": "selected person", "spans": [ { "text": "The person", "range": [ 0, 10 ] }, { "text": "their", "range": [ 68, 73 ] }, { "text": "They", "range": [ 85, 89 ] }, { "text": "they", "range": [ 166, 170 ] } ], "source_mask": "selected_person_mask" }, { "id": "E2", "entity": "another person", "spans": [ { "text": "another person's", "range": [ 34, 50 ] } ], "source_mask": null }, { "id": "E3", "entity": "another person's hand", "spans": [ { "text": "another person's hand", "range": [ 34, 55 ] } ], "source_mask": null }, { "id": "E4", "entity": "selected person's left hand", "spans": [ { "text": "their left hand", "range": [ 68, 83 ] } ], "source_mask": null }, { "id": "E5", "entity": "red nail polish", "spans": [ { "text": "the red nail polish", "range": [ 106, 125 ] } ], "source_mask": null }, { "id": "E6", "entity": "selected finger", "spans": [ { "text": "one", "range": [ 129, 132 ] } ], "source_mask": null }, { "id": "E7", "entity": "fingers", "spans": [ { "text": "the fingers", "range": [ 136, 147 ] } ], "source_mask": null }, { "id": "E8", "entity": "brush", "spans": [ { "text": "the brush", "range": [ 188, 197 ] } ], "source_mask": null }, { "id": "E9", "entity": "bottle", "spans": [ { "text": "the bottle", "range": [ 208, 218 ] } ], "source_mask": null } ], "actions": [ { "id": "A1", "mentions": [ { "text": "leans", "occurrence": 1, "ranges": [ [ 11, 16 ] ], "particle": { "text": null, "occurrence": 1 } } ], "type": "posture", "subject_refs": [ "E1" ], "object_refs": [], "confidence": "high", "temporal_status": "supported", "intervals": [ { "start_s": 0.0, "end_s": 8.0, "source_start_s": 22.83, "source_end_s": 30.83 } ], "evidence": "At clip-relative samples 0–7 seconds, the mask-linked person keeps her upper body bent forward over the other person's hand.", "mask_relation": "subject_matches_selected_person_mask" }, { "id": "A2", "mentions": [ { "text": "holding", "occurrence": 1, "ranges": [ [ 26, 33 ] ], "particle": { "text": null, "occurrence": 1 } } ], "type": "simple", "subject_refs": [ "E1" ], "object_refs": [ "E3" ], "instrument_refs": [ "E4" ], "confidence": "high", "temporal_status": "supported", "intervals": [ { "start_s": 0.0, "end_s": 9.0, "source_start_s": 22.83, "source_end_s": 31.83 } ], "evidence": "Across samples 0–8 seconds, the mask-linked person steadies the other person's extended hand while working on its fingers.", "mask_relation": "subject_matches_selected_person_mask" }, { "id": "A3", "mentions": [ { "text": "apply", "occurrence": 1, "ranges": [ [ 100, 105 ] ], "particle": { "text": null, "occurrence": 1 } } ], "type": "simple", "subject_refs": [ "E1" ], "object_refs": [ "E5" ], "target_refs": [ "E6", "E7" ], "confidence": "high", "temporal_status": "supported", "intervals": [ { "start_s": 0.0, "end_s": 8.0, "source_start_s": 22.83, "source_end_s": 30.83 } ], "evidence": "Samples 0–7 seconds show the mask-linked person repeatedly bringing the polish brush into contact with a fingernail; by sample 8 the brush is being lifted away.", "mask_relation": "subject_matches_selected_person_mask" }, { "id": "A4", "mentions": [ { "text": "sit up", "occurrence": 1, "ranges": [ [ 171, 177 ] ], "particle": { "text": null, "occurrence": 1 } } ], "type": "phrasal", "subject_refs": [ "E1" ], "object_refs": [], "confidence": "high", "temporal_status": "supported", "intervals": [ { "start_s": 8.0, "end_s": 10.0, "source_start_s": 30.83, "source_end_s": 32.83 } ], "evidence": "Between samples 8 and 9 seconds, the mask-linked person raises her torso from the earlier forward-leaning position.", "mask_relation": "subject_matches_selected_person_mask" }, { "id": "A5", "mentions": [ { "text": "place", "occurrence": 1, "ranges": [ [ 182, 187 ] ], "particle": { "text": null, "occurrence": 1 } } ], "type": "simple", "subject_refs": [ "E1" ], "object_refs": [ "E8" ], "target_refs": [ "E9" ], "confidence": "high", "temporal_status": "supported", "intervals": [ { "start_s": 9.0, "end_s": 12.0, "source_start_s": 31.83, "source_end_s": 34.83 } ], "evidence": "Samples 9–11 seconds show the brush and black cap being lowered to the open red bottle at the bottom of the frame; support continues through the clip boundary.", "mask_relation": "subject_matches_selected_person_mask" } ], "notes": [ "The supplied mask track identifies the central person described by “The person,” “their,” “They,” and “they.”", "Only the selected person has a supplied mask. The other person, both hands, the fingers, nail polish, brush, and bottle have no independent source masks, so their source-mask links are null.", "The selected person's manipulating hand is partially missed or occluded in portions of the binary masks, but the clean-frame sequence links its movements to the mask-tracked central person.", "“finishing” is omitted as an aspectual completion predicate framing the nail-polish application rather than a separate physical action span.", "Directional or manner modifiers are excluded from action spans: “forward,” “steady,” “carefully,” and “back.”" ] }, "normalization_repairs": 0 }, { "source_video_id": "v_6gZuc4umTPk", "caption_count_category": 15, "clip": { "caption_index": 4, "caption": "The person screws the cap back onto the red nail polish bottle and sets it down. They then pick up a clear bottle, inspect it, and set it back down. Finally, they pick up a yellow nail polish bottle.", "source_interval_sec": [ 34.83, 44.33 ], "loom_clip_id": "v_6gZuc4umTPk_1045_1330", "clip_frame_interval_4fps": [ 1045, 1330 ], "sample_count_1fps": 10, "sampled_frames": [ { "frame": "00000", "timestamp_sec": 34.83, "offset_sec": 0, "rgb_path": "assets/cat_15/v_6gZuc4umTPk_1045_1330/0000s_00000_rgb.jpg", "mask_path": "assets/cat_15/v_6gZuc4umTPk_1045_1330/0000s_00000_mask.png", "mask_status": "available" }, { "frame": "00028", "timestamp_sec": 35.83, "offset_sec": 1, "rgb_path": "assets/cat_15/v_6gZuc4umTPk_1045_1330/0001s_00028_rgb.jpg", "mask_path": "assets/cat_15/v_6gZuc4umTPk_1045_1330/0001s_00028_mask.png", "mask_status": "available" }, { "frame": "00060", "timestamp_sec": 36.83, "offset_sec": 2, "rgb_path": "assets/cat_15/v_6gZuc4umTPk_1045_1330/0002s_00060_rgb.jpg", "mask_path": "assets/cat_15/v_6gZuc4umTPk_1045_1330/0002s_00060_mask.png", "mask_status": "available" }, { "frame": "00088", "timestamp_sec": 37.83, "offset_sec": 3, "rgb_path": "assets/cat_15/v_6gZuc4umTPk_1045_1330/0003s_00088_rgb.jpg", "mask_path": "assets/cat_15/v_6gZuc4umTPk_1045_1330/0003s_00088_mask.png", "mask_status": "available" }, { "frame": "00120", "timestamp_sec": 38.83, "offset_sec": 4, "rgb_path": "assets/cat_15/v_6gZuc4umTPk_1045_1330/0004s_00120_rgb.jpg", "mask_path": "assets/cat_15/v_6gZuc4umTPk_1045_1330/0004s_00120_mask.png", "mask_status": "available" }, { "frame": "00148", "timestamp_sec": 39.83, "offset_sec": 5, "rgb_path": "assets/cat_15/v_6gZuc4umTPk_1045_1330/0005s_00148_rgb.jpg", "mask_path": "assets/cat_15/v_6gZuc4umTPk_1045_1330/0005s_00148_mask.png", "mask_status": "available" }, { "frame": "00180", "timestamp_sec": 40.83, "offset_sec": 6, "rgb_path": "assets/cat_15/v_6gZuc4umTPk_1045_1330/0006s_00180_rgb.jpg", "mask_path": "assets/cat_15/v_6gZuc4umTPk_1045_1330/0006s_00180_mask.png", "mask_status": "available" }, { "frame": "00208", "timestamp_sec": 41.83, "offset_sec": 7, "rgb_path": "assets/cat_15/v_6gZuc4umTPk_1045_1330/0007s_00208_rgb.jpg", "mask_path": "assets/cat_15/v_6gZuc4umTPk_1045_1330/0007s_00208_mask.png", "mask_status": "available" }, { "frame": "00240", "timestamp_sec": 42.83, "offset_sec": 8, "rgb_path": "assets/cat_15/v_6gZuc4umTPk_1045_1330/0008s_00240_rgb.jpg", "mask_path": "assets/cat_15/v_6gZuc4umTPk_1045_1330/0008s_00240_mask.png", "mask_status": "available" }, { "frame": "00268", "timestamp_sec": 43.83, "offset_sec": 9, "rgb_path": "assets/cat_15/v_6gZuc4umTPk_1045_1330/0009s_00268_rgb.jpg", "mask_path": "assets/cat_15/v_6gZuc4umTPk_1045_1330/0009s_00268_mask.png", "mask_status": "available" } ], "obj_id": [ 0 ], "anno_id": [ 931 ], "fps_inferred_from_clip_interval": 30.0 }, "usage": { "input_tokens": 20205, "input_tokens_details": { "cache_write_tokens": 20202, "cached_tokens": 0 }, "output_tokens": 4424, "output_tokens_details": { "reasoning_tokens": 2588 }, "total_tokens": 24629 }, "model": "gpt-5.6-sol", "effort": "medium", "custom_id": "cat15_cap04_v_6gZuc4umTPk_1045_1330", "annotation": { "caption": "The person screws the cap back onto the red nail polish bottle and sets it down. They then pick up a clear bottle, inspect it, and set it back down. Finally, they pick up a yellow nail polish bottle.", "entity_groups": [ { "id": "person_1", "label": "person", "spans": [ { "text": "The person", "range": [ 0, 10 ] }, { "text": "They", "range": [ 81, 85 ] }, { "text": "they", "range": [ 158, 162 ] } ], "source_mask_link": "selected_person" }, { "id": "cap_1", "label": "cap", "spans": [ { "text": "the cap", "range": [ 18, 25 ] } ], "source_mask_link": null }, { "id": "red_nail_polish_bottle_1", "label": "red nail polish bottle", "spans": [ { "text": "the red nail polish bottle", "range": [ 36, 62 ] }, { "text": "it", "range": [ 72, 74 ] } ], "source_mask_link": null }, { "id": "clear_bottle_1", "label": "clear bottle", "spans": [ { "text": "a clear bottle", "range": [ 99, 113 ] }, { "text": "it", "range": [ 123, 125 ] }, { "text": "it", "range": [ 135, 137 ] } ], "source_mask_link": null }, { "id": "yellow_nail_polish_bottle_1", "label": "yellow nail polish bottle", "spans": [ { "text": "a yellow nail polish bottle", "range": [ 171, 198 ] } ], "source_mask_link": null } ], "actions": [ { "id": "action_1", "mentions": [ { "text": "screws", "ranges": [ [ 11, 17 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "type": "prepositional", "subject_refs": [ "person_1" ], "object_refs": [ "cap_1", "red_nail_polish_bottle_1" ], "confidence": "high", "temporal_status": "supported", "intervals": [ { "start_s": 1.0, "end_s": 4.0, "source_start_s": 35.83, "source_end_s": 38.83 } ], "evidence": "At clip-relative 1–3 seconds, the mask-linked person holds the red bottle and repeatedly twists its cap with both hands.", "mask_relation": "subject matches the supplied selected-person mask", "source_subject_refs": "person_1" }, { "id": "action_2", "mentions": [ { "text": "sets", "ranges": [ [ 67, 71 ] ], "occurrence": 1, "particle": { "text": "down", "occurrence": 1, "range": [ 75, 79 ] } } ], "type": "phrasal", "subject_refs": [ "person_1" ], "object_refs": [ "red_nail_polish_bottle_1" ], "confidence": "medium", "temporal_status": "supported", "intervals": [ { "start_s": 4.0, "end_s": 5.0, "source_start_s": 38.83, "source_end_s": 39.83 } ], "evidence": "At 4 seconds, the mask-linked person's arm moves the red bottle downward toward the blue surface.", "mask_relation": "subject matches the supplied selected-person mask", "source_subject_refs": "person_1" }, { "id": "action_3", "mentions": [ { "text": "pick up", "ranges": [ [ 91, 98 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "type": "phrasal", "subject_refs": [ "person_1" ], "object_refs": [ "clear_bottle_1" ], "confidence": "medium", "temporal_status": "supported", "intervals": [ { "start_s": 5.0, "end_s": 7.0, "source_start_s": 39.83, "source_end_s": 41.83 } ], "evidence": "At 5–6 seconds, a clear bottle is in front of the mask-linked person and her hand closes around and lifts it.", "mask_relation": "subject matches the supplied selected-person mask", "source_subject_refs": "person_1" }, { "id": "action_4", "mentions": [ { "text": "inspect", "ranges": [ [ 115, 122 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "type": "simple", "subject_refs": [ "person_1" ], "object_refs": [ "clear_bottle_1" ], "confidence": "medium", "temporal_status": "supported", "intervals": [ { "start_s": 6.0, "end_s": 7.0, "source_start_s": 40.83, "source_end_s": 41.83 } ], "evidence": "At 6 seconds, the mask-linked person holds the clear bottle upright in front of herself where it can be examined.", "mask_relation": "subject matches the supplied selected-person mask", "source_subject_refs": "person_1" }, { "id": "action_5", "mentions": [ { "text": "set", "ranges": [ [ 131, 134 ] ], "occurrence": 1, "particle": { "text": "down", "occurrence": 2, "range": [ 143, 147 ] } } ], "type": "phrasal", "subject_refs": [ "person_1" ], "object_refs": [ "clear_bottle_1" ], "confidence": "medium", "temporal_status": "supported", "intervals": [ { "start_s": 7.0, "end_s": 8.0, "source_start_s": 41.83, "source_end_s": 42.83 } ], "evidence": "At 7 seconds, the mask-linked person's arm moves across and downward over the surface as the clear bottle is returned.", "mask_relation": "subject matches the supplied selected-person mask", "source_subject_refs": "person_1" }, { "id": "action_6", "mentions": [ { "text": "pick up", "ranges": [ [ 163, 170 ] ], "occurrence": 2, "particle": { "text": null, "occurrence": 1 } } ], "type": "phrasal", "subject_refs": [ "person_1" ], "object_refs": [ "yellow_nail_polish_bottle_1" ], "confidence": "medium", "temporal_status": "supported", "intervals": [ { "start_s": 8.0, "end_s": 9.5, "source_start_s": 42.83, "source_end_s": 44.33 } ], "evidence": "At 8 seconds, the mask-linked person grips a small warm-colored nail-polish bottle; her arm continues the lifting motion through the final sampled frame.", "mask_relation": "subject matches the supplied selected-person mask", "source_subject_refs": "person_1" } ], "notes": [ "The supplied mask consistently identifies the brunette person on the left, who performs the captioned bottle-handling actions.", "The blonde person on the right is not linked to the supplied mask and is not mentioned as a separate entity in the caption.", "No source masks were supplied for the cap or any bottle.", "Bottle colors and the clear bottle are difficult to distinguish in some low-resolution or occluded frames, so the corresponding temporal localization is conservative." ] }, "model_output_raw": { "caption": "The person screws the cap back onto the red nail polish bottle and sets it down. They then pick up a clear bottle, inspect it, and set it back down. Finally, they pick up a yellow nail polish bottle.", "entity_groups": [ { "id": "person_1", "label": "person", "spans": [ { "text": "The person", "range": [ 0, 10 ] }, { "text": "They", "range": [ 81, 85 ] }, { "text": "they", "range": [ 158, 162 ] } ], "source_mask_link": "selected_person_masklet" }, { "id": "cap_1", "label": "cap", "spans": [ { "text": "the cap", "range": [ 18, 25 ] } ], "source_mask_link": null }, { "id": "red_nail_polish_bottle_1", "label": "red nail polish bottle", "spans": [ { "text": "the red nail polish bottle", "range": [ 36, 62 ] }, { "text": "it", "range": [ 72, 74 ] } ], "source_mask_link": null }, { "id": "clear_bottle_1", "label": "clear bottle", "spans": [ { "text": "a clear bottle", "range": [ 99, 113 ] }, { "text": "it", "range": [ 123, 125 ] }, { "text": "it", "range": [ 135, 137 ] } ], "source_mask_link": null }, { "id": "yellow_nail_polish_bottle_1", "label": "yellow nail polish bottle", "spans": [ { "text": "a yellow nail polish bottle", "range": [ 171, 198 ] } ], "source_mask_link": null } ], "actions": [ { "id": "action_1", "mentions": [ { "text": "screws", "occurrence": 1, "ranges": [ [ 11, 17 ] ], "particle": { "text": null, "occurrence": 1 } } ], "type": "prepositional", "subject_ref": "person_1", "object_refs": [ "cap_1", "red_nail_polish_bottle_1" ], "confidence": "high", "temporal_status": "supported", "intervals": [ { "start_s": 1.0, "end_s": 4.0 } ], "evidence": "At clip-relative 1–3 seconds, the mask-linked person holds the red bottle and repeatedly twists its cap with both hands.", "mask_relation": "subject matches the supplied selected-person mask" }, { "id": "action_2", "mentions": [ { "text": "sets", "occurrence": 1, "ranges": [ [ 67, 71 ] ], "particle": { "text": "down", "occurrence": 1, "range": [ 75, 79 ] } } ], "type": "phrasal", "subject_ref": "person_1", "object_refs": [ "red_nail_polish_bottle_1" ], "confidence": "medium", "temporal_status": "supported", "intervals": [ { "start_s": 4.0, "end_s": 5.0 } ], "evidence": "At 4 seconds, the mask-linked person's arm moves the red bottle downward toward the blue surface.", "mask_relation": "subject matches the supplied selected-person mask" }, { "id": "action_3", "mentions": [ { "text": "pick up", "occurrence": 1, "ranges": [ [ 91, 98 ] ], "particle": { "text": null, "occurrence": 1 } } ], "type": "phrasal", "subject_ref": "person_1", "object_refs": [ "clear_bottle_1" ], "confidence": "medium", "temporal_status": "supported", "intervals": [ { "start_s": 5.0, "end_s": 7.0 } ], "evidence": "At 5–6 seconds, a clear bottle is in front of the mask-linked person and her hand closes around and lifts it.", "mask_relation": "subject matches the supplied selected-person mask" }, { "id": "action_4", "mentions": [ { "text": "inspect", "occurrence": 1, "ranges": [ [ 115, 122 ] ], "particle": { "text": null, "occurrence": 1 } } ], "type": "simple", "subject_ref": "person_1", "object_refs": [ "clear_bottle_1" ], "confidence": "medium", "temporal_status": "supported", "intervals": [ { "start_s": 6.0, "end_s": 7.0 } ], "evidence": "At 6 seconds, the mask-linked person holds the clear bottle upright in front of herself where it can be examined.", "mask_relation": "subject matches the supplied selected-person mask" }, { "id": "action_5", "mentions": [ { "text": "set", "occurrence": 1, "ranges": [ [ 131, 134 ] ], "particle": { "text": "down", "occurrence": 2, "range": [ 143, 147 ] } } ], "type": "phrasal", "subject_ref": "person_1", "object_refs": [ "clear_bottle_1" ], "confidence": "medium", "temporal_status": "supported", "intervals": [ { "start_s": 7.0, "end_s": 8.0 } ], "evidence": "At 7 seconds, the mask-linked person's arm moves across and downward over the surface as the clear bottle is returned.", "mask_relation": "subject matches the supplied selected-person mask" }, { "id": "action_6", "mentions": [ { "text": "pick up", "occurrence": 2, "ranges": [ [ 163, 170 ] ], "particle": { "text": null, "occurrence": 1 } } ], "type": "phrasal", "subject_ref": "person_1", "object_refs": [ "yellow_nail_polish_bottle_1" ], "confidence": "medium", "temporal_status": "supported", "intervals": [ { "start_s": 8.0, "end_s": 9.5 } ], "evidence": "At 8 seconds, the mask-linked person grips a small warm-colored nail-polish bottle; her arm continues the lifting motion through the final sampled frame.", "mask_relation": "subject matches the supplied selected-person mask" } ], "notes": [ "The supplied mask consistently identifies the brunette person on the left, who performs the captioned bottle-handling actions.", "The blonde person on the right is not linked to the supplied mask and is not mentioned as a separate entity in the caption.", "No source masks were supplied for the cap or any bottle.", "Bottle colors and the clear bottle are difficult to distinguish in some low-resolution or occluded frames, so the corresponding temporal localization is conservative." ] }, "normalization_repairs": 0 }, { "source_video_id": "v_6gZuc4umTPk", "caption_count_category": 15, "clip": { "caption_index": 5, "caption": "The person opens the yellow nail polish bottle and holds the brush. They then take hold of another person's finger and carefully apply the yellow polish to the nail.", "source_interval_sec": [ 44.33, 64.33 ], "loom_clip_id": "v_6gZuc4umTPk_1330_1930", "clip_frame_interval_4fps": [ 1330, 1930 ], "sample_count_1fps": 20, "sampled_frames": [ { "frame": "00000", "timestamp_sec": 44.33, "offset_sec": 0, "rgb_path": "assets/cat_15/v_6gZuc4umTPk_1330_1930/0000s_00000_rgb.jpg", "mask_path": "assets/cat_15/v_6gZuc4umTPk_1330_1930/0000s_00000_mask.png", "mask_status": "available" }, { "frame": "00028", "timestamp_sec": 45.33, "offset_sec": 1, "rgb_path": "assets/cat_15/v_6gZuc4umTPk_1330_1930/0001s_00028_rgb.jpg", "mask_path": "assets/cat_15/v_6gZuc4umTPk_1330_1930/0001s_00028_mask.png", "mask_status": "available" }, { "frame": "00060", "timestamp_sec": 46.33, "offset_sec": 2, "rgb_path": "assets/cat_15/v_6gZuc4umTPk_1330_1930/0002s_00060_rgb.jpg", "mask_path": "assets/cat_15/v_6gZuc4umTPk_1330_1930/0002s_00060_mask.png", "mask_status": "available" }, { "frame": "00088", "timestamp_sec": 47.33, "offset_sec": 3, "rgb_path": "assets/cat_15/v_6gZuc4umTPk_1330_1930/0003s_00088_rgb.jpg", "mask_path": "assets/cat_15/v_6gZuc4umTPk_1330_1930/0003s_00088_mask.png", "mask_status": "available" }, { "frame": "00120", "timestamp_sec": 48.33, "offset_sec": 4, "rgb_path": "assets/cat_15/v_6gZuc4umTPk_1330_1930/0004s_00120_rgb.jpg", "mask_path": "assets/cat_15/v_6gZuc4umTPk_1330_1930/0004s_00120_mask.png", "mask_status": "available" }, { "frame": "00148", "timestamp_sec": 49.33, "offset_sec": 5, "rgb_path": "assets/cat_15/v_6gZuc4umTPk_1330_1930/0005s_00148_rgb.jpg", "mask_path": "assets/cat_15/v_6gZuc4umTPk_1330_1930/0005s_00148_mask.png", "mask_status": "available" }, { "frame": "00180", "timestamp_sec": 50.33, "offset_sec": 6, "rgb_path": "assets/cat_15/v_6gZuc4umTPk_1330_1930/0006s_00180_rgb.jpg", "mask_path": "assets/cat_15/v_6gZuc4umTPk_1330_1930/0006s_00180_mask.png", "mask_status": "available" }, { "frame": "00208", "timestamp_sec": 51.33, "offset_sec": 7, "rgb_path": "assets/cat_15/v_6gZuc4umTPk_1330_1930/0007s_00208_rgb.jpg", "mask_path": "assets/cat_15/v_6gZuc4umTPk_1330_1930/0007s_00208_mask.png", "mask_status": "available" }, { "frame": "00240", "timestamp_sec": 52.33, "offset_sec": 8, "rgb_path": "assets/cat_15/v_6gZuc4umTPk_1330_1930/0008s_00240_rgb.jpg", "mask_path": "assets/cat_15/v_6gZuc4umTPk_1330_1930/0008s_00240_mask.png", "mask_status": "available" }, { "frame": "00268", "timestamp_sec": 53.33, "offset_sec": 9, "rgb_path": "assets/cat_15/v_6gZuc4umTPk_1330_1930/0009s_00268_rgb.jpg", "mask_path": "assets/cat_15/v_6gZuc4umTPk_1330_1930/0009s_00268_mask.png", "mask_status": "available" }, { "frame": "00300", "timestamp_sec": 54.33, "offset_sec": 10, "rgb_path": "assets/cat_15/v_6gZuc4umTPk_1330_1930/0010s_00300_rgb.jpg", "mask_path": "assets/cat_15/v_6gZuc4umTPk_1330_1930/0010s_00300_mask.png", "mask_status": "available" }, { "frame": "00328", "timestamp_sec": 55.33, "offset_sec": 11, "rgb_path": "assets/cat_15/v_6gZuc4umTPk_1330_1930/0011s_00328_rgb.jpg", "mask_path": "assets/cat_15/v_6gZuc4umTPk_1330_1930/0011s_00328_mask.png", "mask_status": "available" }, { "frame": "00360", "timestamp_sec": 56.33, "offset_sec": 12, "rgb_path": "assets/cat_15/v_6gZuc4umTPk_1330_1930/0012s_00360_rgb.jpg", "mask_path": "assets/cat_15/v_6gZuc4umTPk_1330_1930/0012s_00360_mask.png", "mask_status": "available" }, { "frame": "00388", "timestamp_sec": 57.33, "offset_sec": 13, "rgb_path": "assets/cat_15/v_6gZuc4umTPk_1330_1930/0013s_00388_rgb.jpg", "mask_path": "assets/cat_15/v_6gZuc4umTPk_1330_1930/0013s_00388_mask.png", "mask_status": "available" }, { "frame": "00420", "timestamp_sec": 58.33, "offset_sec": 14, "rgb_path": "assets/cat_15/v_6gZuc4umTPk_1330_1930/0014s_00420_rgb.jpg", "mask_path": "assets/cat_15/v_6gZuc4umTPk_1330_1930/0014s_00420_mask.png", "mask_status": "available" }, { "frame": "00448", "timestamp_sec": 59.33, "offset_sec": 15, "rgb_path": "assets/cat_15/v_6gZuc4umTPk_1330_1930/0015s_00448_rgb.jpg", "mask_path": "assets/cat_15/v_6gZuc4umTPk_1330_1930/0015s_00448_mask.png", "mask_status": "available" }, { "frame": "00480", "timestamp_sec": 60.33, "offset_sec": 16, "rgb_path": "assets/cat_15/v_6gZuc4umTPk_1330_1930/0016s_00480_rgb.jpg", "mask_path": "assets/cat_15/v_6gZuc4umTPk_1330_1930/0016s_00480_mask.png", "mask_status": "available" }, { "frame": "00508", "timestamp_sec": 61.33, "offset_sec": 17, "rgb_path": "assets/cat_15/v_6gZuc4umTPk_1330_1930/0017s_00508_rgb.jpg", "mask_path": "assets/cat_15/v_6gZuc4umTPk_1330_1930/0017s_00508_mask.png", "mask_status": "available" }, { "frame": "00540", "timestamp_sec": 62.33, "offset_sec": 18, "rgb_path": "assets/cat_15/v_6gZuc4umTPk_1330_1930/0018s_00540_rgb.jpg", "mask_path": "assets/cat_15/v_6gZuc4umTPk_1330_1930/0018s_00540_mask.png", "mask_status": "available" }, { "frame": "00568", "timestamp_sec": 63.33, "offset_sec": 19, "rgb_path": "assets/cat_15/v_6gZuc4umTPk_1330_1930/0019s_00568_rgb.jpg", "mask_path": "assets/cat_15/v_6gZuc4umTPk_1330_1930/0019s_00568_mask.png", "mask_status": "available" } ], "obj_id": [ 0 ], "anno_id": [ 932 ], "fps_inferred_from_clip_interval": 30.0 }, "usage": { "input_tokens": 25061, "input_tokens_details": { "cache_write_tokens": 25058, "cached_tokens": 0 }, "output_tokens": 4635, "output_tokens_details": { "reasoning_tokens": 2875 }, "total_tokens": 29696 }, "model": "gpt-5.6-sol", "effort": "medium", "custom_id": "cat15_cap05_v_6gZuc4umTPk_1330_1930", "annotation": { "caption": "The person opens the yellow nail polish bottle and holds the brush. They then take hold of another person's finger and carefully apply the yellow polish to the nail.", "entity_groups": [ { "id": "person_1", "label": "person", "spans": [ { "text": "The person", "range": [ 0, 10 ] }, { "text": "They", "range": [ 68, 72 ] } ], "source_mask_link": "selected_person" }, { "id": "bottle_1", "label": "nail polish bottle", "spans": [ { "text": "the yellow nail polish bottle", "range": [ 17, 46 ] } ], "source_mask_link": null }, { "id": "brush_1", "label": "brush", "spans": [ { "text": "the brush", "range": [ 57, 66 ] } ], "source_mask_link": null }, { "id": "person_2", "label": "person", "spans": [ { "text": "another person's", "range": [ 91, 107 ] } ], "source_mask_link": null }, { "id": "finger_1", "label": "finger", "spans": [ { "text": "another person's finger", "range": [ 91, 114 ] } ], "source_mask_link": null }, { "id": "polish_1", "label": "nail polish", "spans": [ { "text": "the yellow polish", "range": [ 135, 152 ] } ], "source_mask_link": null }, { "id": "nail_1", "label": "nail", "spans": [ { "text": "the nail", "range": [ 156, 164 ] } ], "source_mask_link": null } ], "actions": [ { "id": "action_1", "mentions": [ { "text": "opens", "ranges": [ [ 11, 16 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "type": "simple", "subject_refs": [ "person_1" ], "object_refs": [], "confidence": "medium", "temporal_status": "localized", "intervals": [ { "start_s": 0.0, "end_s": 2.0, "source_start_s": 44.33, "source_end_s": 46.33 } ], "evidence": "At 0–1 seconds, the mask-linked brunette manipulates and twists the yellow bottle or cap in her hands.", "mask_relation": "The action subject matches the supplied selected-person mask.", "source_subject_refs": "person_1" }, { "id": "action_2", "mentions": [ { "text": "holds", "ranges": [ [ 51, 56 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "type": "simple", "subject_refs": [ "person_1" ], "object_refs": [], "confidence": "high", "temporal_status": "localized", "intervals": [ { "start_s": 5.0, "end_s": 20.0, "source_start_s": 49.33, "source_end_s": 64.33 } ], "evidence": "From 5 seconds onward, the mask-linked person visibly holds the yellow-handled polish brush, including through the final sampled frame.", "mask_relation": "The action subject matches the supplied selected-person mask.", "source_subject_refs": "person_1" }, { "id": "action_3", "mentions": [ { "text": "take hold", "ranges": [ [ 78, 87 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "type": "light_verb", "subject_refs": [ "person_1" ], "object_refs": [], "confidence": "medium", "temporal_status": "localized", "intervals": [ { "start_s": 2.0, "end_s": 4.0, "source_start_s": 46.33, "source_end_s": 48.33 }, { "start_s": 11.0, "end_s": 20.0, "source_start_s": 55.33, "source_end_s": 64.33 } ], "evidence": "At 2–3 seconds the mask-linked person reaches for and steadies the other person's hand; from 11 seconds onward she visibly grips and positions a finger while painting.", "mask_relation": "The action subject matches the supplied selected-person mask.", "source_subject_refs": "person_1" }, { "id": "action_4", "mentions": [ { "text": "apply", "ranges": [ [ 129, 134 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "type": "simple", "subject_refs": [ "person_1" ], "object_refs": [], "confidence": "high", "temporal_status": "localized", "intervals": [ { "start_s": 13.0, "end_s": 20.0, "source_start_s": 57.33, "source_end_s": 64.33 } ], "evidence": "At 13–19 seconds, the mask-linked person repeatedly brings the yellow-handled brush into contact with the held fingernail; the action remains visible in the final frame.", "mask_relation": "The action subject matches the supplied selected-person mask.", "source_subject_refs": "person_1" } ], "notes": [ "The supplied mask consistently identifies the brunette person on the left/center, who is the caption's acting person; it does not identify the blonde person on the right.", "Only the acting person has a supplied masklet. The bottle, brush, other person, finger, polish, and nail are text-groundable entities but have no separate source masks.", "The adverbs \"then\" and \"carefully\" are excluded from the action spans.", "Intervals are sparse visual-support intervals derived from the sampled clean RGB frames; intervals ending at 20.0 seconds reach the original clip boundary because the corresponding actions remain visible at the final 19-second sample." ] }, "model_output_raw": { "caption": "The person opens the yellow nail polish bottle and holds the brush. They then take hold of another person's finger and carefully apply the yellow polish to the nail.", "entity_groups": [ { "id": "person_1", "entity": "person", "spans": [ { "start": 0, "end": 10, "text": "The person" }, { "start": 68, "end": 72, "text": "They" } ], "source_mask": "supplied_selected_person_masklet" }, { "id": "bottle_1", "entity": "nail polish bottle", "spans": [ { "start": 17, "end": 46, "text": "the yellow nail polish bottle" } ], "source_mask": null }, { "id": "brush_1", "entity": "brush", "spans": [ { "start": 57, "end": 66, "text": "the brush" } ], "source_mask": null }, { "id": "person_2", "entity": "person", "spans": [ { "start": 91, "end": 107, "text": "another person's" } ], "source_mask": null }, { "id": "finger_1", "entity": "finger", "spans": [ { "start": 91, "end": 114, "text": "another person's finger" } ], "source_mask": null }, { "id": "polish_1", "entity": "nail polish", "spans": [ { "start": 135, "end": 152, "text": "the yellow polish" } ], "source_mask": null }, { "id": "nail_1", "entity": "nail", "spans": [ { "start": 156, "end": 164, "text": "the nail" } ], "source_mask": null } ], "actions": [ { "mentions": [ { "text": "opens", "ranges": [ { "start": 11, "end": 16 } ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "type": "simple", "subject_entity_group": "person_1", "object_entity_groups": [ { "entity_group": "bottle_1", "role": "object" } ], "confidence": "medium", "temporal_status": "localized", "intervals": [ { "start_s": 0.0, "end_s": 2.0, "source_start_s": 44.33, "source_end_s": 46.33 } ], "evidence": "At 0–1 seconds, the mask-linked brunette manipulates and twists the yellow bottle or cap in her hands.", "mask_relation": "The action subject matches the supplied selected-person mask." }, { "mentions": [ { "text": "holds", "ranges": [ { "start": 51, "end": 56 } ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "type": "simple", "subject_entity_group": "person_1", "object_entity_groups": [ { "entity_group": "brush_1", "role": "object" } ], "confidence": "high", "temporal_status": "localized", "intervals": [ { "start_s": 5.0, "end_s": 20.0, "source_start_s": 49.33, "source_end_s": 64.33 } ], "evidence": "From 5 seconds onward, the mask-linked person visibly holds the yellow-handled polish brush, including through the final sampled frame.", "mask_relation": "The action subject matches the supplied selected-person mask." }, { "mentions": [ { "text": "take hold", "ranges": [ { "start": 78, "end": 87 } ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "type": "light_verb", "subject_entity_group": "person_1", "object_entity_groups": [ { "entity_group": "finger_1", "role": "object" }, { "entity_group": "person_2", "role": "possessor_of_object" } ], "confidence": "medium", "temporal_status": "localized", "intervals": [ { "start_s": 2.0, "end_s": 4.0, "source_start_s": 46.33, "source_end_s": 48.33 }, { "start_s": 11.0, "end_s": 20.0, "source_start_s": 55.33, "source_end_s": 64.33 } ], "evidence": "At 2–3 seconds the mask-linked person reaches for and steadies the other person's hand; from 11 seconds onward she visibly grips and positions a finger while painting.", "mask_relation": "The action subject matches the supplied selected-person mask." }, { "mentions": [ { "text": "apply", "ranges": [ { "start": 129, "end": 134 } ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "type": "simple", "subject_entity_group": "person_1", "object_entity_groups": [ { "entity_group": "polish_1", "role": "substance" }, { "entity_group": "nail_1", "role": "target" } ], "confidence": "high", "temporal_status": "localized", "intervals": [ { "start_s": 13.0, "end_s": 20.0, "source_start_s": 57.33, "source_end_s": 64.33 } ], "evidence": "At 13–19 seconds, the mask-linked person repeatedly brings the yellow-handled brush into contact with the held fingernail; the action remains visible in the final frame.", "mask_relation": "The action subject matches the supplied selected-person mask." } ], "notes": [ "The supplied mask consistently identifies the brunette person on the left/center, who is the caption's acting person; it does not identify the blonde person on the right.", "Only the acting person has a supplied masklet. The bottle, brush, other person, finger, polish, and nail are text-groundable entities but have no separate source masks.", "The adverbs \"then\" and \"carefully\" are excluded from the action spans.", "Intervals are sparse visual-support intervals derived from the sampled clean RGB frames; intervals ending at 20.0 seconds reach the original clip boundary because the corresponding actions remain visible at the final 19-second sample." ] }, "normalization_repairs": 0 }, { "source_video_id": "v_6gZuc4umTPk", "caption_count_category": 15, "clip": { "caption_index": 6, "caption": "The person finishes applying the yellow polish, screws the cap back on the bottle, and sets it aside. They then pull the hood of their hoodie over their head.", "source_interval_sec": [ 64.33, 74.83 ], "loom_clip_id": "v_6gZuc4umTPk_1930_2245", "clip_frame_interval_4fps": [ 1930, 2245 ], "sample_count_1fps": 11, "sampled_frames": [ { "frame": "00000", "timestamp_sec": 64.33, "offset_sec": 0, "rgb_path": "assets/cat_15/v_6gZuc4umTPk_1930_2245/0000s_00000_rgb.jpg", "mask_path": "assets/cat_15/v_6gZuc4umTPk_1930_2245/0000s_00000_mask.png", "mask_status": "available" }, { "frame": "00028", "timestamp_sec": 65.33, "offset_sec": 1, "rgb_path": "assets/cat_15/v_6gZuc4umTPk_1930_2245/0001s_00028_rgb.jpg", "mask_path": "assets/cat_15/v_6gZuc4umTPk_1930_2245/0001s_00028_mask.png", "mask_status": "available" }, { "frame": "00060", "timestamp_sec": 66.33, "offset_sec": 2, "rgb_path": "assets/cat_15/v_6gZuc4umTPk_1930_2245/0002s_00060_rgb.jpg", "mask_path": "assets/cat_15/v_6gZuc4umTPk_1930_2245/0002s_00060_mask.png", "mask_status": "available" }, { "frame": "00088", "timestamp_sec": 67.33, "offset_sec": 3, "rgb_path": "assets/cat_15/v_6gZuc4umTPk_1930_2245/0003s_00088_rgb.jpg", "mask_path": "assets/cat_15/v_6gZuc4umTPk_1930_2245/0003s_00088_mask.png", "mask_status": "available" }, { "frame": "00120", "timestamp_sec": 68.33, "offset_sec": 4, "rgb_path": "assets/cat_15/v_6gZuc4umTPk_1930_2245/0004s_00120_rgb.jpg", "mask_path": "assets/cat_15/v_6gZuc4umTPk_1930_2245/0004s_00120_mask.png", "mask_status": "available" }, { "frame": "00148", "timestamp_sec": 69.33, "offset_sec": 5, "rgb_path": "assets/cat_15/v_6gZuc4umTPk_1930_2245/0005s_00148_rgb.jpg", "mask_path": "assets/cat_15/v_6gZuc4umTPk_1930_2245/0005s_00148_mask.png", "mask_status": "available" }, { "frame": "00180", "timestamp_sec": 70.33, "offset_sec": 6, "rgb_path": "assets/cat_15/v_6gZuc4umTPk_1930_2245/0006s_00180_rgb.jpg", "mask_path": "assets/cat_15/v_6gZuc4umTPk_1930_2245/0006s_00180_mask.png", "mask_status": "available" }, { "frame": "00208", "timestamp_sec": 71.33, "offset_sec": 7, "rgb_path": "assets/cat_15/v_6gZuc4umTPk_1930_2245/0007s_00208_rgb.jpg", "mask_path": "assets/cat_15/v_6gZuc4umTPk_1930_2245/0007s_00208_mask.png", "mask_status": "available" }, { "frame": "00240", "timestamp_sec": 72.33, "offset_sec": 8, "rgb_path": "assets/cat_15/v_6gZuc4umTPk_1930_2245/0008s_00240_rgb.jpg", "mask_path": "assets/cat_15/v_6gZuc4umTPk_1930_2245/0008s_00240_mask.png", "mask_status": "available" }, { "frame": "00268", "timestamp_sec": 73.33, "offset_sec": 9, "rgb_path": "assets/cat_15/v_6gZuc4umTPk_1930_2245/0009s_00268_rgb.jpg", "mask_path": "assets/cat_15/v_6gZuc4umTPk_1930_2245/0009s_00268_mask.png", "mask_status": "available" }, { "frame": "00300", "timestamp_sec": 74.33, "offset_sec": 10, "rgb_path": "assets/cat_15/v_6gZuc4umTPk_1930_2245/0010s_00300_rgb.jpg", "mask_path": "assets/cat_15/v_6gZuc4umTPk_1930_2245/0010s_00300_mask.png", "mask_status": "available" } ], "obj_id": [ 0 ], "anno_id": [ 933 ], "fps_inferred_from_clip_interval": 30.0 }, "usage": { "input_tokens": 20680, "input_tokens_details": { "cache_write_tokens": 20677, "cached_tokens": 0 }, "output_tokens": 4244, "output_tokens_details": { "reasoning_tokens": 2588 }, "total_tokens": 24924 }, "model": "gpt-5.6-sol", "effort": "medium", "custom_id": "cat15_cap06_v_6gZuc4umTPk_1930_2245", "annotation": { "caption": "The person finishes applying the yellow polish, screws the cap back on the bottle, and sets it aside. They then pull the hood of their hoodie over their head.", "entity_groups": [ { "id": "person_1", "label": "person", "spans": [ { "text": "The person", "range": [ 0, 10 ] }, { "text": "They", "range": [ 102, 106 ] }, { "text": "their", "range": [ 129, 134 ] }, { "text": "their", "range": [ 147, 152 ] } ], "source_mask_link": "selected_person" }, { "id": "polish_1", "label": "polish", "spans": [ { "text": "the yellow polish", "range": [ 29, 46 ] } ], "source_mask_link": null }, { "id": "cap_1", "label": "cap", "spans": [ { "text": "the cap", "range": [ 55, 62 ] } ], "source_mask_link": null }, { "id": "bottle_1", "label": "bottle", "spans": [ { "text": "the bottle", "range": [ 71, 81 ] }, { "text": "it", "range": [ 92, 94 ] } ], "source_mask_link": null }, { "id": "hood_1", "label": "hood", "spans": [ { "text": "the hood", "range": [ 117, 125 ] } ], "source_mask_link": null }, { "id": "hoodie_1", "label": "hoodie", "spans": [ { "text": "their hoodie", "range": [ 129, 141 ] } ], "source_mask_link": null }, { "id": "head_1", "label": "head", "spans": [ { "text": "their head", "range": [ 147, 157 ] } ], "source_mask_link": null } ], "actions": [ { "id": "action_1", "mentions": [ { "text": "applying", "ranges": [ [ 20, 28 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "type": "simple", "subject_refs": [ "person_1" ], "object_refs": [], "confidence": "medium", "temporal_status": "not_confidently_localized", "intervals": [], "evidence": "At the earliest samples the masked person manipulates hands near the tabletop, but the actual application of polish is not clearly visible.", "mask_relation": "The action subject matches the supplied selected-person mask.", "source_subject_refs": "person_1" }, { "id": "action_2", "mentions": [ { "text": "screws", "ranges": [ [ 48, 54 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "type": "simple", "subject_refs": [ "person_1" ], "object_refs": [], "confidence": "high", "temporal_status": "localized", "intervals": [ { "start_s": 3.0, "end_s": 7.0, "source_start_s": 67.33, "source_end_s": 71.33 } ], "evidence": "From 3.0 through 6.0 seconds, the masked person holds the yellow container in both hands and makes repeated twisting motions.", "mask_relation": "The action subject matches the supplied selected-person mask.", "source_subject_refs": "person_1" }, { "id": "action_3", "mentions": [ { "text": "sets", "ranges": [ [ 87, 91 ] ], "occurrence": 1, "particle": { "text": "aside", "occurrence": 1, "range": [ 95, 100 ] } } ], "type": "phrasal", "subject_refs": [ "person_1" ], "object_refs": [], "confidence": "medium", "temporal_status": "localized", "intervals": [ { "start_s": 7.0, "end_s": 8.0, "source_start_s": 71.33, "source_end_s": 72.33 } ], "evidence": "At 7.0 seconds, the masked person's arm sweeps across the upper frame as the handled yellow item leaves view.", "mask_relation": "The action subject matches the supplied selected-person mask.", "source_subject_refs": "person_1" }, { "id": "action_4", "mentions": [ { "text": "pull", "ranges": [ [ 112, 116 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "type": "simple", "subject_refs": [ "person_1" ], "object_refs": [], "confidence": "high", "temporal_status": "localized", "intervals": [ { "start_s": 8.0, "end_s": 10.5, "source_start_s": 72.33, "source_end_s": 74.83 } ], "evidence": "From 8.0 seconds through the clip boundary, the masked person's patterned hood rises and covers the person's head.", "mask_relation": "The action subject matches the supplied selected-person mask.", "source_subject_refs": "person_1" } ], "notes": [ "The supplied mask consistently identifies the dark-haired person on the left, who performs all captioned actions.", "The blonde person on the right is visible but is not mentioned by the caption and is not represented by the supplied mask.", "No separate source masks were supplied for the polish, cap, bottle, hood, hoodie, or head.", "The aspectual verb \"finishes\" is excluded from the applying-action span, and \"back on\" is excluded from the screwing span as an adverb plus preposition." ] }, "model_output_raw": { "caption": "The person finishes applying the yellow polish, screws the cap back on the bottle, and sets it aside. They then pull the hood of their hoodie over their head.", "entity_groups": [ { "id": "person_1", "label": "person", "spans": [ { "text": "The person", "range": [ 0, 10 ] }, { "text": "They", "range": [ 102, 106 ] }, { "text": "their", "range": [ 129, 134 ] }, { "text": "their", "range": [ 147, 152 ] } ], "source_mask_link": { "mask_id": "selected_person", "description": "The supplied masks track the dark-haired person on the left wearing the patterned hoodie." } }, { "id": "polish_1", "label": "polish", "spans": [ { "text": "the yellow polish", "range": [ 29, 46 ] } ], "source_mask_link": null }, { "id": "cap_1", "label": "cap", "spans": [ { "text": "the cap", "range": [ 55, 62 ] } ], "source_mask_link": null }, { "id": "bottle_1", "label": "bottle", "spans": [ { "text": "the bottle", "range": [ 71, 81 ] }, { "text": "it", "range": [ 92, 94 ] } ], "source_mask_link": null }, { "id": "hood_1", "label": "hood", "spans": [ { "text": "the hood", "range": [ 117, 125 ] } ], "source_mask_link": null }, { "id": "hoodie_1", "label": "hoodie", "spans": [ { "text": "their hoodie", "range": [ 129, 141 ] } ], "source_mask_link": null }, { "id": "head_1", "label": "head", "spans": [ { "text": "their head", "range": [ 147, 157 ] } ], "source_mask_link": null } ], "actions": [ { "id": "action_1", "mentions": [ { "text": "applying", "occurrence": 1, "ranges": [ [ 20, 28 ] ], "particle": { "text": null, "occurrence": 1 } } ], "type": "simple", "subject_ref": "person_1", "object_refs": [ { "entity_group": "polish_1", "role": "theme" } ], "confidence": "medium", "mask_relation": "The action subject matches the supplied selected-person mask.", "temporal": { "status": "not_confidently_localized", "intervals": [] }, "evidence": "At the earliest samples the masked person manipulates hands near the tabletop, but the actual application of polish is not clearly visible." }, { "id": "action_2", "mentions": [ { "text": "screws", "occurrence": 1, "ranges": [ [ 48, 54 ] ], "particle": { "text": null, "occurrence": 1 } } ], "type": "simple", "subject_ref": "person_1", "object_refs": [ { "entity_group": "cap_1", "role": "theme" }, { "entity_group": "bottle_1", "role": "target" } ], "confidence": "high", "mask_relation": "The action subject matches the supplied selected-person mask.", "temporal": { "status": "localized", "intervals": [ { "start_s": 3.0, "end_s": 7.0 } ] }, "evidence": "From 3.0 through 6.0 seconds, the masked person holds the yellow container in both hands and makes repeated twisting motions." }, { "id": "action_3", "mentions": [ { "text": "sets", "occurrence": 1, "ranges": [ [ 87, 91 ] ], "particle": { "text": "aside", "occurrence": 1, "range": [ 95, 100 ] } } ], "type": "phrasal", "subject_ref": "person_1", "object_refs": [ { "entity_group": "bottle_1", "role": "theme" } ], "confidence": "medium", "mask_relation": "The action subject matches the supplied selected-person mask.", "temporal": { "status": "localized", "intervals": [ { "start_s": 7.0, "end_s": 8.0 } ] }, "evidence": "At 7.0 seconds, the masked person's arm sweeps across the upper frame as the handled yellow item leaves view." }, { "id": "action_4", "mentions": [ { "text": "pull", "occurrence": 1, "ranges": [ [ 112, 116 ] ], "particle": { "text": null, "occurrence": 1 } } ], "type": "simple", "subject_ref": "person_1", "object_refs": [ { "entity_group": "hood_1", "role": "theme" }, { "entity_group": "hoodie_1", "role": "whole" }, { "entity_group": "head_1", "role": "path_target" } ], "confidence": "high", "mask_relation": "The action subject matches the supplied selected-person mask.", "temporal": { "status": "localized", "intervals": [ { "start_s": 8.0, "end_s": 10.5 } ] }, "evidence": "From 8.0 seconds through the clip boundary, the masked person's patterned hood rises and covers the person's head." } ], "notes": [ "The supplied mask consistently identifies the dark-haired person on the left, who performs all captioned actions.", "The blonde person on the right is visible but is not mentioned by the caption and is not represented by the supplied mask.", "No separate source masks were supplied for the polish, cap, bottle, hood, hoodie, or head.", "The aspectual verb \"finishes\" is excluded from the applying-action span, and \"back on\" is excluded from the screwing span as an adverb plus preposition." ] }, "normalization_repairs": 0 }, { "source_video_id": "v_6gZuc4umTPk", "caption_count_category": 15, "clip": { "caption_index": 7, "caption": "The person claps their hands twice, picks up a black cup with their right hand, and taps it on the table. They continue a rhythmic sequence of clapping and tapping the cup.", "source_interval_sec": [ 74.83, 85.83 ], "loom_clip_id": "v_6gZuc4umTPk_2245_2575", "clip_frame_interval_4fps": [ 2245, 2575 ], "sample_count_1fps": 11, "sampled_frames": [ { "frame": "00000", "timestamp_sec": 74.83, "offset_sec": 0, "rgb_path": "assets/cat_15/v_6gZuc4umTPk_2245_2575/0000s_00000_rgb.jpg", "mask_path": "assets/cat_15/v_6gZuc4umTPk_2245_2575/0000s_00000_mask.png", "mask_status": "available" }, { "frame": "00028", "timestamp_sec": 75.83, "offset_sec": 1, "rgb_path": "assets/cat_15/v_6gZuc4umTPk_2245_2575/0001s_00028_rgb.jpg", "mask_path": "assets/cat_15/v_6gZuc4umTPk_2245_2575/0001s_00028_mask.png", "mask_status": "available" }, { "frame": "00060", "timestamp_sec": 76.83, "offset_sec": 2, "rgb_path": "assets/cat_15/v_6gZuc4umTPk_2245_2575/0002s_00060_rgb.jpg", "mask_path": "assets/cat_15/v_6gZuc4umTPk_2245_2575/0002s_00060_mask.png", "mask_status": "available" }, { "frame": "00088", "timestamp_sec": 77.83, "offset_sec": 3, "rgb_path": "assets/cat_15/v_6gZuc4umTPk_2245_2575/0003s_00088_rgb.jpg", "mask_path": "assets/cat_15/v_6gZuc4umTPk_2245_2575/0003s_00088_mask.png", "mask_status": "available" }, { "frame": "00120", "timestamp_sec": 78.83, "offset_sec": 4, "rgb_path": "assets/cat_15/v_6gZuc4umTPk_2245_2575/0004s_00120_rgb.jpg", "mask_path": "assets/cat_15/v_6gZuc4umTPk_2245_2575/0004s_00120_mask.png", "mask_status": "available" }, { "frame": "00148", "timestamp_sec": 79.83, "offset_sec": 5, "rgb_path": "assets/cat_15/v_6gZuc4umTPk_2245_2575/0005s_00148_rgb.jpg", "mask_path": "assets/cat_15/v_6gZuc4umTPk_2245_2575/0005s_00148_mask.png", "mask_status": "available" }, { "frame": "00180", "timestamp_sec": 80.83, "offset_sec": 6, "rgb_path": "assets/cat_15/v_6gZuc4umTPk_2245_2575/0006s_00180_rgb.jpg", "mask_path": "assets/cat_15/v_6gZuc4umTPk_2245_2575/0006s_00180_mask.png", "mask_status": "available" }, { "frame": "00208", "timestamp_sec": 81.83, "offset_sec": 7, "rgb_path": "assets/cat_15/v_6gZuc4umTPk_2245_2575/0007s_00208_rgb.jpg", "mask_path": "assets/cat_15/v_6gZuc4umTPk_2245_2575/0007s_00208_mask.png", "mask_status": "available" }, { "frame": "00240", "timestamp_sec": 82.83, "offset_sec": 8, "rgb_path": "assets/cat_15/v_6gZuc4umTPk_2245_2575/0008s_00240_rgb.jpg", "mask_path": "assets/cat_15/v_6gZuc4umTPk_2245_2575/0008s_00240_mask.png", "mask_status": "available" }, { "frame": "00268", "timestamp_sec": 83.83, "offset_sec": 9, "rgb_path": "assets/cat_15/v_6gZuc4umTPk_2245_2575/0009s_00268_rgb.jpg", "mask_path": "assets/cat_15/v_6gZuc4umTPk_2245_2575/0009s_00268_mask.png", "mask_status": "available" }, { "frame": "00300", "timestamp_sec": 84.83, "offset_sec": 10, "rgb_path": "assets/cat_15/v_6gZuc4umTPk_2245_2575/0010s_00300_rgb.jpg", "mask_path": "assets/cat_15/v_6gZuc4umTPk_2245_2575/0010s_00300_mask.png", "mask_status": "available" } ], "obj_id": [ 0 ], "anno_id": [ 935 ], "fps_inferred_from_clip_interval": 30.0 }, "usage": { "input_tokens": 20685, "input_tokens_details": { "cache_write_tokens": 20682, "cached_tokens": 0 }, "output_tokens": 4504, "output_tokens_details": { "reasoning_tokens": 3106 }, "total_tokens": 25189 }, "model": "gpt-5.6-sol", "effort": "medium", "custom_id": "cat15_cap07_v_6gZuc4umTPk_2245_2575", "annotation": { "caption": "The person claps their hands twice, picks up a black cup with their right hand, and taps it on the table. They continue a rhythmic sequence of clapping and tapping the cup.", "entity_groups": [ { "id": "person_1", "label": "person", "spans": [ { "text": "The person", "range": [ 0, 10 ] }, { "text": "their", "range": [ 17, 22 ] }, { "text": "their", "range": [ 62, 67 ] }, { "text": "They", "range": [ 106, 110 ] } ], "source_mask_link": "selected_person" }, { "id": "hands_1", "label": "hands", "spans": [ { "text": "their hands", "range": [ 17, 28 ] } ], "source_mask_link": null }, { "id": "cup_1", "label": "cup", "spans": [ { "text": "a black cup", "range": [ 45, 56 ] }, { "text": "it", "range": [ 89, 91 ] }, { "text": "the cup", "range": [ 164, 171 ] } ], "source_mask_link": null }, { "id": "right_hand_1", "label": "hand", "spans": [ { "text": "their right hand", "range": [ 62, 78 ] } ], "source_mask_link": null }, { "id": "table_1", "label": "table", "spans": [ { "text": "the table", "range": [ 95, 104 ] } ], "source_mask_link": null } ], "actions": [ { "id": "action_1", "mentions": [ { "text": "claps", "ranges": [ [ 11, 16 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } }, { "text": "clapping", "ranges": [ [ 143, 151 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "type": "simple", "subject_refs": [ "person_1" ], "object_refs": [ "hands_1" ], "confidence": "high", "temporal_status": "supported", "intervals": [ { "start_s": 7.0, "end_s": 11.0, "source_start_s": 81.83, "source_end_s": 85.83 } ], "evidence": "At clip-relative 7.0–10.0 seconds (source 81.830–84.830), the masked person repeatedly brings or moves both hands together with visible hand blur. Support reaches the clip boundary, so the interval ends at 11.0 seconds.", "mask_relation": "The action subject matches the supplied selected-person mask.", "source_subject_refs": [ "person_1" ] }, { "id": "action_2", "mentions": [ { "text": "picks up", "ranges": [ [ 36, 44 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "type": "phrasal", "subject_refs": [ "person_1" ], "object_refs": [ "cup_1" ], "confidence": "high", "temporal_status": "uncertain", "intervals": [], "evidence": "The black cup appears by 2.0 seconds near the masked person's moving arm, but no sampled clean frame clearly shows the cup lifted from the table.", "mask_relation": "The action subject matches the supplied selected-person mask.", "source_subject_refs": [ "person_1" ] }, { "id": "action_3", "mentions": [ { "text": "taps", "ranges": [ [ 84, 88 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } }, { "text": "tapping", "ranges": [ [ 156, 163 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "type": "simple", "subject_refs": [ "person_1" ], "object_refs": [ "cup_1" ], "confidence": "high", "temporal_status": "supported", "intervals": [ { "start_s": 4.0, "end_s": 7.0, "source_start_s": 78.83, "source_end_s": 81.83 } ], "evidence": "At clip-relative 4.0–6.0 seconds (source 78.830–80.830), the masked person's hand is beside or contacting the upright cup, with visible motion around the cup and tabletop.", "mask_relation": "The action subject matches the supplied selected-person mask.", "source_subject_refs": [ "person_1" ] } ], "notes": [ "The supplied mask track identifies the patterned-shirt person on the left, who is the caption's person and the subject of all annotated actions.", "The other visible child is not mentioned by the caption and therefore has no entity group.", "No separate masks were supplied for the hands, right hand, cup, or table; their source-mask links are null.", "The later words “clapping” and “tapping” are grouped with the earlier predicates because the second sentence explicitly describes continuation of the same rhythmic action sequence.", "The abstract phrase “a rhythmic sequence” is not a concrete entity, and “continue” is an excluded aspectual predicate." ] }, "model_output_raw": { "caption": "The person claps their hands twice, picks up a black cup with their right hand, and taps it on the table. They continue a rhythmic sequence of clapping and tapping the cup.", "entity_groups": [ { "id": "person_1", "entity_type": "person", "spans": [ { "text": "The person", "range": [ 0, 10 ] }, { "text": "their", "range": [ 17, 22 ] }, { "text": "their", "range": [ 62, 67 ] }, { "text": "They", "range": [ 106, 110 ] } ], "source_mask_link": { "track_id": "selected_person", "sample_ids": [ 1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11 ] } }, { "id": "hands_1", "entity_type": "hands", "spans": [ { "text": "their hands", "range": [ 17, 28 ] } ], "source_mask_link": null }, { "id": "cup_1", "entity_type": "cup", "spans": [ { "text": "a black cup", "range": [ 45, 56 ] }, { "text": "it", "range": [ 89, 91 ] }, { "text": "the cup", "range": [ 164, 171 ] } ], "source_mask_link": null }, { "id": "right_hand_1", "entity_type": "hand", "spans": [ { "text": "their right hand", "range": [ 62, 78 ] } ], "source_mask_link": null }, { "id": "table_1", "entity_type": "table", "spans": [ { "text": "the table", "range": [ 95, 104 ] } ], "source_mask_link": null } ], "actions": [ { "id": "action_1", "mentions": [ { "text": "claps", "ranges": [ [ 11, 16 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } }, { "text": "clapping", "ranges": [ [ 143, 151 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "type": "simple", "subject_refs": [ "person_1" ], "object_refs": [ "hands_1" ], "confidence": "high", "mask_relation": "The action subject matches the supplied selected-person mask.", "temporal_status": "supported", "intervals": [ { "start_s": 7.0, "end_s": 11.0 } ], "evidence": "At clip-relative 7.0–10.0 seconds (source 81.830–84.830), the masked person repeatedly brings or moves both hands together with visible hand blur. Support reaches the clip boundary, so the interval ends at 11.0 seconds." }, { "id": "action_2", "mentions": [ { "text": "picks up", "ranges": [ [ 36, 44 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "type": "phrasal", "subject_refs": [ "person_1" ], "object_refs": [ "cup_1" ], "instrument_refs": [ "right_hand_1" ], "confidence": "high", "mask_relation": "The action subject matches the supplied selected-person mask.", "temporal_status": "uncertain", "intervals": [], "evidence": "The black cup appears by 2.0 seconds near the masked person's moving arm, but no sampled clean frame clearly shows the cup lifted from the table." }, { "id": "action_3", "mentions": [ { "text": "taps", "ranges": [ [ 84, 88 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } }, { "text": "tapping", "ranges": [ [ 156, 163 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "type": "simple", "subject_refs": [ "person_1" ], "object_refs": [ "cup_1" ], "target_refs": [ "table_1" ], "confidence": "high", "mask_relation": "The action subject matches the supplied selected-person mask.", "temporal_status": "supported", "intervals": [ { "start_s": 4.0, "end_s": 7.0 } ], "evidence": "At clip-relative 4.0–6.0 seconds (source 78.830–80.830), the masked person's hand is beside or contacting the upright cup, with visible motion around the cup and tabletop." } ], "notes": [ "The supplied mask track identifies the patterned-shirt person on the left, who is the caption's person and the subject of all annotated actions.", "The other visible child is not mentioned by the caption and therefore has no entity group.", "No separate masks were supplied for the hands, right hand, cup, or table; their source-mask links are null.", "The later words “clapping” and “tapping” are grouped with the earlier predicates because the second sentence explicitly describes continuation of the same rhythmic action sequence.", "The abstract phrase “a rhythmic sequence” is not a concrete entity, and “continue” is an excluded aspectual predicate." ] }, "normalization_repairs": 0 }, { "source_video_id": "v_6gZuc4umTPk", "caption_count_category": 15, "clip": { "caption_index": 8, "caption": "The person puts down the cup and picks up a red marker, which they uncap. After holding it for a moment, they recap the marker and set it down.", "source_interval_sec": [ 85.83, 108.83 ], "loom_clip_id": "v_6gZuc4umTPk_2575_3265", "clip_frame_interval_4fps": [ 2575, 3265 ], "sample_count_1fps": 23, "sampled_frames": [ { "frame": "00000", "timestamp_sec": 85.83, "offset_sec": 0, "rgb_path": "assets/cat_15/v_6gZuc4umTPk_2575_3265/0000s_00000_rgb.jpg", "mask_path": "assets/cat_15/v_6gZuc4umTPk_2575_3265/0000s_00000_mask.png", "mask_status": "available" }, { "frame": "00028", "timestamp_sec": 86.83, "offset_sec": 1, "rgb_path": "assets/cat_15/v_6gZuc4umTPk_2575_3265/0001s_00028_rgb.jpg", "mask_path": "assets/cat_15/v_6gZuc4umTPk_2575_3265/0001s_00028_mask.png", "mask_status": "available" }, { "frame": "00060", "timestamp_sec": 87.83, "offset_sec": 2, "rgb_path": "assets/cat_15/v_6gZuc4umTPk_2575_3265/0002s_00060_rgb.jpg", "mask_path": "assets/cat_15/v_6gZuc4umTPk_2575_3265/0002s_00060_mask.png", "mask_status": "available" }, { "frame": "00088", "timestamp_sec": 88.83, "offset_sec": 3, "rgb_path": "assets/cat_15/v_6gZuc4umTPk_2575_3265/0003s_00088_rgb.jpg", "mask_path": "assets/cat_15/v_6gZuc4umTPk_2575_3265/0003s_00088_mask.png", "mask_status": "available" }, { "frame": "00120", "timestamp_sec": 89.83, "offset_sec": 4, "rgb_path": "assets/cat_15/v_6gZuc4umTPk_2575_3265/0004s_00120_rgb.jpg", "mask_path": "assets/cat_15/v_6gZuc4umTPk_2575_3265/0004s_00120_mask.png", "mask_status": "available" }, { "frame": "00148", "timestamp_sec": 90.83, "offset_sec": 5, "rgb_path": "assets/cat_15/v_6gZuc4umTPk_2575_3265/0005s_00148_rgb.jpg", "mask_path": "assets/cat_15/v_6gZuc4umTPk_2575_3265/0005s_00148_mask.png", "mask_status": "available" }, { "frame": "00180", "timestamp_sec": 91.83, "offset_sec": 6, "rgb_path": "assets/cat_15/v_6gZuc4umTPk_2575_3265/0006s_00180_rgb.jpg", "mask_path": "assets/cat_15/v_6gZuc4umTPk_2575_3265/0006s_00180_mask.png", "mask_status": "available" }, { "frame": "00208", "timestamp_sec": 92.83, "offset_sec": 7, "rgb_path": "assets/cat_15/v_6gZuc4umTPk_2575_3265/0007s_00208_rgb.jpg", "mask_path": "assets/cat_15/v_6gZuc4umTPk_2575_3265/0007s_00208_mask.png", "mask_status": "available" }, { "frame": "00240", "timestamp_sec": 93.83, "offset_sec": 8, "rgb_path": "assets/cat_15/v_6gZuc4umTPk_2575_3265/0008s_00240_rgb.jpg", "mask_path": "assets/cat_15/v_6gZuc4umTPk_2575_3265/0008s_00240_mask.png", "mask_status": "available" }, { "frame": "00268", "timestamp_sec": 94.83, "offset_sec": 9, "rgb_path": "assets/cat_15/v_6gZuc4umTPk_2575_3265/0009s_00268_rgb.jpg", "mask_path": "assets/cat_15/v_6gZuc4umTPk_2575_3265/0009s_00268_mask.png", "mask_status": "available" }, { "frame": "00300", "timestamp_sec": 95.83, "offset_sec": 10, "rgb_path": "assets/cat_15/v_6gZuc4umTPk_2575_3265/0010s_00300_rgb.jpg", "mask_path": "assets/cat_15/v_6gZuc4umTPk_2575_3265/0010s_00300_mask.png", "mask_status": "available" }, { "frame": "00328", "timestamp_sec": 96.83, "offset_sec": 11, "rgb_path": "assets/cat_15/v_6gZuc4umTPk_2575_3265/0011s_00328_rgb.jpg", "mask_path": "assets/cat_15/v_6gZuc4umTPk_2575_3265/0011s_00328_mask.png", "mask_status": "available" }, { "frame": "00360", "timestamp_sec": 97.83, "offset_sec": 12, "rgb_path": "assets/cat_15/v_6gZuc4umTPk_2575_3265/0012s_00360_rgb.jpg", "mask_path": "assets/cat_15/v_6gZuc4umTPk_2575_3265/0012s_00360_mask.png", "mask_status": "available" }, { "frame": "00388", "timestamp_sec": 98.83, "offset_sec": 13, "rgb_path": "assets/cat_15/v_6gZuc4umTPk_2575_3265/0013s_00388_rgb.jpg", "mask_path": "assets/cat_15/v_6gZuc4umTPk_2575_3265/0013s_00388_mask.png", "mask_status": "available" }, { "frame": "00420", "timestamp_sec": 99.83, "offset_sec": 14, "rgb_path": "assets/cat_15/v_6gZuc4umTPk_2575_3265/0014s_00420_rgb.jpg", "mask_path": "assets/cat_15/v_6gZuc4umTPk_2575_3265/0014s_00420_mask.png", "mask_status": "available" }, { "frame": "00448", "timestamp_sec": 100.83, "offset_sec": 15, "rgb_path": "assets/cat_15/v_6gZuc4umTPk_2575_3265/0015s_00448_rgb.jpg", "mask_path": "assets/cat_15/v_6gZuc4umTPk_2575_3265/0015s_00448_mask.png", "mask_status": "available" }, { "frame": "00480", "timestamp_sec": 101.83, "offset_sec": 16, "rgb_path": "assets/cat_15/v_6gZuc4umTPk_2575_3265/0016s_00480_rgb.jpg", "mask_path": "assets/cat_15/v_6gZuc4umTPk_2575_3265/0016s_00480_mask.png", "mask_status": "available" }, { "frame": "00508", "timestamp_sec": 102.83, "offset_sec": 17, "rgb_path": "assets/cat_15/v_6gZuc4umTPk_2575_3265/0017s_00508_rgb.jpg", "mask_path": "assets/cat_15/v_6gZuc4umTPk_2575_3265/0017s_00508_mask.png", "mask_status": "available" }, { "frame": "00540", "timestamp_sec": 103.83, "offset_sec": 18, "rgb_path": "assets/cat_15/v_6gZuc4umTPk_2575_3265/0018s_00540_rgb.jpg", "mask_path": "assets/cat_15/v_6gZuc4umTPk_2575_3265/0018s_00540_mask.png", "mask_status": "available" }, { "frame": "00568", "timestamp_sec": 104.83, "offset_sec": 19, "rgb_path": "assets/cat_15/v_6gZuc4umTPk_2575_3265/0019s_00568_rgb.jpg", "mask_path": "assets/cat_15/v_6gZuc4umTPk_2575_3265/0019s_00568_mask.png", "mask_status": "available" }, { "frame": "00600", "timestamp_sec": 105.83, "offset_sec": 20, "rgb_path": "assets/cat_15/v_6gZuc4umTPk_2575_3265/0020s_00600_rgb.jpg", "mask_path": "assets/cat_15/v_6gZuc4umTPk_2575_3265/0020s_00600_mask.png", "mask_status": "available" }, { "frame": "00628", "timestamp_sec": 106.83, "offset_sec": 21, "rgb_path": "assets/cat_15/v_6gZuc4umTPk_2575_3265/0021s_00628_rgb.jpg", "mask_path": "assets/cat_15/v_6gZuc4umTPk_2575_3265/0021s_00628_mask.png", "mask_status": "available" }, { "frame": "00660", "timestamp_sec": 107.83, "offset_sec": 22, "rgb_path": "assets/cat_15/v_6gZuc4umTPk_2575_3265/0022s_00660_rgb.jpg", "mask_path": "assets/cat_15/v_6gZuc4umTPk_2575_3265/0022s_00660_mask.png", "mask_status": "available" } ], "obj_id": [ 0 ], "anno_id": [ 936 ], "fps_inferred_from_clip_interval": 30.0 }, "usage": { "input_tokens": 26525, "input_tokens_details": { "cache_write_tokens": 26522, "cached_tokens": 0 }, "output_tokens": 3725, "output_tokens_details": { "reasoning_tokens": 2070 }, "total_tokens": 30250 }, "model": "gpt-5.6-sol", "effort": "medium", "custom_id": "cat15_cap08_v_6gZuc4umTPk_2575_3265", "annotation": { "caption": "The person puts down the cup and picks up a red marker, which they uncap. After holding it for a moment, they recap the marker and set it down.", "entity_groups": [ { "id": "person_1", "label": "person", "spans": [ { "text": "The person", "range": [ 0, 10 ] }, { "text": "they", "range": [ 62, 66 ] }, { "text": "they", "range": [ 105, 109 ] } ], "source_mask_link": "selected_person" }, { "id": "cup_1", "label": "cup", "spans": [ { "text": "the cup", "range": [ 21, 28 ] } ], "source_mask_link": null }, { "id": "marker_1", "label": "marker", "spans": [ { "text": "a red marker", "range": [ 42, 54 ] }, { "text": "which", "range": [ 56, 61 ] }, { "text": "it", "range": [ 88, 90 ] }, { "text": "the marker", "range": [ 116, 126 ] }, { "text": "it", "range": [ 135, 137 ] } ], "source_mask_link": null } ], "actions": [ { "id": "action_1", "mentions": [ { "text": "puts down", "ranges": [ [ 11, 20 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "type": "phrasal", "subject_refs": [ "person_1" ], "object_refs": [ "cup_1" ], "confidence": "medium", "temporal_status": "supported", "intervals": [ { "start_s": 0.0, "end_s": 1.0, "source_start_s": 85.83, "source_end_s": 86.83 } ], "evidence": "At the first sampled frame, the masked person has a hand beside the upright dark cup-like object on the table; subsequent frames show attention shifting to the marker.", "mask_relation": "subject matches the supplied selected-person mask", "source_subject_refs": "person_1" }, { "id": "action_2", "mentions": [ { "text": "picks up", "ranges": [ [ 33, 41 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "type": "phrasal", "subject_refs": [ "person_1" ], "object_refs": [ "marker_1" ], "confidence": "medium", "temporal_status": "supported", "intervals": [ { "start_s": 1.0, "end_s": 3.0, "source_start_s": 86.83, "source_end_s": 88.83 } ], "evidence": "The masked person raises and holds a marker-like object horizontally in samples at 1 and 2 seconds.", "mask_relation": "subject matches the supplied selected-person mask", "source_subject_refs": "person_1" }, { "id": "action_3", "mentions": [ { "text": "uncap", "ranges": [ [ 67, 72 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "type": "simple", "subject_refs": [ "person_1" ], "object_refs": [ "marker_1" ], "confidence": "medium", "temporal_status": "supported", "intervals": [ { "start_s": 5.0, "end_s": 12.0, "source_start_s": 90.83, "source_end_s": 97.83 } ], "evidence": "Across samples 5–11 seconds, the masked person's hands twist and separate parts of the red marker; by 11 seconds its open end is visible.", "mask_relation": "subject matches the supplied selected-person mask", "source_subject_refs": "person_1" }, { "id": "action_4", "mentions": [ { "text": "holding", "ranges": [ [ 80, 87 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "type": "simple", "subject_refs": [ "person_1" ], "object_refs": [ "marker_1" ], "confidence": "high", "temporal_status": "supported", "intervals": [ { "start_s": 12.0, "end_s": 17.0, "source_start_s": 97.83, "source_end_s": 102.83 } ], "evidence": "The masked person visibly holds the uncapped red marker near the central dark cup-like object throughout samples 12–16 seconds.", "mask_relation": "subject matches the supplied selected-person mask", "source_subject_refs": "person_1" }, { "id": "action_5", "mentions": [ { "text": "recap", "ranges": [ [ 110, 115 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "type": "simple", "subject_refs": [ "person_1" ], "object_refs": [ "marker_1" ], "confidence": "medium", "temporal_status": "supported", "intervals": [ { "start_s": 17.0, "end_s": 22.0, "source_start_s": 102.83, "source_end_s": 107.83 } ], "evidence": "In samples 17–21 seconds, the masked person's hands bring a small cap-like part to the marker and press or twist it into place.", "mask_relation": "subject matches the supplied selected-person mask", "source_subject_refs": "person_1" }, { "id": "action_6", "mentions": [ { "text": "set", "ranges": [ [ 131, 134 ] ], "occurrence": 1, "particle": { "text": "down", "occurrence": 1, "range": [ 138, 142 ] } } ], "type": "phrasal", "subject_refs": [ "person_1" ], "object_refs": [ "marker_1" ], "confidence": "low", "temporal_status": "not_supported_in_sampled_frames", "intervals": [], "evidence": "The final samples still show the masked person handling the marker; no sampled clean frame clearly shows the completed placement.", "mask_relation": "subject matches the supplied selected-person mask", "source_subject_refs": "person_1" } ], "notes": [ "The supplied masklet consistently identifies the older person on the left, who performs the captioned actions.", "A second, younger person is visible on the right but has no caption mention and is not linked to the supplied mask.", "No separate masks were supplied for the cup or marker.", "Temporal intervals are sparse sampled-frame support rather than exact continuous-video action boundaries." ] }, "model_output_raw": { "caption": "The person puts down the cup and picks up a red marker, which they uncap. After holding it for a moment, they recap the marker and set it down.", "entity_groups": [ { "id": "person_1", "entity_type": "person", "spans": [ { "text": "The person", "range": [ 0, 10 ] }, { "text": "they", "range": [ 62, 66 ] }, { "text": "they", "range": [ 105, 109 ] } ], "source_mask": "selected_person_masklet" }, { "id": "cup_1", "entity_type": "cup", "spans": [ { "text": "the cup", "range": [ 21, 28 ] } ], "source_mask": null }, { "id": "marker_1", "entity_type": "marker", "spans": [ { "text": "a red marker", "range": [ 42, 54 ] }, { "text": "which", "range": [ 56, 61 ] }, { "text": "it", "range": [ 88, 90 ] }, { "text": "the marker", "range": [ 116, 126 ] }, { "text": "it", "range": [ 135, 137 ] } ], "source_mask": null } ], "actions": [ { "id": "action_1", "mentions": [ { "text": "puts down", "occurrence": 1, "ranges": [ [ 11, 20 ] ], "particle": { "text": null, "occurrence": 1 } } ], "type": "phrasal", "subject": "person_1", "objects": [ "cup_1" ], "confidence": "medium", "mask_relation": "subject matches the supplied selected-person mask", "temporal_status": "supported", "intervals": [ { "clip_relative_s": [ 0.0, 1.0 ], "source_video_s": [ 85.83, 86.83 ] } ], "evidence": "At the first sampled frame, the masked person has a hand beside the upright dark cup-like object on the table; subsequent frames show attention shifting to the marker." }, { "id": "action_2", "mentions": [ { "text": "picks up", "occurrence": 1, "ranges": [ [ 33, 41 ] ], "particle": { "text": null, "occurrence": 1 } } ], "type": "phrasal", "subject": "person_1", "objects": [ "marker_1" ], "confidence": "medium", "mask_relation": "subject matches the supplied selected-person mask", "temporal_status": "supported", "intervals": [ { "clip_relative_s": [ 1.0, 3.0 ], "source_video_s": [ 86.83, 88.83 ] } ], "evidence": "The masked person raises and holds a marker-like object horizontally in samples at 1 and 2 seconds." }, { "id": "action_3", "mentions": [ { "text": "uncap", "occurrence": 1, "ranges": [ [ 67, 72 ] ], "particle": { "text": null, "occurrence": 1 } } ], "type": "simple", "subject": "person_1", "objects": [ "marker_1" ], "confidence": "medium", "mask_relation": "subject matches the supplied selected-person mask", "temporal_status": "supported", "intervals": [ { "clip_relative_s": [ 5.0, 12.0 ], "source_video_s": [ 90.83, 97.83 ] } ], "evidence": "Across samples 5–11 seconds, the masked person's hands twist and separate parts of the red marker; by 11 seconds its open end is visible." }, { "id": "action_4", "mentions": [ { "text": "holding", "occurrence": 1, "ranges": [ [ 80, 87 ] ], "particle": { "text": null, "occurrence": 1 } } ], "type": "simple", "subject": "person_1", "objects": [ "marker_1" ], "confidence": "high", "mask_relation": "subject matches the supplied selected-person mask", "temporal_status": "supported", "intervals": [ { "clip_relative_s": [ 12.0, 17.0 ], "source_video_s": [ 97.83, 102.83 ] } ], "evidence": "The masked person visibly holds the uncapped red marker near the central dark cup-like object throughout samples 12–16 seconds." }, { "id": "action_5", "mentions": [ { "text": "recap", "occurrence": 1, "ranges": [ [ 110, 115 ] ], "particle": { "text": null, "occurrence": 1 } } ], "type": "simple", "subject": "person_1", "objects": [ "marker_1" ], "confidence": "medium", "mask_relation": "subject matches the supplied selected-person mask", "temporal_status": "supported", "intervals": [ { "clip_relative_s": [ 17.0, 22.0 ], "source_video_s": [ 102.83, 107.83 ] } ], "evidence": "In samples 17–21 seconds, the masked person's hands bring a small cap-like part to the marker and press or twist it into place." }, { "id": "action_6", "mentions": [ { "text": "set", "occurrence": 1, "ranges": [ [ 131, 134 ] ], "particle": { "text": "down", "occurrence": 1, "range": [ 138, 142 ] } } ], "type": "phrasal", "subject": "person_1", "objects": [ "marker_1" ], "confidence": "low", "mask_relation": "subject matches the supplied selected-person mask", "temporal_status": "not_supported_in_sampled_frames", "intervals": [], "evidence": "The final samples still show the masked person handling the marker; no sampled clean frame clearly shows the completed placement." } ], "notes": [ "The supplied masklet consistently identifies the older person on the left, who performs the captioned actions.", "A second, younger person is visible on the right but has no caption mention and is not linked to the supplied mask.", "No separate masks were supplied for the cup or marker.", "Temporal intervals are sparse sampled-frame support rather than exact continuous-video action boundaries." ] }, "normalization_repairs": 0 }, { "source_video_id": "v_6gZuc4umTPk", "caption_count_category": 15, "clip": { "caption_index": 9, "caption": "The person picks up a green nail polish bottle and opens it. They hold the brush up, look at it, and talk while gesturing with the brush.", "source_interval_sec": [ 108.83, 127.33 ], "loom_clip_id": "v_6gZuc4umTPk_3265_3820", "clip_frame_interval_4fps": [ 3265, 3820 ], "sample_count_1fps": 19, "sampled_frames": [ { "frame": "00000", "timestamp_sec": 108.83, "offset_sec": 0, "rgb_path": "assets/cat_15/v_6gZuc4umTPk_3265_3820/0000s_00000_rgb.jpg", "mask_path": "assets/cat_15/v_6gZuc4umTPk_3265_3820/0000s_00000_mask.png", "mask_status": "available" }, { "frame": "00028", "timestamp_sec": 109.83, "offset_sec": 1, "rgb_path": "assets/cat_15/v_6gZuc4umTPk_3265_3820/0001s_00028_rgb.jpg", "mask_path": "assets/cat_15/v_6gZuc4umTPk_3265_3820/0001s_00028_mask.png", "mask_status": "available" }, { "frame": "00060", "timestamp_sec": 110.83, "offset_sec": 2, "rgb_path": "assets/cat_15/v_6gZuc4umTPk_3265_3820/0002s_00060_rgb.jpg", "mask_path": "assets/cat_15/v_6gZuc4umTPk_3265_3820/0002s_00060_mask.png", "mask_status": "available" }, { "frame": "00088", "timestamp_sec": 111.83, "offset_sec": 3, "rgb_path": "assets/cat_15/v_6gZuc4umTPk_3265_3820/0003s_00088_rgb.jpg", "mask_path": "assets/cat_15/v_6gZuc4umTPk_3265_3820/0003s_00088_mask.png", "mask_status": "available" }, { "frame": "00120", "timestamp_sec": 112.83, "offset_sec": 4, "rgb_path": "assets/cat_15/v_6gZuc4umTPk_3265_3820/0004s_00120_rgb.jpg", "mask_path": "assets/cat_15/v_6gZuc4umTPk_3265_3820/0004s_00120_mask.png", "mask_status": "available" }, { "frame": "00148", "timestamp_sec": 113.83, "offset_sec": 5, "rgb_path": "assets/cat_15/v_6gZuc4umTPk_3265_3820/0005s_00148_rgb.jpg", "mask_path": "assets/cat_15/v_6gZuc4umTPk_3265_3820/0005s_00148_mask.png", "mask_status": "available" }, { "frame": "00180", "timestamp_sec": 114.83, "offset_sec": 6, "rgb_path": "assets/cat_15/v_6gZuc4umTPk_3265_3820/0006s_00180_rgb.jpg", "mask_path": "assets/cat_15/v_6gZuc4umTPk_3265_3820/0006s_00180_mask.png", "mask_status": "available" }, { "frame": "00208", "timestamp_sec": 115.83, "offset_sec": 7, "rgb_path": "assets/cat_15/v_6gZuc4umTPk_3265_3820/0007s_00208_rgb.jpg", "mask_path": "assets/cat_15/v_6gZuc4umTPk_3265_3820/0007s_00208_mask.png", "mask_status": "available" }, { "frame": "00240", "timestamp_sec": 116.83, "offset_sec": 8, "rgb_path": "assets/cat_15/v_6gZuc4umTPk_3265_3820/0008s_00240_rgb.jpg", "mask_path": "assets/cat_15/v_6gZuc4umTPk_3265_3820/0008s_00240_mask.png", "mask_status": "available" }, { "frame": "00268", "timestamp_sec": 117.83, "offset_sec": 9, "rgb_path": "assets/cat_15/v_6gZuc4umTPk_3265_3820/0009s_00268_rgb.jpg", "mask_path": "assets/cat_15/v_6gZuc4umTPk_3265_3820/0009s_00268_mask.png", "mask_status": "available" }, { "frame": "00300", "timestamp_sec": 118.83, "offset_sec": 10, "rgb_path": "assets/cat_15/v_6gZuc4umTPk_3265_3820/0010s_00300_rgb.jpg", "mask_path": "assets/cat_15/v_6gZuc4umTPk_3265_3820/0010s_00300_mask.png", "mask_status": "available" }, { "frame": "00328", "timestamp_sec": 119.83, "offset_sec": 11, "rgb_path": "assets/cat_15/v_6gZuc4umTPk_3265_3820/0011s_00328_rgb.jpg", "mask_path": "assets/cat_15/v_6gZuc4umTPk_3265_3820/0011s_00328_mask.png", "mask_status": "available" }, { "frame": "00360", "timestamp_sec": 120.83, "offset_sec": 12, "rgb_path": "assets/cat_15/v_6gZuc4umTPk_3265_3820/0012s_00360_rgb.jpg", "mask_path": "assets/cat_15/v_6gZuc4umTPk_3265_3820/0012s_00360_mask.png", "mask_status": "available" }, { "frame": "00388", "timestamp_sec": 121.83, "offset_sec": 13, "rgb_path": "assets/cat_15/v_6gZuc4umTPk_3265_3820/0013s_00388_rgb.jpg", "mask_path": "assets/cat_15/v_6gZuc4umTPk_3265_3820/0013s_00388_mask.png", "mask_status": "available" }, { "frame": "00420", "timestamp_sec": 122.83, "offset_sec": 14, "rgb_path": "assets/cat_15/v_6gZuc4umTPk_3265_3820/0014s_00420_rgb.jpg", "mask_path": "assets/cat_15/v_6gZuc4umTPk_3265_3820/0014s_00420_mask.png", "mask_status": "available" }, { "frame": "00448", "timestamp_sec": 123.83, "offset_sec": 15, "rgb_path": "assets/cat_15/v_6gZuc4umTPk_3265_3820/0015s_00448_rgb.jpg", "mask_path": "assets/cat_15/v_6gZuc4umTPk_3265_3820/0015s_00448_mask.png", "mask_status": "available" }, { "frame": "00480", "timestamp_sec": 124.83, "offset_sec": 16, "rgb_path": "assets/cat_15/v_6gZuc4umTPk_3265_3820/0016s_00480_rgb.jpg", "mask_path": "assets/cat_15/v_6gZuc4umTPk_3265_3820/0016s_00480_mask.png", "mask_status": "available" }, { "frame": "00508", "timestamp_sec": 125.83, "offset_sec": 17, "rgb_path": "assets/cat_15/v_6gZuc4umTPk_3265_3820/0017s_00508_rgb.jpg", "mask_path": "assets/cat_15/v_6gZuc4umTPk_3265_3820/0017s_00508_mask.png", "mask_status": "available" }, { "frame": "00540", "timestamp_sec": 126.83, "offset_sec": 18, "rgb_path": "assets/cat_15/v_6gZuc4umTPk_3265_3820/0018s_00540_rgb.jpg", "mask_path": "assets/cat_15/v_6gZuc4umTPk_3265_3820/0018s_00540_mask.png", "mask_status": "available" } ], "obj_id": [ 0 ], "anno_id": [ 937 ], "fps_inferred_from_clip_interval": 30.0 }, "usage": { "input_tokens": 24575, "input_tokens_details": { "cache_write_tokens": 24572, "cached_tokens": 0 }, "output_tokens": 3918, "output_tokens_details": { "reasoning_tokens": 2200 }, "total_tokens": 28493 }, "model": "gpt-5.6-sol", "effort": "medium", "custom_id": "cat15_cap09_v_6gZuc4umTPk_3265_3820", "annotation": { "caption": "The person picks up a green nail polish bottle and opens it. They hold the brush up, look at it, and talk while gesturing with the brush.", "entity_groups": [ { "id": "entity_person", "label": "person", "spans": [ { "text": "The person", "range": [ 0, 10 ] }, { "text": "They", "range": [ 61, 65 ] } ], "source_mask_link": "selected_person" }, { "id": "entity_bottle", "label": "nail_polish_bottle", "spans": [ { "text": "a green nail polish bottle", "range": [ 20, 46 ] }, { "text": "it", "range": [ 57, 59 ] } ], "source_mask_link": null }, { "id": "entity_brush", "label": "brush", "spans": [ { "text": "the brush", "range": [ 71, 80 ] }, { "text": "it", "range": [ 93, 95 ] }, { "text": "the brush", "range": [ 127, 136 ] } ], "source_mask_link": null } ], "actions": [ { "id": "action_pick_up", "mentions": [ { "text": "picks up", "ranges": [ [ 11, 19 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "type": "phrasal", "subject_refs": [ "entity_person" ], "object_refs": [ "entity_bottle" ], "confidence": "medium", "temporal_status": "localized_sparse_support", "intervals": [ { "start_s": 10.0, "end_s": 13.0, "source_start_s": 118.83, "source_end_s": 121.83 } ], "evidence": "Across samples 11–13, the masked person grasps and raises the small green bottle into her hands.", "mask_relation": "The action subject matches the supplied selected-person mask.", "source_subject_refs": [ "entity_person" ] }, { "id": "action_open", "mentions": [ { "text": "opens", "ranges": [ [ 51, 56 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "type": "simple", "subject_refs": [ "entity_person" ], "object_refs": [ "entity_bottle" ], "confidence": "high", "temporal_status": "localized_sparse_support", "intervals": [ { "start_s": 13.0, "end_s": 18.0, "source_start_s": 121.83, "source_end_s": 126.83 } ], "evidence": "Samples 14–18 show the masked person holding the bottle and twisting or pulling its cap and brush assembly.", "mask_relation": "The action subject matches the supplied selected-person mask.", "source_subject_refs": [ "entity_person" ] }, { "id": "action_hold", "mentions": [ { "text": "hold", "ranges": [ [ 66, 70 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "type": "simple", "subject_refs": [ "entity_person" ], "object_refs": [ "entity_brush" ], "confidence": "medium", "temporal_status": "localized_sparse_support", "intervals": [ { "start_s": 13.0, "end_s": 18.5, "source_start_s": 121.83, "source_end_s": 127.33 } ], "evidence": "In samples 14–19, the masked person keeps the extracted brush or cap-and-brush component raised in her hand.", "mask_relation": "The action subject matches the supplied selected-person mask.", "source_subject_refs": [ "entity_person" ] }, { "id": "action_look", "mentions": [ { "text": "look", "ranges": [ [ 85, 89 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "type": "prepositional", "subject_refs": [ "entity_person" ], "object_refs": [ "entity_brush" ], "confidence": "medium", "temporal_status": "localized_sparse_support", "intervals": [ { "start_s": 13.0, "end_s": 18.0, "source_start_s": 121.83, "source_end_s": 126.83 } ], "evidence": "Samples 14–18 show the masked person's head and gaze directed toward the bottle and raised brush in her hands.", "mask_relation": "The action subject matches the supplied selected-person mask.", "source_subject_refs": [ "entity_person" ] }, { "id": "action_talk", "mentions": [ { "text": "talk", "ranges": [ [ 101, 105 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "type": "simple", "subject_refs": [ "entity_person" ], "object_refs": [], "confidence": "medium", "temporal_status": "localized_sparse_support", "intervals": [ { "start_s": 13.0, "end_s": 18.5, "source_start_s": 121.83, "source_end_s": 127.33 } ], "evidence": "During samples 14–19, the masked person's mouth repeatedly changes between open and closed positions while she handles the brush.", "mask_relation": "The action subject matches the supplied selected-person mask.", "source_subject_refs": [ "entity_person" ] }, { "id": "action_gesture", "mentions": [ { "text": "gesturing", "ranges": [ [ 112, 121 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "type": "simple", "subject_refs": [ "entity_person" ], "object_refs": [ "entity_brush" ], "confidence": "medium", "temporal_status": "localized_sparse_support", "intervals": [ { "start_s": 15.0, "end_s": 18.5, "source_start_s": 123.83, "source_end_s": 127.33 } ], "evidence": "Samples 16–19 show the masked person moving the raised brush-bearing hand while speaking.", "mask_relation": "The action subject matches the supplied selected-person mask.", "source_subject_refs": [ "entity_person" ] } ], "notes": [ "Only the selected older person has a supplied masklet; the nail-polish bottle and brush have no separate source masks.", "A second child is visible on the right but is not mentioned as a separate caption entity and is not identified by the supplied mask.", "Action intervals are sparse support inferred from the sampled clean RGB frames and are not exact continuous-video boundaries." ] }, "model_output_raw": { "caption": "The person picks up a green nail polish bottle and opens it. They hold the brush up, look at it, and talk while gesturing with the brush.", "entity_groups": [ { "id": "entity_person", "entity_type": "person", "spans": [ { "text": "The person", "range": [ 0, 10 ] }, { "text": "They", "range": [ 61, 65 ] } ], "source_mask_link": { "masklet_id": "selected_person", "description": "The supplied masks consistently identify the older person on the left." } }, { "id": "entity_bottle", "entity_type": "nail_polish_bottle", "spans": [ { "text": "a green nail polish bottle", "range": [ 20, 46 ] }, { "text": "it", "range": [ 57, 59 ] } ], "source_mask_link": null }, { "id": "entity_brush", "entity_type": "brush", "spans": [ { "text": "the brush", "range": [ 71, 80 ] }, { "text": "it", "range": [ 93, 95 ] }, { "text": "the brush", "range": [ 127, 136 ] } ], "source_mask_link": null } ], "actions": [ { "id": "action_pick_up", "mentions": [ { "text": "picks up", "ranges": [ [ 11, 19 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "type": "phrasal", "subject_refs": [ "entity_person" ], "object_refs": [ "entity_bottle" ], "confidence": "medium", "temporal_status": "localized_sparse_support", "intervals": [ { "start_s": 10.0, "end_s": 13.0, "source_start_s": 118.83, "source_end_s": 121.83 } ], "evidence": "Across samples 11–13, the masked person grasps and raises the small green bottle into her hands.", "mask_relation": "The action subject matches the supplied selected-person mask." }, { "id": "action_open", "mentions": [ { "text": "opens", "ranges": [ [ 51, 56 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "type": "simple", "subject_refs": [ "entity_person" ], "object_refs": [ "entity_bottle" ], "confidence": "high", "temporal_status": "localized_sparse_support", "intervals": [ { "start_s": 13.0, "end_s": 18.0, "source_start_s": 121.83, "source_end_s": 126.83 } ], "evidence": "Samples 14–18 show the masked person holding the bottle and twisting or pulling its cap and brush assembly.", "mask_relation": "The action subject matches the supplied selected-person mask." }, { "id": "action_hold", "mentions": [ { "text": "hold", "ranges": [ [ 66, 70 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "type": "simple", "subject_refs": [ "entity_person" ], "object_refs": [ "entity_brush" ], "confidence": "medium", "temporal_status": "localized_sparse_support", "intervals": [ { "start_s": 13.0, "end_s": 18.5, "source_start_s": 121.83, "source_end_s": 127.33 } ], "evidence": "In samples 14–19, the masked person keeps the extracted brush or cap-and-brush component raised in her hand.", "mask_relation": "The action subject matches the supplied selected-person mask." }, { "id": "action_look", "mentions": [ { "text": "look", "ranges": [ [ 85, 89 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "type": "prepositional", "subject_refs": [ "entity_person" ], "object_refs": [ "entity_brush" ], "confidence": "medium", "temporal_status": "localized_sparse_support", "intervals": [ { "start_s": 13.0, "end_s": 18.0, "source_start_s": 121.83, "source_end_s": 126.83 } ], "evidence": "Samples 14–18 show the masked person's head and gaze directed toward the bottle and raised brush in her hands.", "mask_relation": "The action subject matches the supplied selected-person mask." }, { "id": "action_talk", "mentions": [ { "text": "talk", "ranges": [ [ 101, 105 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "type": "simple", "subject_refs": [ "entity_person" ], "object_refs": [], "confidence": "medium", "temporal_status": "localized_sparse_support", "intervals": [ { "start_s": 13.0, "end_s": 18.5, "source_start_s": 121.83, "source_end_s": 127.33 } ], "evidence": "During samples 14–19, the masked person's mouth repeatedly changes between open and closed positions while she handles the brush.", "mask_relation": "The action subject matches the supplied selected-person mask." }, { "id": "action_gesture", "mentions": [ { "text": "gesturing", "ranges": [ [ 112, 121 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "type": "simple", "subject_refs": [ "entity_person" ], "object_refs": [ "entity_brush" ], "confidence": "medium", "temporal_status": "localized_sparse_support", "intervals": [ { "start_s": 15.0, "end_s": 18.5, "source_start_s": 123.83, "source_end_s": 127.33 } ], "evidence": "Samples 16–19 show the masked person moving the raised brush-bearing hand while speaking.", "mask_relation": "The action subject matches the supplied selected-person mask." } ], "notes": [ "Only the selected older person has a supplied masklet; the nail-polish bottle and brush have no separate source masks.", "A second child is visible on the right but is not mentioned as a separate caption entity and is not identified by the supplied mask.", "Action intervals are sparse support inferred from the sampled clean RGB frames and are not exact continuous-video boundaries." ] }, "normalization_repairs": 0 }, { "source_video_id": "v_6gZuc4umTPk", "caption_count_category": 15, "clip": { "caption_index": 10, "caption": "The person puts the brush back into the green nail polish bottle and closes it. They set the bottle down and then bend forward, leaning far down towards the table.", "source_interval_sec": [ 127.33, 150.83 ], "loom_clip_id": "v_6gZuc4umTPk_3820_4525", "clip_frame_interval_4fps": [ 3820, 4525 ], "sample_count_1fps": 24, "sampled_frames": [ { "frame": "00000", "timestamp_sec": 127.33, "offset_sec": 0, "rgb_path": "assets/cat_15/v_6gZuc4umTPk_3820_4525/0000s_00000_rgb.jpg", "mask_path": "assets/cat_15/v_6gZuc4umTPk_3820_4525/0000s_00000_mask.png", "mask_status": "available" }, { "frame": "00028", "timestamp_sec": 128.33, "offset_sec": 1, "rgb_path": "assets/cat_15/v_6gZuc4umTPk_3820_4525/0001s_00028_rgb.jpg", "mask_path": "assets/cat_15/v_6gZuc4umTPk_3820_4525/0001s_00028_mask.png", "mask_status": "available" }, { "frame": "00060", "timestamp_sec": 129.33, "offset_sec": 2, "rgb_path": "assets/cat_15/v_6gZuc4umTPk_3820_4525/0002s_00060_rgb.jpg", "mask_path": "assets/cat_15/v_6gZuc4umTPk_3820_4525/0002s_00060_mask.png", "mask_status": "available" }, { "frame": "00088", "timestamp_sec": 130.33, "offset_sec": 3, "rgb_path": "assets/cat_15/v_6gZuc4umTPk_3820_4525/0003s_00088_rgb.jpg", "mask_path": "assets/cat_15/v_6gZuc4umTPk_3820_4525/0003s_00088_mask.png", "mask_status": "available" }, { "frame": "00120", "timestamp_sec": 131.33, "offset_sec": 4, "rgb_path": "assets/cat_15/v_6gZuc4umTPk_3820_4525/0004s_00120_rgb.jpg", "mask_path": "assets/cat_15/v_6gZuc4umTPk_3820_4525/0004s_00120_mask.png", "mask_status": "available" }, { "frame": "00148", "timestamp_sec": 132.33, "offset_sec": 5, "rgb_path": "assets/cat_15/v_6gZuc4umTPk_3820_4525/0005s_00148_rgb.jpg", "mask_path": "assets/cat_15/v_6gZuc4umTPk_3820_4525/0005s_00148_mask.png", "mask_status": "available" }, { "frame": "00180", "timestamp_sec": 133.33, "offset_sec": 6, "rgb_path": "assets/cat_15/v_6gZuc4umTPk_3820_4525/0006s_00180_rgb.jpg", "mask_path": "assets/cat_15/v_6gZuc4umTPk_3820_4525/0006s_00180_mask.png", "mask_status": "available" }, { "frame": "00208", "timestamp_sec": 134.33, "offset_sec": 7, "rgb_path": "assets/cat_15/v_6gZuc4umTPk_3820_4525/0007s_00208_rgb.jpg", "mask_path": "assets/cat_15/v_6gZuc4umTPk_3820_4525/0007s_00208_mask.png", "mask_status": "available" }, { "frame": "00240", "timestamp_sec": 135.33, "offset_sec": 8, "rgb_path": "assets/cat_15/v_6gZuc4umTPk_3820_4525/0008s_00240_rgb.jpg", "mask_path": "assets/cat_15/v_6gZuc4umTPk_3820_4525/0008s_00240_mask.png", "mask_status": "available" }, { "frame": "00268", "timestamp_sec": 136.33, "offset_sec": 9, "rgb_path": "assets/cat_15/v_6gZuc4umTPk_3820_4525/0009s_00268_rgb.jpg", "mask_path": "assets/cat_15/v_6gZuc4umTPk_3820_4525/0009s_00268_mask.png", "mask_status": "available" }, { "frame": "00300", "timestamp_sec": 137.33, "offset_sec": 10, "rgb_path": "assets/cat_15/v_6gZuc4umTPk_3820_4525/0010s_00300_rgb.jpg", "mask_path": "assets/cat_15/v_6gZuc4umTPk_3820_4525/0010s_00300_mask.png", "mask_status": "available" }, { "frame": "00328", "timestamp_sec": 138.33, "offset_sec": 11, "rgb_path": "assets/cat_15/v_6gZuc4umTPk_3820_4525/0011s_00328_rgb.jpg", "mask_path": "assets/cat_15/v_6gZuc4umTPk_3820_4525/0011s_00328_mask.png", "mask_status": "available" }, { "frame": "00360", "timestamp_sec": 139.33, "offset_sec": 12, "rgb_path": "assets/cat_15/v_6gZuc4umTPk_3820_4525/0012s_00360_rgb.jpg", "mask_path": "assets/cat_15/v_6gZuc4umTPk_3820_4525/0012s_00360_mask.png", "mask_status": "available" }, { "frame": "00388", "timestamp_sec": 140.33, "offset_sec": 13, "rgb_path": "assets/cat_15/v_6gZuc4umTPk_3820_4525/0013s_00388_rgb.jpg", "mask_path": "assets/cat_15/v_6gZuc4umTPk_3820_4525/0013s_00388_mask.png", "mask_status": "available" }, { "frame": "00420", "timestamp_sec": 141.33, "offset_sec": 14, "rgb_path": "assets/cat_15/v_6gZuc4umTPk_3820_4525/0014s_00420_rgb.jpg", "mask_path": "assets/cat_15/v_6gZuc4umTPk_3820_4525/0014s_00420_mask.png", "mask_status": "available" }, { "frame": "00448", "timestamp_sec": 142.33, "offset_sec": 15, "rgb_path": "assets/cat_15/v_6gZuc4umTPk_3820_4525/0015s_00448_rgb.jpg", "mask_path": "assets/cat_15/v_6gZuc4umTPk_3820_4525/0015s_00448_mask.png", "mask_status": "available" }, { "frame": "00480", "timestamp_sec": 143.33, "offset_sec": 16, "rgb_path": "assets/cat_15/v_6gZuc4umTPk_3820_4525/0016s_00480_rgb.jpg", "mask_path": "assets/cat_15/v_6gZuc4umTPk_3820_4525/0016s_00480_mask.png", "mask_status": "available" }, { "frame": "00508", "timestamp_sec": 144.33, "offset_sec": 17, "rgb_path": "assets/cat_15/v_6gZuc4umTPk_3820_4525/0017s_00508_rgb.jpg", "mask_path": "assets/cat_15/v_6gZuc4umTPk_3820_4525/0017s_00508_mask.png", "mask_status": "available" }, { "frame": "00540", "timestamp_sec": 145.33, "offset_sec": 18, "rgb_path": "assets/cat_15/v_6gZuc4umTPk_3820_4525/0018s_00540_rgb.jpg", "mask_path": "assets/cat_15/v_6gZuc4umTPk_3820_4525/0018s_00540_mask.png", "mask_status": "available" }, { "frame": "00568", "timestamp_sec": 146.33, "offset_sec": 19, "rgb_path": "assets/cat_15/v_6gZuc4umTPk_3820_4525/0019s_00568_rgb.jpg", "mask_path": "assets/cat_15/v_6gZuc4umTPk_3820_4525/0019s_00568_mask.png", "mask_status": "available" }, { "frame": "00600", "timestamp_sec": 147.33, "offset_sec": 20, "rgb_path": "assets/cat_15/v_6gZuc4umTPk_3820_4525/0020s_00600_rgb.jpg", "mask_path": "assets/cat_15/v_6gZuc4umTPk_3820_4525/0020s_00600_mask.png", "mask_status": "available" }, { "frame": "00628", "timestamp_sec": 148.33, "offset_sec": 21, "rgb_path": "assets/cat_15/v_6gZuc4umTPk_3820_4525/0021s_00628_rgb.jpg", "mask_path": "assets/cat_15/v_6gZuc4umTPk_3820_4525/0021s_00628_mask.png", "mask_status": "available" }, { "frame": "00660", "timestamp_sec": 149.33, "offset_sec": 22, "rgb_path": "assets/cat_15/v_6gZuc4umTPk_3820_4525/0022s_00660_rgb.jpg", "mask_path": "assets/cat_15/v_6gZuc4umTPk_3820_4525/0022s_00660_mask.png", "mask_status": "available" }, { "frame": "00688", "timestamp_sec": 150.33, "offset_sec": 23, "rgb_path": "assets/cat_15/v_6gZuc4umTPk_3820_4525/0023s_00688_rgb.jpg", "mask_path": "assets/cat_15/v_6gZuc4umTPk_3820_4525/0023s_00688_mask.png", "mask_status": "available" } ], "obj_id": [ 0 ], "anno_id": [ 938 ], "fps_inferred_from_clip_interval": 30.0 }, "usage": { "input_tokens": 27011, "input_tokens_details": { "cache_write_tokens": 27008, "cached_tokens": 0 }, "output_tokens": 4730, "output_tokens_details": { "reasoning_tokens": 3082 }, "total_tokens": 31741 }, "model": "gpt-5.6-sol", "effort": "medium", "custom_id": "cat15_cap10_v_6gZuc4umTPk_3820_4525", "annotation": { "caption": "The person puts the brush back into the green nail polish bottle and closes it. They set the bottle down and then bend forward, leaning far down towards the table.", "entity_groups": [ { "id": "entity_person", "label": "person", "spans": [ { "text": "The person", "range": [ 0, 10 ] }, { "text": "They", "range": [ 80, 84 ] } ], "source_mask_link": "selected_person" }, { "id": "entity_brush", "label": "brush", "spans": [ { "text": "the brush", "range": [ 16, 25 ] } ], "source_mask_link": null }, { "id": "entity_bottle", "label": "nail polish bottle", "spans": [ { "text": "the green nail polish bottle", "range": [ 36, 64 ] }, { "text": "it", "range": [ 76, 78 ] }, { "text": "the bottle", "range": [ 89, 99 ] } ], "source_mask_link": null }, { "id": "entity_table", "label": "table", "spans": [ { "text": "the table", "range": [ 153, 162 ] } ], "source_mask_link": null } ], "actions": [ { "id": "action_put_brush_back", "mentions": [ { "text": "puts", "ranges": [ [ 11, 15 ] ], "occurrence": 1, "particle": { "text": "back", "occurrence": 1, "range": { "start": 26, "end": 30 } } } ], "type": "phrasal", "subject_refs": [ "entity_person" ], "object_refs": [ "entity_brush", "entity_bottle" ], "confidence": "high", "temporal_status": "localized", "intervals": [ { "start_s": 9.0, "end_s": 11.0, "source_start_s": 136.32999999999998, "source_end_s": 138.32999999999998 } ], "evidence": "At 9–10 seconds, the masked person moves the brush/cap over and into the opening of the green bottle.", "mask_relation": "subject matches the supplied selected-person mask", "source_subject_refs": "entity_person" }, { "id": "action_close_bottle", "mentions": [ { "text": "closes", "ranges": [ [ 69, 75 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1, "ranges": [] } } ], "type": "simple", "subject_refs": [ "entity_person" ], "object_refs": [ "entity_bottle" ], "confidence": "medium", "temporal_status": "localized", "intervals": [ { "start_s": 10.0, "end_s": 11.0, "source_start_s": 137.32999999999998, "source_end_s": 138.32999999999998 } ], "evidence": "At 10 seconds, the masked person's hands manipulate the bottle opening; by 11 seconds the assembled bottle is being held.", "mask_relation": "subject matches the supplied selected-person mask", "source_subject_refs": "entity_person" }, { "id": "action_set_bottle_down", "mentions": [ { "text": "set", "ranges": [ [ 85, 88 ] ], "occurrence": 1, "particle": { "text": "down", "occurrence": 1, "range": { "start": 100, "end": 104 } } } ], "type": "phrasal", "subject_refs": [ "entity_person" ], "object_refs": [ "entity_bottle" ], "confidence": "medium", "temporal_status": "localized_with_sparse_support", "intervals": [ { "start_s": 12.0, "end_s": 14.0, "source_start_s": 139.32999999999998, "source_end_s": 141.32999999999998 } ], "evidence": "At 12–13 seconds the masked person carries the bottle toward the left edge; by 14 seconds it is no longer held visibly and the hand is withdrawing.", "mask_relation": "subject matches the supplied selected-person mask", "source_subject_refs": "entity_person" }, { "id": "action_bend_and_lean", "mentions": [ { "text": "bend", "ranges": [ [ 114, 118 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1, "ranges": [] } }, { "text": "leaning", "ranges": [ [ 128, 135 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1, "ranges": [] } } ], "type": "posture", "subject_refs": [ "entity_person" ], "object_refs": [], "confidence": "high", "temporal_status": "localized", "intervals": [ { "start_s": 15.0, "end_s": 23.0, "source_start_s": 142.32999999999998, "source_end_s": 150.32999999999998 } ], "evidence": "From 15 seconds the masked person lowers her upper body, remaining bent and leaning far over the blue tabletop through 22 seconds.", "mask_relation": "subject matches the supplied selected-person mask", "source_subject_refs": "entity_person" } ], "notes": [ "The supplied mask consistently identifies the dark-haired person performing the captioned actions.", "The visible blonde person has no text mention in the caption and is not assigned an entity group.", "No source masks were supplied for the brush, nail polish bottle, or table.", "The bottle's placement surface is partly off-frame, so the set-down timing is supported indirectly by its movement out of the person's hand." ] }, "model_output_raw": { "caption": "The person puts the brush back into the green nail polish bottle and closes it. They set the bottle down and then bend forward, leaning far down towards the table.", "entity_groups": [ { "id": "entity_person", "label": "person", "spans": [ { "text": "The person", "start": 0, "end": 10 }, { "text": "They", "start": 80, "end": 84 } ], "source_mask": "selected_person_track" }, { "id": "entity_brush", "label": "brush", "spans": [ { "text": "the brush", "start": 16, "end": 25 } ], "source_mask": null }, { "id": "entity_bottle", "label": "nail polish bottle", "spans": [ { "text": "the green nail polish bottle", "start": 36, "end": 64 }, { "text": "it", "start": 76, "end": 78 }, { "text": "the bottle", "start": 89, "end": 99 } ], "source_mask": null }, { "id": "entity_table", "label": "table", "spans": [ { "text": "the table", "start": 153, "end": 162 } ], "source_mask": null } ], "actions": [ { "id": "action_put_brush_back", "mentions": [ { "text": "puts", "occurrence": 1, "ranges": [ { "start": 11, "end": 15 } ], "particle": { "text": "back", "occurrence": 1, "range": { "start": 26, "end": 30 } } } ], "type": "phrasal", "subject_ref": "entity_person", "object_refs": [ { "entity_group_id": "entity_brush", "role": "theme" }, { "entity_group_id": "entity_bottle", "role": "destination" } ], "confidence": "high", "mask_relation": "subject matches the supplied selected-person mask", "temporal_status": "localized", "intervals": [ { "start_s": 9.0, "end_s": 11.0, "source_start_s": 136.33, "source_end_s": 138.33 } ], "evidence": "At 9–10 seconds, the masked person moves the brush/cap over and into the opening of the green bottle." }, { "id": "action_close_bottle", "mentions": [ { "text": "closes", "occurrence": 1, "ranges": [ { "start": 69, "end": 75 } ], "particle": { "text": null, "occurrence": 1, "ranges": [] } } ], "type": "simple", "subject_ref": "entity_person", "object_refs": [ { "entity_group_id": "entity_bottle", "role": "patient" } ], "confidence": "medium", "mask_relation": "subject matches the supplied selected-person mask", "temporal_status": "localized", "intervals": [ { "start_s": 10.0, "end_s": 11.0, "source_start_s": 137.33, "source_end_s": 138.33 } ], "evidence": "At 10 seconds, the masked person's hands manipulate the bottle opening; by 11 seconds the assembled bottle is being held." }, { "id": "action_set_bottle_down", "mentions": [ { "text": "set", "occurrence": 1, "ranges": [ { "start": 85, "end": 88 } ], "particle": { "text": "down", "occurrence": 1, "range": { "start": 100, "end": 104 } } } ], "type": "phrasal", "subject_ref": "entity_person", "object_refs": [ { "entity_group_id": "entity_bottle", "role": "theme" } ], "confidence": "medium", "mask_relation": "subject matches the supplied selected-person mask", "temporal_status": "localized_with_sparse_support", "intervals": [ { "start_s": 12.0, "end_s": 14.0, "source_start_s": 139.33, "source_end_s": 141.33 } ], "evidence": "At 12–13 seconds the masked person carries the bottle toward the left edge; by 14 seconds it is no longer held visibly and the hand is withdrawing." }, { "id": "action_bend_and_lean", "mentions": [ { "text": "bend", "occurrence": 1, "ranges": [ { "start": 114, "end": 118 } ], "particle": { "text": null, "occurrence": 1, "ranges": [] } }, { "text": "leaning", "occurrence": 1, "ranges": [ { "start": 128, "end": 135 } ], "particle": { "text": null, "occurrence": 1, "ranges": [] } } ], "type": "posture", "subject_ref": "entity_person", "object_refs": [], "related_entity_refs": [ { "entity_group_id": "entity_table", "role": "directional_reference" } ], "confidence": "high", "mask_relation": "subject matches the supplied selected-person mask", "temporal_status": "localized", "intervals": [ { "start_s": 15.0, "end_s": 23.0, "source_start_s": 142.33, "source_end_s": 150.33 } ], "evidence": "From 15 seconds the masked person lowers her upper body, remaining bent and leaning far over the blue tabletop through 22 seconds." } ], "notes": [ "The supplied mask consistently identifies the dark-haired person performing the captioned actions.", "The visible blonde person has no text mention in the caption and is not assigned an entity group.", "No source masks were supplied for the brush, nail polish bottle, or table.", "The bottle's placement surface is partly off-frame, so the set-down timing is supported indirectly by its movement out of the person's hand." ] }, "normalization_repairs": 0 }, { "source_video_id": "v_6gZuc4umTPk", "caption_count_category": 15, "clip": { "caption_index": 11, "caption": "The person sits up and picks up a bottle of purple nail polish. They open the bottle and place the cap down on the table before looking around.", "source_interval_sec": [ 150.83, 166.33 ], "loom_clip_id": "v_6gZuc4umTPk_4525_4990", "clip_frame_interval_4fps": [ 4525, 4990 ], "sample_count_1fps": 16, "sampled_frames": [ { "frame": "00000", "timestamp_sec": 150.83, "offset_sec": 0, "rgb_path": "assets/cat_15/v_6gZuc4umTPk_4525_4990/0000s_00000_rgb.jpg", "mask_path": "assets/cat_15/v_6gZuc4umTPk_4525_4990/0000s_00000_mask.png", "mask_status": "available" }, { "frame": "00028", "timestamp_sec": 151.83, "offset_sec": 1, "rgb_path": "assets/cat_15/v_6gZuc4umTPk_4525_4990/0001s_00028_rgb.jpg", "mask_path": "assets/cat_15/v_6gZuc4umTPk_4525_4990/0001s_00028_mask.png", "mask_status": "available" }, { "frame": "00060", "timestamp_sec": 152.83, "offset_sec": 2, "rgb_path": "assets/cat_15/v_6gZuc4umTPk_4525_4990/0002s_00060_rgb.jpg", "mask_path": "assets/cat_15/v_6gZuc4umTPk_4525_4990/0002s_00060_mask.png", "mask_status": "available" }, { "frame": "00088", "timestamp_sec": 153.83, "offset_sec": 3, "rgb_path": "assets/cat_15/v_6gZuc4umTPk_4525_4990/0003s_00088_rgb.jpg", "mask_path": "assets/cat_15/v_6gZuc4umTPk_4525_4990/0003s_00088_mask.png", "mask_status": "available" }, { "frame": "00120", "timestamp_sec": 154.83, "offset_sec": 4, "rgb_path": "assets/cat_15/v_6gZuc4umTPk_4525_4990/0004s_00120_rgb.jpg", "mask_path": "assets/cat_15/v_6gZuc4umTPk_4525_4990/0004s_00120_mask.png", "mask_status": "available" }, { "frame": "00148", "timestamp_sec": 155.83, "offset_sec": 5, "rgb_path": "assets/cat_15/v_6gZuc4umTPk_4525_4990/0005s_00148_rgb.jpg", "mask_path": "assets/cat_15/v_6gZuc4umTPk_4525_4990/0005s_00148_mask.png", "mask_status": "available" }, { "frame": "00180", "timestamp_sec": 156.83, "offset_sec": 6, "rgb_path": "assets/cat_15/v_6gZuc4umTPk_4525_4990/0006s_00180_rgb.jpg", "mask_path": "assets/cat_15/v_6gZuc4umTPk_4525_4990/0006s_00180_mask.png", "mask_status": "available" }, { "frame": "00208", "timestamp_sec": 157.83, "offset_sec": 7, "rgb_path": "assets/cat_15/v_6gZuc4umTPk_4525_4990/0007s_00208_rgb.jpg", "mask_path": "assets/cat_15/v_6gZuc4umTPk_4525_4990/0007s_00208_mask.png", "mask_status": "available" }, { "frame": "00240", "timestamp_sec": 158.83, "offset_sec": 8, "rgb_path": "assets/cat_15/v_6gZuc4umTPk_4525_4990/0008s_00240_rgb.jpg", "mask_path": "assets/cat_15/v_6gZuc4umTPk_4525_4990/0008s_00240_mask.png", "mask_status": "available" }, { "frame": "00268", "timestamp_sec": 159.83, "offset_sec": 9, "rgb_path": "assets/cat_15/v_6gZuc4umTPk_4525_4990/0009s_00268_rgb.jpg", "mask_path": "assets/cat_15/v_6gZuc4umTPk_4525_4990/0009s_00268_mask.png", "mask_status": "available" }, { "frame": "00300", "timestamp_sec": 160.83, "offset_sec": 10, "rgb_path": "assets/cat_15/v_6gZuc4umTPk_4525_4990/0010s_00300_rgb.jpg", "mask_path": "assets/cat_15/v_6gZuc4umTPk_4525_4990/0010s_00300_mask.png", "mask_status": "available" }, { "frame": "00328", "timestamp_sec": 161.83, "offset_sec": 11, "rgb_path": "assets/cat_15/v_6gZuc4umTPk_4525_4990/0011s_00328_rgb.jpg", "mask_path": "assets/cat_15/v_6gZuc4umTPk_4525_4990/0011s_00328_mask.png", "mask_status": "available" }, { "frame": "00360", "timestamp_sec": 162.83, "offset_sec": 12, "rgb_path": "assets/cat_15/v_6gZuc4umTPk_4525_4990/0012s_00360_rgb.jpg", "mask_path": "assets/cat_15/v_6gZuc4umTPk_4525_4990/0012s_00360_mask.png", "mask_status": "available" }, { "frame": "00388", "timestamp_sec": 163.83, "offset_sec": 13, "rgb_path": "assets/cat_15/v_6gZuc4umTPk_4525_4990/0013s_00388_rgb.jpg", "mask_path": "assets/cat_15/v_6gZuc4umTPk_4525_4990/0013s_00388_mask.png", "mask_status": "available" }, { "frame": "00420", "timestamp_sec": 164.83, "offset_sec": 14, "rgb_path": "assets/cat_15/v_6gZuc4umTPk_4525_4990/0014s_00420_rgb.jpg", "mask_path": "assets/cat_15/v_6gZuc4umTPk_4525_4990/0014s_00420_mask.png", "mask_status": "available" }, { "frame": "00448", "timestamp_sec": 165.83, "offset_sec": 15, "rgb_path": "assets/cat_15/v_6gZuc4umTPk_4525_4990/0015s_00448_rgb.jpg", "mask_path": "assets/cat_15/v_6gZuc4umTPk_4525_4990/0015s_00448_mask.png", "mask_status": "available" } ], "obj_id": [ 0 ], "anno_id": [ 939 ], "fps_inferred_from_clip_interval": 30.0 }, "usage": { "input_tokens": 23112, "input_tokens_details": { "cache_write_tokens": 23109, "cached_tokens": 0 }, "output_tokens": 4567, "output_tokens_details": { "reasoning_tokens": 3099 }, "total_tokens": 27679 }, "model": "gpt-5.6-sol", "effort": "medium", "custom_id": "cat15_cap11_v_6gZuc4umTPk_4525_4990", "annotation": { "caption": "The person sits up and picks up a bottle of purple nail polish. They open the bottle and place the cap down on the table before looking around.", "entity_groups": [ { "id": "person_1", "label": "person", "spans": [ { "text": "The person", "range": [ 0, 10 ] }, { "text": "They", "range": [ 64, 68 ] } ], "source_mask_link": "selected_person" }, { "id": "bottle_1", "label": "bottle", "spans": [ { "text": "a bottle", "range": [ 32, 40 ] }, { "text": "the bottle", "range": [ 74, 84 ] } ], "source_mask_link": null }, { "id": "cap_1", "label": "cap", "spans": [ { "text": "the cap", "range": [ 95, 102 ] } ], "source_mask_link": null }, { "id": "table_1", "label": "table", "spans": [ { "text": "the table", "range": [ 111, 120 ] } ], "source_mask_link": null } ], "actions": [ { "id": "action_1", "mentions": [ { "text": "sits up", "ranges": [ [ 11, 18 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "type": "posture", "subject_refs": [ "person_1" ], "object_refs": [], "confidence": "high", "temporal_status": "localized", "intervals": [ { "start_s": 0.0, "end_s": 2.0, "source_start_s": 150.83, "source_end_s": 152.83 } ], "evidence": "At samples 1–2 (0.000–1.000 s), the masked person changes from leaning close over the table to a more upright seated posture.", "mask_relation": "The action subject matches the supplied selected-person mask.", "source_subject_refs": "person_1" }, { "id": "action_2", "mentions": [ { "text": "picks up", "ranges": [ [ 23, 31 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "type": "phrasal", "subject_refs": [ "person_1" ], "object_refs": [ "bottle_1" ], "confidence": "high", "temporal_status": "localized", "intervals": [ { "start_s": 1.0, "end_s": 3.0, "source_start_s": 151.83, "source_end_s": 153.83 } ], "evidence": "Samples 2–3 (1.000–2.000 s) show the masked person reaching and then holding a small object between both hands.", "mask_relation": "The action subject matches the supplied selected-person mask.", "source_subject_refs": "person_1" }, { "id": "action_3", "mentions": [ { "text": "open", "ranges": [ [ 69, 73 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "type": "simple", "subject_refs": [ "person_1" ], "object_refs": [ "bottle_1" ], "confidence": "high", "temporal_status": "localized", "intervals": [ { "start_s": 2.0, "end_s": 7.0, "source_start_s": 152.83, "source_end_s": 157.83 } ], "evidence": "Across samples 3–7 (2.000–6.000 s), the masked person repeatedly manipulates the small object with both hands in a twisting and separating motion consistent with opening it.", "mask_relation": "The action subject matches the supplied selected-person mask.", "source_subject_refs": "person_1" }, { "id": "action_4", "mentions": [ { "text": "place", "ranges": [ [ 89, 94 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "type": "simple", "subject_refs": [ "person_1" ], "object_refs": [ "cap_1" ], "confidence": "high", "temporal_status": "uncertain", "intervals": [], "evidence": "Samples 13–15 show the masked person lowering an object toward the blue table and an object standing on the table, but the sparse frames do not clearly establish that the object placed is the bottle cap rather than the bottle.", "mask_relation": "The action subject matches the supplied selected-person mask.", "source_subject_refs": "person_1" }, { "id": "action_5", "mentions": [ { "text": "looking around", "ranges": [ [ 128, 142 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "type": "phrasal", "subject_refs": [ "person_1" ], "object_refs": [], "confidence": "high", "temporal_status": "uncertain", "intervals": [], "evidence": "No sampled clean frame clearly shows the masked person's head surveying multiple directions after the placement event; the head is largely cropped in the final samples.", "mask_relation": "The action subject matches the supplied selected-person mask.", "source_subject_refs": "person_1" } ], "notes": [ "The supplied masklet consistently identifies the older person on the left, corresponding to both “The person” and “They.”", "No source masks are supplied for the bottle, cap, or table.", "The unmentioned child visible on the right is not assigned an entity group.", "The phrase “a bottle of purple nail polish” is treated as a container-focused reference: the bottle span is “a bottle,” while the postmodifying content phrase is not separately grounded." ] }, "model_output_raw": { "caption": "The person sits up and picks up a bottle of purple nail polish. They open the bottle and place the cap down on the table before looking around.", "entity_groups": [ { "id": "person_1", "label": "person", "spans": [ { "text": "The person", "range": [ 0, 10 ] }, { "text": "They", "range": [ 64, 68 ] } ], "source_mask": "supplied_selected_person_masklet" }, { "id": "bottle_1", "label": "bottle", "spans": [ { "text": "a bottle", "range": [ 32, 40 ] }, { "text": "the bottle", "range": [ 74, 84 ] } ], "source_mask": null }, { "id": "cap_1", "label": "cap", "spans": [ { "text": "the cap", "range": [ 95, 102 ] } ], "source_mask": null }, { "id": "table_1", "label": "table", "spans": [ { "text": "the table", "range": [ 111, 120 ] } ], "source_mask": null } ], "actions": [ { "id": "action_1", "mentions": [ { "text": "sits up", "occurrence": 1, "ranges": [ [ 11, 18 ] ], "particle": { "text": null, "occurrence": 1 } } ], "type": "posture", "subject_entity_id": "person_1", "object_entity_ids": [], "confidence": "high", "temporal_status": "localized", "intervals": [ { "start_s": 0.0, "end_s": 2.0 } ], "evidence": "At samples 1–2 (0.000–1.000 s), the masked person changes from leaning close over the table to a more upright seated posture.", "mask_relation": "The action subject matches the supplied selected-person mask." }, { "id": "action_2", "mentions": [ { "text": "picks up", "occurrence": 1, "ranges": [ [ 23, 31 ] ], "particle": { "text": null, "occurrence": 1 } } ], "type": "phrasal", "subject_entity_id": "person_1", "object_entity_ids": [ "bottle_1" ], "confidence": "high", "temporal_status": "localized", "intervals": [ { "start_s": 1.0, "end_s": 3.0 } ], "evidence": "Samples 2–3 (1.000–2.000 s) show the masked person reaching and then holding a small object between both hands.", "mask_relation": "The action subject matches the supplied selected-person mask." }, { "id": "action_3", "mentions": [ { "text": "open", "occurrence": 1, "ranges": [ [ 69, 73 ] ], "particle": { "text": null, "occurrence": 1 } } ], "type": "simple", "subject_entity_id": "person_1", "object_entity_ids": [ "bottle_1" ], "confidence": "high", "temporal_status": "localized", "intervals": [ { "start_s": 2.0, "end_s": 7.0 } ], "evidence": "Across samples 3–7 (2.000–6.000 s), the masked person repeatedly manipulates the small object with both hands in a twisting and separating motion consistent with opening it.", "mask_relation": "The action subject matches the supplied selected-person mask." }, { "id": "action_4", "mentions": [ { "text": "place", "occurrence": 1, "ranges": [ [ 89, 94 ] ], "particle": { "text": null, "occurrence": 1 } } ], "type": "simple", "subject_entity_id": "person_1", "object_entity_ids": [ "cap_1" ], "location_entity_ids": [ "table_1" ], "confidence": "high", "temporal_status": "uncertain", "intervals": [], "evidence": "Samples 13–15 show the masked person lowering an object toward the blue table and an object standing on the table, but the sparse frames do not clearly establish that the object placed is the bottle cap rather than the bottle.", "mask_relation": "The action subject matches the supplied selected-person mask." }, { "id": "action_5", "mentions": [ { "text": "looking around", "occurrence": 1, "ranges": [ [ 128, 142 ] ], "particle": { "text": null, "occurrence": 1 } } ], "type": "phrasal", "subject_entity_id": "person_1", "object_entity_ids": [], "confidence": "high", "temporal_status": "uncertain", "intervals": [], "evidence": "No sampled clean frame clearly shows the masked person's head surveying multiple directions after the placement event; the head is largely cropped in the final samples.", "mask_relation": "The action subject matches the supplied selected-person mask." } ], "notes": [ "The supplied masklet consistently identifies the older person on the left, corresponding to both “The person” and “They.”", "No source masks are supplied for the bottle, cap, or table.", "The unmentioned child visible on the right is not assigned an entity group.", "The phrase “a bottle of purple nail polish” is treated as a container-focused reference: the bottle span is “a bottle,” while the postmodifying content phrase is not separately grounded." ] }, "normalization_repairs": 0 }, { "source_video_id": "v_6gZuc4umTPk", "caption_count_category": 15, "clip": { "caption_index": 12, "caption": "The person reaches over to take another person's hand. They hold the hand steady and begin to apply the purple nail polish to a finger.", "source_interval_sec": [ 166.33, 191.83 ], "loom_clip_id": "v_6gZuc4umTPk_4990_5755", "clip_frame_interval_4fps": [ 4990, 5755 ], "sample_count_1fps": 26, "sampled_frames": [ { "frame": "00000", "timestamp_sec": 166.33, "offset_sec": 0, "rgb_path": "assets/cat_15/v_6gZuc4umTPk_4990_5755/0000s_00000_rgb.jpg", "mask_path": "assets/cat_15/v_6gZuc4umTPk_4990_5755/0000s_00000_mask.png", "mask_status": "available" }, { "frame": "00028", "timestamp_sec": 167.33, "offset_sec": 1, "rgb_path": "assets/cat_15/v_6gZuc4umTPk_4990_5755/0001s_00028_rgb.jpg", "mask_path": "assets/cat_15/v_6gZuc4umTPk_4990_5755/0001s_00028_mask.png", "mask_status": "available" }, { "frame": "00060", "timestamp_sec": 168.33, "offset_sec": 2, "rgb_path": "assets/cat_15/v_6gZuc4umTPk_4990_5755/0002s_00060_rgb.jpg", "mask_path": "assets/cat_15/v_6gZuc4umTPk_4990_5755/0002s_00060_mask.png", "mask_status": "available" }, { "frame": "00088", "timestamp_sec": 169.33, "offset_sec": 3, "rgb_path": "assets/cat_15/v_6gZuc4umTPk_4990_5755/0003s_00088_rgb.jpg", "mask_path": "assets/cat_15/v_6gZuc4umTPk_4990_5755/0003s_00088_mask.png", "mask_status": "available" }, { "frame": "00120", "timestamp_sec": 170.33, "offset_sec": 4, "rgb_path": "assets/cat_15/v_6gZuc4umTPk_4990_5755/0004s_00120_rgb.jpg", "mask_path": "assets/cat_15/v_6gZuc4umTPk_4990_5755/0004s_00120_mask.png", "mask_status": "available" }, { "frame": "00148", "timestamp_sec": 171.33, "offset_sec": 5, "rgb_path": "assets/cat_15/v_6gZuc4umTPk_4990_5755/0005s_00148_rgb.jpg", "mask_path": "assets/cat_15/v_6gZuc4umTPk_4990_5755/0005s_00148_mask.png", "mask_status": "available" }, { "frame": "00180", "timestamp_sec": 172.33, "offset_sec": 6, "rgb_path": "assets/cat_15/v_6gZuc4umTPk_4990_5755/0006s_00180_rgb.jpg", "mask_path": "assets/cat_15/v_6gZuc4umTPk_4990_5755/0006s_00180_mask.png", "mask_status": "available" }, { "frame": "00208", "timestamp_sec": 173.33, "offset_sec": 7, "rgb_path": "assets/cat_15/v_6gZuc4umTPk_4990_5755/0007s_00208_rgb.jpg", "mask_path": "assets/cat_15/v_6gZuc4umTPk_4990_5755/0007s_00208_mask.png", "mask_status": "available" }, { "frame": "00240", "timestamp_sec": 174.33, "offset_sec": 8, "rgb_path": "assets/cat_15/v_6gZuc4umTPk_4990_5755/0008s_00240_rgb.jpg", "mask_path": "assets/cat_15/v_6gZuc4umTPk_4990_5755/0008s_00240_mask.png", "mask_status": "available" }, { "frame": "00268", "timestamp_sec": 175.33, "offset_sec": 9, "rgb_path": "assets/cat_15/v_6gZuc4umTPk_4990_5755/0009s_00268_rgb.jpg", "mask_path": "assets/cat_15/v_6gZuc4umTPk_4990_5755/0009s_00268_mask.png", "mask_status": "available" }, { "frame": "00300", "timestamp_sec": 176.33, "offset_sec": 10, "rgb_path": "assets/cat_15/v_6gZuc4umTPk_4990_5755/0010s_00300_rgb.jpg", "mask_path": "assets/cat_15/v_6gZuc4umTPk_4990_5755/0010s_00300_mask.png", "mask_status": "available" }, { "frame": "00328", "timestamp_sec": 177.33, "offset_sec": 11, "rgb_path": "assets/cat_15/v_6gZuc4umTPk_4990_5755/0011s_00328_rgb.jpg", "mask_path": "assets/cat_15/v_6gZuc4umTPk_4990_5755/0011s_00328_mask.png", "mask_status": "available" }, { "frame": "00360", "timestamp_sec": 178.33, "offset_sec": 12, "rgb_path": "assets/cat_15/v_6gZuc4umTPk_4990_5755/0012s_00360_rgb.jpg", "mask_path": "assets/cat_15/v_6gZuc4umTPk_4990_5755/0012s_00360_mask.png", "mask_status": "available" }, { "frame": "00388", "timestamp_sec": 179.33, "offset_sec": 13, "rgb_path": "assets/cat_15/v_6gZuc4umTPk_4990_5755/0013s_00388_rgb.jpg", "mask_path": "assets/cat_15/v_6gZuc4umTPk_4990_5755/0013s_00388_mask.png", "mask_status": "available" }, { "frame": "00420", "timestamp_sec": 180.33, "offset_sec": 14, "rgb_path": "assets/cat_15/v_6gZuc4umTPk_4990_5755/0014s_00420_rgb.jpg", "mask_path": "assets/cat_15/v_6gZuc4umTPk_4990_5755/0014s_00420_mask.png", "mask_status": "available" }, { "frame": "00448", "timestamp_sec": 181.33, "offset_sec": 15, "rgb_path": "assets/cat_15/v_6gZuc4umTPk_4990_5755/0015s_00448_rgb.jpg", "mask_path": "assets/cat_15/v_6gZuc4umTPk_4990_5755/0015s_00448_mask.png", "mask_status": "available" }, { "frame": "00480", "timestamp_sec": 182.33, "offset_sec": 16, "rgb_path": "assets/cat_15/v_6gZuc4umTPk_4990_5755/0016s_00480_rgb.jpg", "mask_path": "assets/cat_15/v_6gZuc4umTPk_4990_5755/0016s_00480_mask.png", "mask_status": "available" }, { "frame": "00508", "timestamp_sec": 183.33, "offset_sec": 17, "rgb_path": "assets/cat_15/v_6gZuc4umTPk_4990_5755/0017s_00508_rgb.jpg", "mask_path": "assets/cat_15/v_6gZuc4umTPk_4990_5755/0017s_00508_mask.png", "mask_status": "available" }, { "frame": "00540", "timestamp_sec": 184.33, "offset_sec": 18, "rgb_path": "assets/cat_15/v_6gZuc4umTPk_4990_5755/0018s_00540_rgb.jpg", "mask_path": "assets/cat_15/v_6gZuc4umTPk_4990_5755/0018s_00540_mask.png", "mask_status": "available" }, { "frame": "00568", "timestamp_sec": 185.33, "offset_sec": 19, "rgb_path": "assets/cat_15/v_6gZuc4umTPk_4990_5755/0019s_00568_rgb.jpg", "mask_path": "assets/cat_15/v_6gZuc4umTPk_4990_5755/0019s_00568_mask.png", "mask_status": "available" }, { "frame": "00600", "timestamp_sec": 186.33, "offset_sec": 20, "rgb_path": "assets/cat_15/v_6gZuc4umTPk_4990_5755/0020s_00600_rgb.jpg", "mask_path": "assets/cat_15/v_6gZuc4umTPk_4990_5755/0020s_00600_mask.png", "mask_status": "available" }, { "frame": "00628", "timestamp_sec": 187.33, "offset_sec": 21, "rgb_path": "assets/cat_15/v_6gZuc4umTPk_4990_5755/0021s_00628_rgb.jpg", "mask_path": "assets/cat_15/v_6gZuc4umTPk_4990_5755/0021s_00628_mask.png", "mask_status": "available" }, { "frame": "00660", "timestamp_sec": 188.33, "offset_sec": 22, "rgb_path": "assets/cat_15/v_6gZuc4umTPk_4990_5755/0022s_00660_rgb.jpg", "mask_path": "assets/cat_15/v_6gZuc4umTPk_4990_5755/0022s_00660_mask.png", "mask_status": "available" }, { "frame": "00688", "timestamp_sec": 189.33, "offset_sec": 23, "rgb_path": "assets/cat_15/v_6gZuc4umTPk_4990_5755/0023s_00688_rgb.jpg", "mask_path": "assets/cat_15/v_6gZuc4umTPk_4990_5755/0023s_00688_mask.png", "mask_status": "available" }, { "frame": "00720", "timestamp_sec": 190.33, "offset_sec": 24, "rgb_path": "assets/cat_15/v_6gZuc4umTPk_4990_5755/0024s_00720_rgb.jpg", "mask_path": "assets/cat_15/v_6gZuc4umTPk_4990_5755/0024s_00720_mask.png", "mask_status": "available" }, { "frame": "00748", "timestamp_sec": 191.33, "offset_sec": 25, "rgb_path": "assets/cat_15/v_6gZuc4umTPk_4990_5755/0025s_00748_rgb.jpg", "mask_path": "assets/cat_15/v_6gZuc4umTPk_4990_5755/0025s_00748_mask.png", "mask_status": "available" } ], "obj_id": [ 0 ], "anno_id": [ 941 ], "fps_inferred_from_clip_interval": 30.0 }, "usage": { "input_tokens": 27979, "input_tokens_details": { "cache_write_tokens": 27976, "cached_tokens": 0 }, "output_tokens": 3904, "output_tokens_details": { "reasoning_tokens": 2523 }, "total_tokens": 31883 }, "model": "gpt-5.6-sol", "effort": "medium", "custom_id": "cat15_cap12_v_6gZuc4umTPk_4990_5755", "annotation": { "caption": "The person reaches over to take another person's hand. They hold the hand steady and begin to apply the purple nail polish to a finger.", "entity_groups": [ { "id": "person_1", "label": "person", "spans": [ { "text": "The person", "range": [ 0, 10 ] }, { "text": "They", "range": [ 55, 59 ] } ], "source_mask_link": "selected_person" }, { "id": "person_2", "label": "person", "spans": [ { "text": "another person's", "range": [ 32, 48 ] } ], "source_mask_link": null }, { "id": "hand_1", "label": "hand", "spans": [ { "text": "another person's hand", "range": [ 32, 53 ] }, { "text": "the hand", "range": [ 65, 73 ] } ], "source_mask_link": null }, { "id": "nail_polish_1", "label": "nail polish", "spans": [ { "text": "the purple nail polish", "range": [ 100, 122 ] } ], "source_mask_link": null }, { "id": "finger_1", "label": "finger", "spans": [ { "text": "a finger", "range": [ 126, 134 ] } ], "source_mask_link": null } ], "actions": [ { "id": "action_1", "mentions": [ { "text": "reaches", "ranges": [ [ 11, 18 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "type": "simple", "subject_refs": [ "person_1" ], "object_refs": [ "hand_1" ], "confidence": "medium", "temporal_status": "localized_from_sparse_samples", "intervals": [ { "start_s": 0.0, "end_s": 3.0, "source_start_s": 166.33, "source_end_s": 169.33 } ], "evidence": "At clip-relative 0–2 seconds, the mask-linked person extends an arm across the table toward the other person's hand.", "mask_relation": "The action subject matches the supplied selected-person mask.", "source_subject_refs": [ "person_1" ] }, { "id": "action_2", "mentions": [ { "text": "take", "ranges": [ [ 27, 31 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "type": "simple", "subject_refs": [ "person_1" ], "object_refs": [ "hand_1" ], "confidence": "medium", "temporal_status": "localized_from_sparse_samples", "intervals": [ { "start_s": 2.0, "end_s": 4.0, "source_start_s": 168.33, "source_end_s": 170.33 } ], "evidence": "At 2–3 seconds, the mask-linked person reaches the other person's extended hand and makes grasping contact.", "mask_relation": "The action subject matches the supplied selected-person mask.", "source_subject_refs": [ "person_1" ] }, { "id": "action_3", "mentions": [ { "text": "hold", "ranges": [ [ 60, 64 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "type": "simple", "subject_refs": [ "person_1" ], "object_refs": [ "hand_1" ], "confidence": "high", "temporal_status": "localized_from_sparse_samples", "intervals": [ { "start_s": 3.0, "end_s": 4.0, "source_start_s": 169.33, "source_end_s": 170.33 }, { "start_s": 17.0, "end_s": 25.5, "source_start_s": 183.33, "source_end_s": 191.83 } ], "evidence": "The selected person briefly holds the recipient's hand at 3 seconds and visibly stabilizes the fingers throughout the polish application from 17 seconds to the clip boundary.", "mask_relation": "The action subject matches the supplied selected-person mask.", "source_subject_refs": [ "person_1" ] }, { "id": "action_4", "mentions": [ { "text": "apply", "ranges": [ [ 94, 99 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "type": "simple", "subject_refs": [ "person_1" ], "object_refs": [ "nail_polish_1", "finger_1" ], "confidence": "high", "temporal_status": "localized_from_sparse_samples", "intervals": [ { "start_s": 18.0, "end_s": 25.5, "source_start_s": 184.33, "source_end_s": 191.83 } ], "evidence": "From 18 seconds through the final 25-second sample, the mask-linked person repeatedly brings the polish applicator into contact with the other person's fingernails.", "mask_relation": "The action subject matches the supplied selected-person mask.", "source_subject_refs": [ "person_1" ] } ], "notes": [ "The supplied mask consistently identifies the dark-haired person on the left, who performs the captioned actions.", "The other person, hand, finger, and nail polish are text-groundable entities but have no independently supplied source masks.", "\"over\" is directional and excluded from the reach span. \"begin\" is an aspectual framing verb and is excluded; the embedded action \"apply\" is annotated.", "Temporal intervals are sparse sampled-frame support, not exact continuous-video onset and offset boundaries." ] }, "model_output_raw": { "caption": "The person reaches over to take another person's hand. They hold the hand steady and begin to apply the purple nail polish to a finger.", "entity_groups": [ { "id": "person_1", "label": "person", "spans": [ { "text": "The person", "range": [ 0, 10 ] }, { "text": "They", "range": [ 55, 59 ] } ], "source_mask": "selected_person_masklet" }, { "id": "person_2", "label": "person", "spans": [ { "text": "another person's", "range": [ 32, 48 ] } ], "source_mask": null }, { "id": "hand_1", "label": "hand", "spans": [ { "text": "another person's hand", "range": [ 32, 53 ] }, { "text": "the hand", "range": [ 65, 73 ] } ], "source_mask": null }, { "id": "nail_polish_1", "label": "nail polish", "spans": [ { "text": "the purple nail polish", "range": [ 100, 122 ] } ], "source_mask": null }, { "id": "finger_1", "label": "finger", "spans": [ { "text": "a finger", "range": [ 126, 134 ] } ], "source_mask": null } ], "actions": [ { "id": "action_1", "mentions": [ { "text": "reaches", "ranges": [ [ 11, 18 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "type": "simple", "subject_refs": [ "person_1" ], "object_refs": [ "hand_1" ], "confidence": "medium", "temporal_status": "localized_from_sparse_samples", "intervals": [ { "start_s": 0.0, "end_s": 3.0 } ], "evidence": "At clip-relative 0–2 seconds, the mask-linked person extends an arm across the table toward the other person's hand.", "mask_relation": "The action subject matches the supplied selected-person mask." }, { "id": "action_2", "mentions": [ { "text": "take", "ranges": [ [ 27, 31 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "type": "simple", "subject_refs": [ "person_1" ], "object_refs": [ "hand_1" ], "confidence": "medium", "temporal_status": "localized_from_sparse_samples", "intervals": [ { "start_s": 2.0, "end_s": 4.0 } ], "evidence": "At 2–3 seconds, the mask-linked person reaches the other person's extended hand and makes grasping contact.", "mask_relation": "The action subject matches the supplied selected-person mask." }, { "id": "action_3", "mentions": [ { "text": "hold", "ranges": [ [ 60, 64 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "type": "simple", "subject_refs": [ "person_1" ], "object_refs": [ "hand_1" ], "confidence": "high", "temporal_status": "localized_from_sparse_samples", "intervals": [ { "start_s": 3.0, "end_s": 4.0 }, { "start_s": 17.0, "end_s": 25.5 } ], "evidence": "The selected person briefly holds the recipient's hand at 3 seconds and visibly stabilizes the fingers throughout the polish application from 17 seconds to the clip boundary.", "mask_relation": "The action subject matches the supplied selected-person mask." }, { "id": "action_4", "mentions": [ { "text": "apply", "ranges": [ [ 94, 99 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "type": "simple", "subject_refs": [ "person_1" ], "object_refs": [ "nail_polish_1", "finger_1" ], "confidence": "high", "temporal_status": "localized_from_sparse_samples", "intervals": [ { "start_s": 18.0, "end_s": 25.5 } ], "evidence": "From 18 seconds through the final 25-second sample, the mask-linked person repeatedly brings the polish applicator into contact with the other person's fingernails.", "mask_relation": "The action subject matches the supplied selected-person mask." } ], "notes": [ "The supplied mask consistently identifies the dark-haired person on the left, who performs the captioned actions.", "The other person, hand, finger, and nail polish are text-groundable entities but have no independently supplied source masks.", "\"over\" is directional and excluded from the reach span. \"begin\" is an aspectual framing verb and is excluded; the embedded action \"apply\" is annotated.", "Temporal intervals are sparse sampled-frame support, not exact continuous-video onset and offset boundaries." ] }, "normalization_repairs": 0 }, { "source_video_id": "v_6gZuc4umTPk", "caption_count_category": 15, "clip": { "caption_index": 13, "caption": "The person holds the other person's hand up, blows on the freshly painted nails, and examines their work. They then release the hand and look forward.", "source_interval_sec": [ 191.83, 212.33 ], "loom_clip_id": "v_6gZuc4umTPk_5755_6370", "clip_frame_interval_4fps": [ 5755, 6370 ], "sample_count_1fps": 21, "sampled_frames": [ { "frame": "00000", "timestamp_sec": 191.83, "offset_sec": 0, "rgb_path": "assets/cat_15/v_6gZuc4umTPk_5755_6370/0000s_00000_rgb.jpg", "mask_path": "assets/cat_15/v_6gZuc4umTPk_5755_6370/0000s_00000_mask.png", "mask_status": "available" }, { "frame": "00028", "timestamp_sec": 192.83, "offset_sec": 1, "rgb_path": "assets/cat_15/v_6gZuc4umTPk_5755_6370/0001s_00028_rgb.jpg", "mask_path": "assets/cat_15/v_6gZuc4umTPk_5755_6370/0001s_00028_mask.png", "mask_status": "available" }, { "frame": "00060", "timestamp_sec": 193.83, "offset_sec": 2, "rgb_path": "assets/cat_15/v_6gZuc4umTPk_5755_6370/0002s_00060_rgb.jpg", "mask_path": "assets/cat_15/v_6gZuc4umTPk_5755_6370/0002s_00060_mask.png", "mask_status": "available" }, { "frame": "00088", "timestamp_sec": 194.83, "offset_sec": 3, "rgb_path": "assets/cat_15/v_6gZuc4umTPk_5755_6370/0003s_00088_rgb.jpg", "mask_path": "assets/cat_15/v_6gZuc4umTPk_5755_6370/0003s_00088_mask.png", "mask_status": "available" }, { "frame": "00120", "timestamp_sec": 195.83, "offset_sec": 4, "rgb_path": "assets/cat_15/v_6gZuc4umTPk_5755_6370/0004s_00120_rgb.jpg", "mask_path": "assets/cat_15/v_6gZuc4umTPk_5755_6370/0004s_00120_mask.png", "mask_status": "available" }, { "frame": "00148", "timestamp_sec": 196.83, "offset_sec": 5, "rgb_path": "assets/cat_15/v_6gZuc4umTPk_5755_6370/0005s_00148_rgb.jpg", "mask_path": "assets/cat_15/v_6gZuc4umTPk_5755_6370/0005s_00148_mask.png", "mask_status": "available" }, { "frame": "00180", "timestamp_sec": 197.83, "offset_sec": 6, "rgb_path": "assets/cat_15/v_6gZuc4umTPk_5755_6370/0006s_00180_rgb.jpg", "mask_path": "assets/cat_15/v_6gZuc4umTPk_5755_6370/0006s_00180_mask.png", "mask_status": "available" }, { "frame": "00208", "timestamp_sec": 198.83, "offset_sec": 7, "rgb_path": "assets/cat_15/v_6gZuc4umTPk_5755_6370/0007s_00208_rgb.jpg", "mask_path": "assets/cat_15/v_6gZuc4umTPk_5755_6370/0007s_00208_mask.png", "mask_status": "available" }, { "frame": "00240", "timestamp_sec": 199.83, "offset_sec": 8, "rgb_path": "assets/cat_15/v_6gZuc4umTPk_5755_6370/0008s_00240_rgb.jpg", "mask_path": "assets/cat_15/v_6gZuc4umTPk_5755_6370/0008s_00240_mask.png", "mask_status": "available" }, { "frame": "00268", "timestamp_sec": 200.83, "offset_sec": 9, "rgb_path": "assets/cat_15/v_6gZuc4umTPk_5755_6370/0009s_00268_rgb.jpg", "mask_path": "assets/cat_15/v_6gZuc4umTPk_5755_6370/0009s_00268_mask.png", "mask_status": "available" }, { "frame": "00300", "timestamp_sec": 201.83, "offset_sec": 10, "rgb_path": "assets/cat_15/v_6gZuc4umTPk_5755_6370/0010s_00300_rgb.jpg", "mask_path": "assets/cat_15/v_6gZuc4umTPk_5755_6370/0010s_00300_mask.png", "mask_status": "available" }, { "frame": "00328", "timestamp_sec": 202.83, "offset_sec": 11, "rgb_path": "assets/cat_15/v_6gZuc4umTPk_5755_6370/0011s_00328_rgb.jpg", "mask_path": "assets/cat_15/v_6gZuc4umTPk_5755_6370/0011s_00328_mask.png", "mask_status": "available" }, { "frame": "00360", "timestamp_sec": 203.83, "offset_sec": 12, "rgb_path": "assets/cat_15/v_6gZuc4umTPk_5755_6370/0012s_00360_rgb.jpg", "mask_path": "assets/cat_15/v_6gZuc4umTPk_5755_6370/0012s_00360_mask.png", "mask_status": "available" }, { "frame": "00388", "timestamp_sec": 204.83, "offset_sec": 13, "rgb_path": "assets/cat_15/v_6gZuc4umTPk_5755_6370/0013s_00388_rgb.jpg", "mask_path": "assets/cat_15/v_6gZuc4umTPk_5755_6370/0013s_00388_mask.png", "mask_status": "available" }, { "frame": "00420", "timestamp_sec": 205.83, "offset_sec": 14, "rgb_path": "assets/cat_15/v_6gZuc4umTPk_5755_6370/0014s_00420_rgb.jpg", "mask_path": "assets/cat_15/v_6gZuc4umTPk_5755_6370/0014s_00420_mask.png", "mask_status": "available" }, { "frame": "00448", "timestamp_sec": 206.83, "offset_sec": 15, "rgb_path": "assets/cat_15/v_6gZuc4umTPk_5755_6370/0015s_00448_rgb.jpg", "mask_path": "assets/cat_15/v_6gZuc4umTPk_5755_6370/0015s_00448_mask.png", "mask_status": "available" }, { "frame": "00480", "timestamp_sec": 207.83, "offset_sec": 16, "rgb_path": "assets/cat_15/v_6gZuc4umTPk_5755_6370/0016s_00480_rgb.jpg", "mask_path": "assets/cat_15/v_6gZuc4umTPk_5755_6370/0016s_00480_mask.png", "mask_status": "available" }, { "frame": "00508", "timestamp_sec": 208.83, "offset_sec": 17, "rgb_path": "assets/cat_15/v_6gZuc4umTPk_5755_6370/0017s_00508_rgb.jpg", "mask_path": "assets/cat_15/v_6gZuc4umTPk_5755_6370/0017s_00508_mask.png", "mask_status": "available" }, { "frame": "00540", "timestamp_sec": 209.83, "offset_sec": 18, "rgb_path": "assets/cat_15/v_6gZuc4umTPk_5755_6370/0018s_00540_rgb.jpg", "mask_path": "assets/cat_15/v_6gZuc4umTPk_5755_6370/0018s_00540_mask.png", "mask_status": "available" }, { "frame": "00568", "timestamp_sec": 210.83, "offset_sec": 19, "rgb_path": "assets/cat_15/v_6gZuc4umTPk_5755_6370/0019s_00568_rgb.jpg", "mask_path": "assets/cat_15/v_6gZuc4umTPk_5755_6370/0019s_00568_mask.png", "mask_status": "available" }, { "frame": "00600", "timestamp_sec": 211.83, "offset_sec": 20, "rgb_path": "assets/cat_15/v_6gZuc4umTPk_5755_6370/0020s_00600_rgb.jpg", "mask_path": "assets/cat_15/v_6gZuc4umTPk_5755_6370/0020s_00600_mask.png", "mask_status": "available" } ], "obj_id": [ 0 ], "anno_id": [ 942 ], "fps_inferred_from_clip_interval": 30.0 }, "usage": { "input_tokens": 25547, "input_tokens_details": { "cache_write_tokens": 25544, "cached_tokens": 0 }, "output_tokens": 4559, "output_tokens_details": { "reasoning_tokens": 2971 }, "total_tokens": 30106 }, "model": "gpt-5.6-sol", "effort": "medium", "custom_id": "cat15_cap13_v_6gZuc4umTPk_5755_6370", "annotation": { "caption": "The person holds the other person's hand up, blows on the freshly painted nails, and examines their work. They then release the hand and look forward.", "entity_groups": [ { "id": "E1", "label": "person", "spans": [ { "text": "The person", "range": [ 0, 10 ] }, { "text": "their", "range": [ 94, 99 ] }, { "text": "They", "range": [ 106, 110 ] } ], "source_mask_link": "selected_person" }, { "id": "E2", "label": "other person", "spans": [ { "text": "the other person's", "range": [ 17, 35 ] } ], "source_mask_link": null }, { "id": "E3", "label": "other person's hand", "spans": [ { "text": "the other person's hand", "range": [ 17, 40 ] }, { "text": "the hand", "range": [ 124, 132 ] } ], "source_mask_link": null }, { "id": "E4", "label": "nails", "spans": [ { "text": "the freshly painted nails", "range": [ 54, 79 ] } ], "source_mask_link": null } ], "actions": [ { "id": "A1", "mentions": [ { "text": "holds", "ranges": [ [ 11, 16 ] ], "occurrence": 1, "particle": { "text": "up", "occurrence": 1, "ranges": [ [ 41, 43 ] ] } } ], "type": "phrasal", "subject_refs": [ "E1" ], "object_refs": [ "E3" ], "confidence": "high", "temporal_status": "supported", "intervals": [ { "start_s": 2.0, "end_s": 17.0, "source_start_s": 193.83, "source_end_s": 208.83 } ], "evidence": "From samples at 2–16 seconds, the mask-linked brunette person grips and keeps the other person's hand raised.", "mask_relation": "subject_matches_supplied_selected_person_mask", "source_subject_refs": "E1" }, { "id": "A2", "mentions": [ { "text": "blows", "ranges": [ [ 45, 50 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1, "ranges": [] } } ], "type": "prepositional", "subject_refs": [ "E1" ], "object_refs": [ "E4" ], "confidence": "medium", "temporal_status": "supported", "intervals": [ { "start_s": 3.0, "end_s": 6.0, "source_start_s": 194.83, "source_end_s": 197.83 } ], "evidence": "At 3–5 seconds, the mask-linked person holds the painted fingers immediately in front of her mouth with a blowing posture.", "mask_relation": "subject_matches_supplied_selected_person_mask", "source_subject_refs": "E1" }, { "id": "A3", "mentions": [ { "text": "examines", "ranges": [ [ 85, 93 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1, "ranges": [] } } ], "type": "simple", "subject_refs": [ "E1" ], "object_refs": [ "E4" ], "confidence": "high", "temporal_status": "supported", "intervals": [ { "start_s": 6.0, "end_s": 17.0, "source_start_s": 197.83, "source_end_s": 208.83 } ], "evidence": "Across samples at 6–16 seconds, the mask-linked person repeatedly looks closely at and turns the painted fingers.", "mask_relation": "subject_matches_supplied_selected_person_mask", "source_subject_refs": "E1" }, { "id": "A4", "mentions": [ { "text": "release", "ranges": [ [ 116, 123 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1, "ranges": [] } } ], "type": "simple", "subject_refs": [ "E1" ], "object_refs": [ "E3" ], "confidence": "high", "temporal_status": "supported", "intervals": [ { "start_s": 17.0, "end_s": 20.0, "source_start_s": 208.83, "source_end_s": 211.83 } ], "evidence": "At 17–19 seconds, the raised hand is lowered and the mask-linked person's grip separates from it.", "mask_relation": "subject_matches_supplied_selected_person_mask", "source_subject_refs": "E1" }, { "id": "A5", "mentions": [ { "text": "look", "ranges": [ [ 137, 141 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1, "ranges": [] } } ], "type": "simple", "subject_refs": [ "E1" ], "object_refs": [], "confidence": "medium", "temporal_status": "supported", "intervals": [ { "start_s": 20.0, "end_s": 20.5, "source_start_s": 211.83, "source_end_s": 212.33 } ], "evidence": "In the final sample at 20 seconds, the mask-linked person has raised her gaze toward the forward viewpoint.", "mask_relation": "subject_matches_supplied_selected_person_mask", "source_subject_refs": "E1" } ], "notes": [ "The supplied mask consistently identifies the brunette person on the left, corresponding to “The person” and singular “They.”", "No separate masks were supplied for the other person, the hand, or the nails, so their source-mask links are null.", "“their work” is treated as an abstract description of the painted-nail result rather than as a separate concrete entity; the examination target is linked to the nails group.", "“painted” describes the nails' resulting state and is not annotated as a separate action.", "Temporal intervals are sparse support windows derived from the sampled clean RGB frames; the final look interval is clipped to the 20.5-second clip boundary." ] }, "model_output_raw": { "caption": "The person holds the other person's hand up, blows on the freshly painted nails, and examines their work. They then release the hand and look forward.", "entity_groups": [ { "id": "E1", "entity": "person", "spans": [ { "text": "The person", "range": [ 0, 10 ] }, { "text": "their", "range": [ 94, 99 ] }, { "text": "They", "range": [ 106, 110 ] } ], "source_mask": "supplied_selected_person_mask" }, { "id": "E2", "entity": "other person", "spans": [ { "text": "the other person's", "range": [ 17, 35 ] } ], "source_mask": null }, { "id": "E3", "entity": "other person's hand", "spans": [ { "text": "the other person's hand", "range": [ 17, 40 ] }, { "text": "the hand", "range": [ 124, 132 ] } ], "source_mask": null }, { "id": "E4", "entity": "nails", "spans": [ { "text": "the freshly painted nails", "range": [ 54, 79 ] } ], "source_mask": null } ], "actions": [ { "id": "A1", "mentions": [ { "text": "holds", "occurrence": 1, "ranges": [ [ 11, 16 ] ], "particle": { "text": "up", "occurrence": 1, "ranges": [ [ 41, 43 ] ] } } ], "type": "phrasal", "subject": "E1", "objects": [ "E3" ], "confidence": "high", "mask_relation": "subject_matches_supplied_selected_person_mask", "temporal_status": "supported", "intervals": [ { "clip_relative_s": [ 2.0, 17.0 ], "source_video_s": [ 193.83, 208.83 ] } ], "evidence": "From samples at 2–16 seconds, the mask-linked brunette person grips and keeps the other person's hand raised." }, { "id": "A2", "mentions": [ { "text": "blows", "occurrence": 1, "ranges": [ [ 45, 50 ] ], "particle": { "text": null, "occurrence": 1, "ranges": [] } } ], "type": "prepositional", "subject": "E1", "objects": [ "E4" ], "confidence": "medium", "mask_relation": "subject_matches_supplied_selected_person_mask", "temporal_status": "supported", "intervals": [ { "clip_relative_s": [ 3.0, 6.0 ], "source_video_s": [ 194.83, 197.83 ] } ], "evidence": "At 3–5 seconds, the mask-linked person holds the painted fingers immediately in front of her mouth with a blowing posture." }, { "id": "A3", "mentions": [ { "text": "examines", "occurrence": 1, "ranges": [ [ 85, 93 ] ], "particle": { "text": null, "occurrence": 1, "ranges": [] } } ], "type": "simple", "subject": "E1", "objects": [ "E4" ], "object_expression": { "text": "their work", "range": [ 94, 104 ] }, "confidence": "high", "mask_relation": "subject_matches_supplied_selected_person_mask", "temporal_status": "supported", "intervals": [ { "clip_relative_s": [ 6.0, 17.0 ], "source_video_s": [ 197.83, 208.83 ] } ], "evidence": "Across samples at 6–16 seconds, the mask-linked person repeatedly looks closely at and turns the painted fingers." }, { "id": "A4", "mentions": [ { "text": "release", "occurrence": 1, "ranges": [ [ 116, 123 ] ], "particle": { "text": null, "occurrence": 1, "ranges": [] } } ], "type": "simple", "subject": "E1", "objects": [ "E3" ], "confidence": "high", "mask_relation": "subject_matches_supplied_selected_person_mask", "temporal_status": "supported", "intervals": [ { "clip_relative_s": [ 17.0, 20.0 ], "source_video_s": [ 208.83, 211.83 ] } ], "evidence": "At 17–19 seconds, the raised hand is lowered and the mask-linked person's grip separates from it." }, { "id": "A5", "mentions": [ { "text": "look", "occurrence": 1, "ranges": [ [ 137, 141 ] ], "particle": { "text": null, "occurrence": 1, "ranges": [] } } ], "type": "simple", "subject": "E1", "objects": [], "confidence": "medium", "mask_relation": "subject_matches_supplied_selected_person_mask", "temporal_status": "supported", "intervals": [ { "clip_relative_s": [ 20.0, 20.5 ], "source_video_s": [ 211.83, 212.33 ] } ], "evidence": "In the final sample at 20 seconds, the mask-linked person has raised her gaze toward the forward viewpoint." } ], "notes": [ "The supplied mask consistently identifies the brunette person on the left, corresponding to “The person” and singular “They.”", "No separate masks were supplied for the other person, the hand, or the nails, so their source-mask links are null.", "“their work” is treated as an abstract description of the painted-nail result rather than as a separate concrete entity; the examination target is linked to the nails group.", "“painted” describes the nails' resulting state and is not annotated as a separate action.", "Temporal intervals are sparse support windows derived from the sampled clean RGB frames; the final look interval is clipped to the 20.5-second clip boundary." ] }, "normalization_repairs": 0 }, { "source_video_id": "v_6gZuc4umTPk", "caption_count_category": 15, "clip": { "caption_index": 14, "caption": "The person picks up a small purple object and places it into a small plaid box. They then pick up the purple nail polish bottle and close the box.", "source_interval_sec": [ 212.33, 224.33 ], "loom_clip_id": "v_6gZuc4umTPk_6370_6730", "clip_frame_interval_4fps": [ 6370, 6730 ], "sample_count_1fps": 12, "sampled_frames": [ { "frame": "00000", "timestamp_sec": 212.33, "offset_sec": 0, "rgb_path": "assets/cat_15/v_6gZuc4umTPk_6370_6730/0000s_00000_rgb.jpg", "mask_path": "assets/cat_15/v_6gZuc4umTPk_6370_6730/0000s_00000_mask.png", "mask_status": "available" }, { "frame": "00028", "timestamp_sec": 213.33, "offset_sec": 1, "rgb_path": "assets/cat_15/v_6gZuc4umTPk_6370_6730/0001s_00028_rgb.jpg", "mask_path": "assets/cat_15/v_6gZuc4umTPk_6370_6730/0001s_00028_mask.png", "mask_status": "available" }, { "frame": "00060", "timestamp_sec": 214.33, "offset_sec": 2, "rgb_path": "assets/cat_15/v_6gZuc4umTPk_6370_6730/0002s_00060_rgb.jpg", "mask_path": "assets/cat_15/v_6gZuc4umTPk_6370_6730/0002s_00060_mask.png", "mask_status": "available" }, { "frame": "00088", "timestamp_sec": 215.33, "offset_sec": 3, "rgb_path": "assets/cat_15/v_6gZuc4umTPk_6370_6730/0003s_00088_rgb.jpg", "mask_path": "assets/cat_15/v_6gZuc4umTPk_6370_6730/0003s_00088_mask.png", "mask_status": "available" }, { "frame": "00120", "timestamp_sec": 216.33, "offset_sec": 4, "rgb_path": "assets/cat_15/v_6gZuc4umTPk_6370_6730/0004s_00120_rgb.jpg", "mask_path": "assets/cat_15/v_6gZuc4umTPk_6370_6730/0004s_00120_mask.png", "mask_status": "available" }, { "frame": "00148", "timestamp_sec": 217.33, "offset_sec": 5, "rgb_path": "assets/cat_15/v_6gZuc4umTPk_6370_6730/0005s_00148_rgb.jpg", "mask_path": "assets/cat_15/v_6gZuc4umTPk_6370_6730/0005s_00148_mask.png", "mask_status": "available" }, { "frame": "00180", "timestamp_sec": 218.33, "offset_sec": 6, "rgb_path": "assets/cat_15/v_6gZuc4umTPk_6370_6730/0006s_00180_rgb.jpg", "mask_path": "assets/cat_15/v_6gZuc4umTPk_6370_6730/0006s_00180_mask.png", "mask_status": "available" }, { "frame": "00208", "timestamp_sec": 219.33, "offset_sec": 7, "rgb_path": "assets/cat_15/v_6gZuc4umTPk_6370_6730/0007s_00208_rgb.jpg", "mask_path": "assets/cat_15/v_6gZuc4umTPk_6370_6730/0007s_00208_mask.png", "mask_status": "available" }, { "frame": "00240", "timestamp_sec": 220.33, "offset_sec": 8, "rgb_path": "assets/cat_15/v_6gZuc4umTPk_6370_6730/0008s_00240_rgb.jpg", "mask_path": "assets/cat_15/v_6gZuc4umTPk_6370_6730/0008s_00240_mask.png", "mask_status": "available" }, { "frame": "00268", "timestamp_sec": 221.33, "offset_sec": 9, "rgb_path": "assets/cat_15/v_6gZuc4umTPk_6370_6730/0009s_00268_rgb.jpg", "mask_path": "assets/cat_15/v_6gZuc4umTPk_6370_6730/0009s_00268_mask.png", "mask_status": "available" }, { "frame": "00300", "timestamp_sec": 222.33, "offset_sec": 10, "rgb_path": "assets/cat_15/v_6gZuc4umTPk_6370_6730/0010s_00300_rgb.jpg", "mask_path": "assets/cat_15/v_6gZuc4umTPk_6370_6730/0010s_00300_mask.png", "mask_status": "available" }, { "frame": "00328", "timestamp_sec": 223.33, "offset_sec": 11, "rgb_path": "assets/cat_15/v_6gZuc4umTPk_6370_6730/0011s_00328_rgb.jpg", "mask_path": "assets/cat_15/v_6gZuc4umTPk_6370_6730/0011s_00328_mask.png", "mask_status": "available" } ], "obj_id": [ 0 ], "anno_id": [ 943 ], "fps_inferred_from_clip_interval": 30.0 }, "usage": { "input_tokens": 21165, "input_tokens_details": { "cache_write_tokens": 21162, "cached_tokens": 0 }, "output_tokens": 3466, "output_tokens_details": { "reasoning_tokens": 2070 }, "total_tokens": 24631 }, "model": "gpt-5.6-sol", "effort": "medium", "custom_id": "cat15_cap14_v_6gZuc4umTPk_6370_6730", "annotation": { "caption": "The person picks up a small purple object and places it into a small plaid box. They then pick up the purple nail polish bottle and close the box.", "entity_groups": [ { "id": "entity_person_1", "label": "person", "spans": [ { "text": "The person", "range": [ 0, 10 ] }, { "text": "They", "range": [ 80, 84 ] } ], "source_mask_link": "selected_person" }, { "id": "entity_purple_object_1", "label": "object", "spans": [ { "text": "a small purple object", "range": [ 20, 41 ] }, { "text": "it", "range": [ 53, 55 ] } ], "source_mask_link": null }, { "id": "entity_box_1", "label": "box", "spans": [ { "text": "a small plaid box", "range": [ 61, 78 ] }, { "text": "the box", "range": [ 138, 145 ] } ], "source_mask_link": null }, { "id": "entity_nail_polish_bottle_1", "label": "bottle", "spans": [ { "text": "the purple nail polish bottle", "range": [ 98, 127 ] } ], "source_mask_link": null } ], "actions": [ { "id": "action_pick_up_object_1", "mentions": [ { "text": "picks up", "ranges": [ [ 11, 19 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "type": "phrasal", "subject_refs": [ "entity_person_1" ], "object_refs": [ "entity_purple_object_1" ], "confidence": "medium", "temporal_status": "localized", "intervals": [ { "start_s": 2.0, "end_s": 4.0, "source_start_s": 214.33, "source_end_s": 216.33 } ], "evidence": "At 2.0 seconds the masked person reaches toward the foreground, and at 3.0 seconds she visibly holds up a small purple object.", "mask_relation": "subject matches the supplied selected-person mask", "source_subject_refs": "entity_person_1" }, { "id": "action_place_object_1", "mentions": [ { "text": "places", "ranges": [ [ 46, 52 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "type": "simple", "subject_refs": [ "entity_person_1" ], "object_refs": [ "entity_purple_object_1" ], "confidence": "medium", "temporal_status": "localized", "intervals": [ { "start_s": 4.0, "end_s": 8.0, "source_start_s": 216.33, "source_end_s": 220.33 } ], "evidence": "From 4.0 through 7.0 seconds, the masked person lowers both hands and manipulates the object at the box area.", "mask_relation": "subject matches the supplied selected-person mask", "source_subject_refs": "entity_person_1" }, { "id": "action_pick_up_bottle_1", "mentions": [ { "text": "pick up", "ranges": [ [ 90, 97 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "type": "phrasal", "subject_refs": [ "entity_person_1" ], "object_refs": [ "entity_nail_polish_bottle_1" ], "confidence": "medium", "temporal_status": "localized", "intervals": [ { "start_s": 9.0, "end_s": 11.0, "source_start_s": 221.33, "source_end_s": 223.33 } ], "evidence": "The masked person reaches down at 9.0 seconds and visibly holds a purple bottle-like item at 10.0 seconds.", "mask_relation": "subject matches the supplied selected-person mask", "source_subject_refs": "entity_person_1" }, { "id": "action_close_box_1", "mentions": [ { "text": "close", "ranges": [ [ 132, 137 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "type": "simple", "subject_refs": [ "entity_person_1" ], "object_refs": [ "entity_box_1" ], "confidence": "medium", "temporal_status": "localized", "intervals": [ { "start_s": 11.0, "end_s": 12.0, "source_start_s": 223.33, "source_end_s": 224.33 } ], "evidence": "At 11.0 seconds the masked person makes a rapid two-handed closing motion around the box; support reaches the clip boundary.", "mask_relation": "subject matches the supplied selected-person mask", "source_subject_refs": "entity_person_1" } ], "notes": [ "The supplied masklet consistently identifies the person described by “The person” and “They.”", "No separate masks were supplied for the small purple object, plaid box, or nail polish bottle, so their source-mask links are null.", "An additional unmasked arm is visible in the earliest frames, but no separate person is mentioned in the caption and no action was transferred from that arm to the masked subject.", "Temporal intervals are sparse sampled-frame support intervals rather than exact continuous-video action boundaries." ] }, "model_output_raw": { "caption": "The person picks up a small purple object and places it into a small plaid box. They then pick up the purple nail polish bottle and close the box.", "entity_groups": [ { "id": "entity_person_1", "entity_type": "person", "spans": [ { "text": "The person", "range": [ 0, 10 ] }, { "text": "They", "range": [ 80, 84 ] } ], "source_mask": "selected_person_masklet" }, { "id": "entity_purple_object_1", "entity_type": "object", "spans": [ { "text": "a small purple object", "range": [ 20, 41 ] }, { "text": "it", "range": [ 53, 55 ] } ], "source_mask": null }, { "id": "entity_box_1", "entity_type": "box", "spans": [ { "text": "a small plaid box", "range": [ 61, 78 ] }, { "text": "the box", "range": [ 138, 145 ] } ], "source_mask": null }, { "id": "entity_nail_polish_bottle_1", "entity_type": "bottle", "spans": [ { "text": "the purple nail polish bottle", "range": [ 98, 127 ] } ], "source_mask": null } ], "actions": [ { "id": "action_pick_up_object_1", "mentions": [ { "text": "picks up", "occurrence": 1, "ranges": [ [ 11, 19 ] ], "particle": { "text": null, "occurrence": 1 } } ], "type": "phrasal", "subject_ref": "entity_person_1", "object_refs": [ "entity_purple_object_1" ], "confidence": "medium", "temporal_status": "localized", "intervals": [ { "start_s": 2.0, "end_s": 4.0, "source_start_s": 214.33, "source_end_s": 216.33 } ], "mask_relation": "subject matches the supplied selected-person mask", "evidence": "At 2.0 seconds the masked person reaches toward the foreground, and at 3.0 seconds she visibly holds up a small purple object." }, { "id": "action_place_object_1", "mentions": [ { "text": "places", "occurrence": 1, "ranges": [ [ 46, 52 ] ], "particle": { "text": null, "occurrence": 1 } } ], "type": "simple", "subject_ref": "entity_person_1", "object_refs": [ "entity_purple_object_1" ], "destination_ref": "entity_box_1", "confidence": "medium", "temporal_status": "localized", "intervals": [ { "start_s": 4.0, "end_s": 8.0, "source_start_s": 216.33, "source_end_s": 220.33 } ], "mask_relation": "subject matches the supplied selected-person mask", "evidence": "From 4.0 through 7.0 seconds, the masked person lowers both hands and manipulates the object at the box area." }, { "id": "action_pick_up_bottle_1", "mentions": [ { "text": "pick up", "occurrence": 1, "ranges": [ [ 90, 97 ] ], "particle": { "text": null, "occurrence": 1 } } ], "type": "phrasal", "subject_ref": "entity_person_1", "object_refs": [ "entity_nail_polish_bottle_1" ], "confidence": "medium", "temporal_status": "localized", "intervals": [ { "start_s": 9.0, "end_s": 11.0, "source_start_s": 221.33, "source_end_s": 223.33 } ], "mask_relation": "subject matches the supplied selected-person mask", "evidence": "The masked person reaches down at 9.0 seconds and visibly holds a purple bottle-like item at 10.0 seconds." }, { "id": "action_close_box_1", "mentions": [ { "text": "close", "occurrence": 1, "ranges": [ [ 132, 137 ] ], "particle": { "text": null, "occurrence": 1 } } ], "type": "simple", "subject_ref": "entity_person_1", "object_refs": [ "entity_box_1" ], "confidence": "medium", "temporal_status": "localized", "intervals": [ { "start_s": 11.0, "end_s": 12.0, "source_start_s": 223.33, "source_end_s": 224.33 } ], "mask_relation": "subject matches the supplied selected-person mask", "evidence": "At 11.0 seconds the masked person makes a rapid two-handed closing motion around the box; support reaches the clip boundary." } ], "notes": [ "The supplied masklet consistently identifies the person described by “The person” and “They.”", "No separate masks were supplied for the small purple object, plaid box, or nail polish bottle, so their source-mask links are null.", "An additional unmasked arm is visible in the earliest frames, but no separate person is mentioned in the caption and no action was transferred from that arm to the masked subject.", "Temporal intervals are sparse sampled-frame support intervals rather than exact continuous-video action boundaries." ] }, "normalization_repairs": 0 }, { "source_video_id": "v_6gZuc4umTPk", "caption_count_category": 15, "clip": { "caption_index": 15, "caption": "The person picks up a yellow object and puts it into the box. They then give a thumbs-up gesture before waving goodbye.", "source_interval_sec": [ 224.33, 233.0 ], "loom_clip_id": "v_6gZuc4umTPk_6730_6990", "clip_frame_interval_4fps": [ 6730, 6990 ], "sample_count_1fps": 9, "sampled_frames": [ { "frame": "00000", "timestamp_sec": 224.33, "offset_sec": 0, "rgb_path": "assets/cat_15/v_6gZuc4umTPk_6730_6990/0000s_00000_rgb.jpg", "mask_path": "assets/cat_15/v_6gZuc4umTPk_6730_6990/0000s_00000_mask.png", "mask_status": "available" }, { "frame": "00028", "timestamp_sec": 225.33, "offset_sec": 1, "rgb_path": "assets/cat_15/v_6gZuc4umTPk_6730_6990/0001s_00028_rgb.jpg", "mask_path": "assets/cat_15/v_6gZuc4umTPk_6730_6990/0001s_00028_mask.png", "mask_status": "available" }, { "frame": "00060", "timestamp_sec": 226.33, "offset_sec": 2, "rgb_path": "assets/cat_15/v_6gZuc4umTPk_6730_6990/0002s_00060_rgb.jpg", "mask_path": "assets/cat_15/v_6gZuc4umTPk_6730_6990/0002s_00060_mask.png", "mask_status": "available" }, { "frame": "00088", "timestamp_sec": 227.33, "offset_sec": 3, "rgb_path": "assets/cat_15/v_6gZuc4umTPk_6730_6990/0003s_00088_rgb.jpg", "mask_path": "assets/cat_15/v_6gZuc4umTPk_6730_6990/0003s_00088_mask.png", "mask_status": "available" }, { "frame": "00120", "timestamp_sec": 228.33, "offset_sec": 4, "rgb_path": "assets/cat_15/v_6gZuc4umTPk_6730_6990/0004s_00120_rgb.jpg", "mask_path": "assets/cat_15/v_6gZuc4umTPk_6730_6990/0004s_00120_mask.png", "mask_status": "available" }, { "frame": "00148", "timestamp_sec": 229.33, "offset_sec": 5, "rgb_path": "assets/cat_15/v_6gZuc4umTPk_6730_6990/0005s_00148_rgb.jpg", "mask_path": "assets/cat_15/v_6gZuc4umTPk_6730_6990/0005s_00148_mask.png", "mask_status": "available" }, { "frame": "00180", "timestamp_sec": 230.33, "offset_sec": 6, "rgb_path": "assets/cat_15/v_6gZuc4umTPk_6730_6990/0006s_00180_rgb.jpg", "mask_path": "assets/cat_15/v_6gZuc4umTPk_6730_6990/0006s_00180_mask.png", "mask_status": "available" }, { "frame": "00208", "timestamp_sec": 231.33, "offset_sec": 7, "rgb_path": "assets/cat_15/v_6gZuc4umTPk_6730_6990/0007s_00208_rgb.jpg", "mask_path": "assets/cat_15/v_6gZuc4umTPk_6730_6990/0007s_00208_mask.png", "mask_status": "available" }, { "frame": "00240", "timestamp_sec": 232.33, "offset_sec": 8, "rgb_path": "assets/cat_15/v_6gZuc4umTPk_6730_6990/0008s_00240_rgb.jpg", "mask_path": "assets/cat_15/v_6gZuc4umTPk_6730_6990/0008s_00240_mask.png", "mask_status": "available" } ], "obj_id": [ 0 ], "anno_id": [ 944 ], "fps_inferred_from_clip_interval": 29.9885 }, "usage": { "input_tokens": 19698, "input_tokens_details": { "cache_write_tokens": 19695, "cached_tokens": 0 }, "output_tokens": 3368, "output_tokens_details": { "reasoning_tokens": 2070 }, "total_tokens": 23066 }, "model": "gpt-5.6-sol", "effort": "medium", "custom_id": "cat15_cap15_v_6gZuc4umTPk_6730_6990", "annotation": { "caption": "The person picks up a yellow object and puts it into the box. They then give a thumbs-up gesture before waving goodbye.", "entity_groups": [ { "id": "entity_person", "label": "person", "spans": [ { "text": "The person", "range": [ 0, 10 ] }, { "text": "They", "range": [ 62, 66 ] } ], "source_mask_link": "selected_person" }, { "id": "entity_yellow_object", "label": "yellow object", "spans": [ { "text": "a yellow object", "range": [ 20, 35 ] }, { "text": "it", "range": [ 45, 47 ] } ], "source_mask_link": null }, { "id": "entity_box", "label": "box", "spans": [ { "text": "the box", "range": [ 53, 60 ] } ], "source_mask_link": null } ], "actions": [ { "id": "action_pick_up", "mentions": [ { "text": "picks up", "ranges": [ [ 11, 19 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "type": "phrasal", "subject_refs": [ "entity_person" ], "object_refs": [ "entity_yellow_object" ], "confidence": "high", "temporal_status": "supported", "intervals": [ { "start_s": 1.0, "end_s": 4.0, "source_start_s": 225.33, "source_end_s": 228.33 } ], "evidence": "At 1.0–2.0 seconds the masked person reaches and handles an item near the box; at 3.0 seconds the yellow object is visibly lifted in her hands.", "mask_relation": "subject_matches_selected_person_mask", "source_subject_refs": [ "entity_person" ] }, { "id": "action_put", "mentions": [ { "text": "puts", "ranges": [ [ 40, 44 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "type": "prepositional", "subject_refs": [ "entity_person" ], "object_refs": [ "entity_yellow_object" ], "confidence": "high", "temporal_status": "supported", "intervals": [ { "start_s": 3.0, "end_s": 5.0, "source_start_s": 227.33, "source_end_s": 229.33 } ], "evidence": "At 3.0 seconds the masked person holds the yellow object at the open box, and by 4.0 seconds it is no longer visible while her hands remain over the box.", "mask_relation": "subject_matches_selected_person_mask", "source_subject_refs": [ "entity_person" ] }, { "id": "action_thumbs_up", "mentions": [ { "text": "give a thumbs-up gesture", "ranges": [ [ 72, 96 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "type": "light_verb", "subject_refs": [ "entity_person" ], "object_refs": [], "confidence": "high", "temporal_status": "supported", "intervals": [ { "start_s": 5.0, "end_s": 6.0, "source_start_s": 229.33, "source_end_s": 230.33 } ], "evidence": "At 5.0 seconds the masked person clearly holds a raised thumb toward the viewer.", "mask_relation": "subject_matches_selected_person_mask", "source_subject_refs": [ "entity_person" ] }, { "id": "action_wave", "mentions": [ { "text": "waving", "ranges": [ [ 104, 110 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "type": "simple", "subject_refs": [ "entity_person" ], "object_refs": [], "confidence": "high", "temporal_status": "supported", "intervals": [ { "start_s": 7.0, "end_s": 8.669999999999987, "source_start_s": 231.33, "source_end_s": 233.0 } ], "evidence": "At 7.0 seconds the masked person's hands move close to the camera, and at 8.0 seconds an open hand is clearly shown in a goodbye wave continuing to the clip boundary.", "mask_relation": "subject_matches_selected_person_mask", "source_subject_refs": [ "entity_person" ] } ], "notes": [ "The supplied masklet consistently identifies the person described by both “The person” and “They.”", "No separate source masks were supplied for the yellow object or the box.", "The phrase “a thumbs-up gesture” names an action rather than a separate tangible entity, so it is included in the light-verb action span but not assigned an entity group.", "Temporal intervals are sparse clip-relative support inferred only from the sampled clean RGB frames." ] }, "model_output_raw": { "caption": "The person picks up a yellow object and puts it into the box. They then give a thumbs-up gesture before waving goodbye.", "entity_groups": [ { "id": "entity_person", "label": "person", "spans": [ { "text": "The person", "start": 0, "end": 10 }, { "text": "They", "start": 62, "end": 66 } ], "source_mask_link": "selected_person_masklet" }, { "id": "entity_yellow_object", "label": "yellow object", "spans": [ { "text": "a yellow object", "start": 20, "end": 35 }, { "text": "it", "start": 45, "end": 47 } ], "source_mask_link": null }, { "id": "entity_box", "label": "box", "spans": [ { "text": "the box", "start": 53, "end": 60 } ], "source_mask_link": null } ], "actions": [ { "id": "action_pick_up", "mentions": [ { "text": "picks up", "ranges": [ { "start": 11, "end": 19 } ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "type": "phrasal", "subject_refs": [ "entity_person" ], "object_refs": [ "entity_yellow_object" ], "confidence": "high", "temporal_status": "supported", "intervals": [ { "start_s": 1.0, "end_s": 4.0 } ], "evidence": "At 1.0–2.0 seconds the masked person reaches and handles an item near the box; at 3.0 seconds the yellow object is visibly lifted in her hands.", "mask_relation": "subject_matches_selected_person_mask" }, { "id": "action_put", "mentions": [ { "text": "puts", "ranges": [ { "start": 40, "end": 44 } ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "type": "prepositional", "subject_refs": [ "entity_person" ], "object_refs": [ "entity_yellow_object" ], "related_entity_refs": [ { "entity_id": "entity_box", "role": "destination" } ], "confidence": "high", "temporal_status": "supported", "intervals": [ { "start_s": 3.0, "end_s": 5.0 } ], "evidence": "At 3.0 seconds the masked person holds the yellow object at the open box, and by 4.0 seconds it is no longer visible while her hands remain over the box.", "mask_relation": "subject_matches_selected_person_mask" }, { "id": "action_thumbs_up", "mentions": [ { "text": "give a thumbs-up gesture", "ranges": [ { "start": 72, "end": 96 } ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "type": "light_verb", "subject_refs": [ "entity_person" ], "object_refs": [], "confidence": "high", "temporal_status": "supported", "intervals": [ { "start_s": 5.0, "end_s": 6.0 } ], "evidence": "At 5.0 seconds the masked person clearly holds a raised thumb toward the viewer.", "mask_relation": "subject_matches_selected_person_mask" }, { "id": "action_wave", "mentions": [ { "text": "waving", "ranges": [ { "start": 104, "end": 110 } ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "type": "simple", "subject_refs": [ "entity_person" ], "object_refs": [], "confidence": "high", "temporal_status": "supported", "intervals": [ { "start_s": 7.0, "end_s": 8.67 } ], "evidence": "At 7.0 seconds the masked person's hands move close to the camera, and at 8.0 seconds an open hand is clearly shown in a goodbye wave continuing to the clip boundary.", "mask_relation": "subject_matches_selected_person_mask" } ], "notes": [ "The supplied masklet consistently identifies the person described by both “The person” and “They.”", "No separate source masks were supplied for the yellow object or the box.", "The phrase “a thumbs-up gesture” names an action rather than a separate tangible entity, so it is included in the light-verb action span but not assigned an entity group.", "Temporal intervals are sparse clip-relative support inferred only from the sampled clean RGB frames." ] }, "normalization_repairs": 0 } ], "video_action_groups": [ { "source_video_id": "v_--veKG73Di4", "grouping_rule": "Exact mention phrase set after case/whitespace normalization; action type is retained as metadata but does not split groups.", "actions": [ { "action": "holding", "types": [ "simple" ], "segments": [ { "caption_index": 1, "loom_clip_id": "v_--veKG73Di4_1_300", "action_id": "action_2", "mentions": [ { "text": "holding", "ranges": [ [ 74, 81 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "interval": { "start_s": 0.0, "end_s": 10.0, "source_start_s": 0.0, "source_end_s": 10.0 }, "confidence": "high", "temporal_status": "supported" }, { "caption_index": 4, "loom_clip_id": "v_--veKG73Di4_959_1753", "action_id": "action_3", "mentions": [ { "text": "holding", "ranges": [ [ 71, 78 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "interval": { "start_s": 0.0, "end_s": 2.0, "source_start_s": 32.0, "source_end_s": 34.0 }, "confidence": "medium", "temporal_status": "localized" } ], "id": "action_group_001", "segment_count": 2, "caption_indices": [ 1, 4 ] }, { "action": "smiles", "types": [ "simple" ], "segments": [ { "caption_index": 1, "loom_clip_id": "v_--veKG73Di4_1_300", "action_id": "action_3", "mentions": [ { "text": "smiles", "ranges": [ [ 151, 157 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "interval": { "start_s": 0.0, "end_s": 10.0, "source_start_s": 0.0, "source_end_s": 10.0 }, "confidence": "high", "temporal_status": "supported" } ], "id": "action_group_002", "segment_count": 1, "caption_indices": [ 1 ] }, { "action": "stands", "types": [ "posture" ], "segments": [ { "caption_index": 1, "loom_clip_id": "v_--veKG73Di4_1_300", "action_id": "action_1", "mentions": [ { "text": "stands", "ranges": [ [ 61, 67 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "interval": { "start_s": 0.0, "end_s": 10.0, "source_start_s": 0.0, "source_end_s": 10.0 }, "confidence": "high", "temporal_status": "supported" } ], "id": "action_group_003", "segment_count": 1, "caption_indices": [ 1 ] }, { "action": "raises", "types": [ "simple" ], "segments": [ { "caption_index": 1, "loom_clip_id": "v_--veKG73Di4_1_300", "action_id": "action_4", "mentions": [ { "text": "raises", "ranges": [ [ 159, 165 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "interval": { "start_s": 2.0, "end_s": 3.0, "source_start_s": 2.0, "source_end_s": 3.0 }, "confidence": "high", "temporal_status": "localized_sparse" } ], "id": "action_group_004", "segment_count": 1, "caption_indices": [ 1 ] }, { "action": "wave", "types": [ "simple" ], "segments": [ { "caption_index": 1, "loom_clip_id": "v_--veKG73Di4_1_300", "action_id": "action_5", "mentions": [ { "text": "wave", "ranges": [ [ 183, 187 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "interval": { "start_s": 2.0, "end_s": 4.0, "source_start_s": 2.0, "source_end_s": 4.0 }, "confidence": "high", "temporal_status": "localized_sparse" } ], "id": "action_group_005", "segment_count": 1, "caption_indices": [ 1 ] }, { "action": "lowers", "types": [ "simple" ], "segments": [ { "caption_index": 1, "loom_clip_id": "v_--veKG73Di4_1_300", "action_id": "action_6", "mentions": [ { "text": "lowers", "ranges": [ [ 198, 204 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "interval": { "start_s": 4.0, "end_s": 5.0, "source_start_s": 4.0, "source_end_s": 5.0 }, "confidence": "high", "temporal_status": "localized_sparse" } ], "id": "action_group_006", "segment_count": 1, "caption_indices": [ 1 ] }, { "action": "turns", "types": [ "simple" ], "segments": [ { "caption_index": 2, "loom_clip_id": "v_--veKG73Di4_300_764", "action_id": "A1", "mentions": [ { "text": "turns", "ranges": [ [ 4, 9 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "interval": { "start_s": 0.0, "end_s": 2.0, "source_start_s": 10.0, "source_end_s": 12.0 }, "confidence": "high", "temporal_status": "localized" }, { "caption_index": 2, "loom_clip_id": "v_--veKG73Di4_300_764", "action_id": "A5", "mentions": [ { "text": "turns", "ranges": [ [ 119, 124 ] ], "occurrence": 2, "particle": { "text": null, "occurrence": 1 } } ], "interval": { "start_s": 10.0, "end_s": 12.0, "source_start_s": 20.0, "source_end_s": 22.0 }, "confidence": "medium", "temporal_status": "localized" }, { "caption_index": 5, "loom_clip_id": "v_--veKG73Di4_1753_2832", "action_id": "action_6", "mentions": [ { "text": "turns", "ranges": [ [ 117, 122 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "interval": { "start_s": 4.0, "end_s": 36.0, "source_start_s": 62.5, "source_end_s": 94.5 }, "confidence": "high", "temporal_status": "localized" } ], "id": "action_group_007", "segment_count": 3, "caption_indices": [ 2, 5 ] }, { "action": "face", "types": [ "posture" ], "segments": [ { "caption_index": 2, "loom_clip_id": "v_--veKG73Di4_300_764", "action_id": "A2", "mentions": [ { "text": "face", "ranges": [ [ 35, 39 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "interval": { "start_s": 1.0, "end_s": 11.0, "source_start_s": 11.0, "source_end_s": 21.0 }, "confidence": "high", "temporal_status": "localized" } ], "id": "action_group_008", "segment_count": 1, "caption_indices": [ 2 ] }, { "action": "gesturing", "types": [ "simple" ], "segments": [ { "caption_index": 2, "loom_clip_id": "v_--veKG73Di4_300_764", "action_id": "A3", "mentions": [ { "text": "gesturing", "ranges": [ [ 64, 73 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "interval": { "start_s": 3.0, "end_s": 11.0, "source_start_s": 13.0, "source_end_s": 21.0 }, "confidence": "high", "temporal_status": "localized" } ], "id": "action_group_009", "segment_count": 1, "caption_indices": [ 2 ] }, { "action": "unscrew", "types": [ "simple" ], "segments": [ { "caption_index": 2, "loom_clip_id": "v_--veKG73Di4_300_764", "action_id": "A7", "mentions": [ { "text": "unscrew", "ranges": [ [ 162, 169 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "interval": { "start_s": 11.0, "end_s": 15.0, "source_start_s": 21.0, "source_end_s": 25.0 }, "confidence": "high", "temporal_status": "localized" } ], "id": "action_group_010", "segment_count": 1, "caption_indices": [ 2 ] }, { "action": "uses", "types": [ "simple" ], "segments": [ { "caption_index": 2, "loom_clip_id": "v_--veKG73Di4_300_764", "action_id": "A6", "mentions": [ { "text": "uses", "ranges": [ [ 143, 147 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "interval": { "start_s": 11.0, "end_s": 15.0, "source_start_s": 21.0, "source_end_s": 25.0 }, "confidence": "high", "temporal_status": "localized" } ], "id": "action_group_011", "segment_count": 1, "caption_indices": [ 2 ] }, { "action": "hold", "types": [ "simple" ], "segments": [ { "caption_index": 2, "loom_clip_id": "v_--veKG73Di4_300_764", "action_id": "A8", "mentions": [ { "text": "hold", "ranges": [ [ 212, 216 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "interval": { "start_s": 15.0, "end_s": 15.5, "source_start_s": 25.0, "source_end_s": 25.5 }, "confidence": "high", "temporal_status": "localized" } ], "id": "action_group_012", "segment_count": 1, "caption_indices": [ 2 ] }, { "action": "turns back", "types": [ "phrasal" ], "segments": [ { "caption_index": 3, "loom_clip_id": "v_--veKG73Di4_764_959", "action_id": "action_1", "mentions": [ { "text": "turns back", "ranges": [ [ 77, 87 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "interval": { "start_s": 2.0, "end_s": 3.0, "source_start_s": 27.5, "source_end_s": 28.5 }, "confidence": "medium", "temporal_status": "localized_sparse" } ], "id": "action_group_013", "segment_count": 1, "caption_indices": [ 3 ] }, { "action": "clapping", "types": [ "simple" ], "segments": [ { "caption_index": 3, "loom_clip_id": "v_--veKG73Di4_764_959", "action_id": "action_2", "mentions": [ { "text": "clapping", "ranges": [ [ 99, 107 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } }, { "text": "clapping", "ranges": [ [ 157, 165 ] ], "occurrence": 2, "particle": { "text": null, "occurrence": 1 } } ], "interval": { "start_s": 3.0, "end_s": 6.5, "source_start_s": 28.5, "source_end_s": 32.0 }, "confidence": "medium", "temporal_status": "localized_sparse" } ], "id": "action_group_014", "segment_count": 1, "caption_indices": [ 3 ] }, { "action": "sways", "types": [ "simple" ], "segments": [ { "caption_index": 3, "loom_clip_id": "v_--veKG73Di4_764_959", "action_id": "action_3", "mentions": [ { "text": "sways", "ranges": [ [ 136, 141 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "interval": { "start_s": 3.0, "end_s": 6.5, "source_start_s": 28.5, "source_end_s": 32.0 }, "confidence": "medium", "temporal_status": "localized_sparse" } ], "id": "action_group_015", "segment_count": 1, "caption_indices": [ 3 ] }, { "action": "dancing · dances", "types": [ "simple" ], "segments": [ { "caption_index": 4, "loom_clip_id": "v_--veKG73Di4_959_1753", "action_id": "action_1", "mentions": [ { "text": "dancing", "ranges": [ [ 11, 18 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } }, { "text": "dances", "ranges": [ [ 215, 221 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "interval": { "start_s": 0.0, "end_s": 26.5, "source_start_s": 32.0, "source_end_s": 58.5 }, "confidence": "high", "temporal_status": "localized" } ], "id": "action_group_016", "segment_count": 1, "caption_indices": [ 4 ] }, { "action": "moving", "types": [ "simple" ], "segments": [ { "caption_index": 4, "loom_clip_id": "v_--veKG73Di4_959_1753", "action_id": "action_2", "mentions": [ { "text": "moving", "ranges": [ [ 34, 40 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "interval": { "start_s": 0.0, "end_s": 26.5, "source_start_s": 32.0, "source_end_s": 58.5 }, "confidence": "high", "temporal_status": "localized" } ], "id": "action_group_017", "segment_count": 1, "caption_indices": [ 4 ] }, { "action": "raising", "types": [ "simple" ], "segments": [ { "caption_index": 4, "loom_clip_id": "v_--veKG73Di4_959_1753", "action_id": "action_4", "mentions": [ { "text": "raising", "ranges": [ [ 121, 128 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "interval": { "start_s": 1.0, "end_s": 5.0, "source_start_s": 33.0, "source_end_s": 37.0 }, "confidence": "high", "temporal_status": "localized" }, { "caption_index": 4, "loom_clip_id": "v_--veKG73Di4_959_1753", "action_id": "action_4", "mentions": [ { "text": "raising", "ranges": [ [ 121, 128 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "interval": { "start_s": 8.0, "end_s": 11.0, "source_start_s": 40.0, "source_end_s": 43.0 }, "confidence": "high", "temporal_status": "localized" }, { "caption_index": 4, "loom_clip_id": "v_--veKG73Di4_959_1753", "action_id": "action_4", "mentions": [ { "text": "raising", "ranges": [ [ 121, 128 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "interval": { "start_s": 17.0, "end_s": 22.0, "source_start_s": 49.0, "source_end_s": 54.0 }, "confidence": "high", "temporal_status": "localized" }, { "caption_index": 4, "loom_clip_id": "v_--veKG73Di4_959_1753", "action_id": "action_4", "mentions": [ { "text": "raising", "ranges": [ [ 121, 128 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "interval": { "start_s": 23.0, "end_s": 25.0, "source_start_s": 55.0, "source_end_s": 57.0 }, "confidence": "high", "temporal_status": "localized" } ], "id": "action_group_018", "segment_count": 4, "caption_indices": [ 4 ] }, { "action": "swaying", "types": [ "simple" ], "segments": [ { "caption_index": 4, "loom_clip_id": "v_--veKG73Di4_959_1753", "action_id": "action_6", "mentions": [ { "text": "swaying", "ranges": [ [ 164, 171 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "interval": { "start_s": 5.0, "end_s": 26.5, "source_start_s": 37.0, "source_end_s": 58.5 }, "confidence": "high", "temporal_status": "localized" } ], "id": "action_group_019", "segment_count": 1, "caption_indices": [ 4 ] }, { "action": "turning", "types": [ "simple" ], "segments": [ { "caption_index": 4, "loom_clip_id": "v_--veKG73Di4_959_1753", "action_id": "action_5", "mentions": [ { "text": "turning", "ranges": [ [ 139, 146 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "interval": { "start_s": 9.0, "end_s": 18.0, "source_start_s": 41.0, "source_end_s": 50.0 }, "confidence": "high", "temporal_status": "localized" }, { "caption_index": 6, "loom_clip_id": "v_--veKG73Di4_2832_3291", "action_id": "action_4", "mentions": [ { "text": "turning", "ranges": [ [ 126, 133 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "interval": { "start_s": 13.0, "end_s": 14.0, "source_start_s": 107.5, "source_end_s": 108.5 }, "confidence": "medium", "temporal_status": "supported" } ], "id": "action_group_020", "segment_count": 2, "caption_indices": [ 4, 6 ] }, { "action": "dance · dancing", "types": [ "simple" ], "segments": [ { "caption_index": 5, "loom_clip_id": "v_--veKG73Di4_1753_2832", "action_id": "action_1", "mentions": [ { "text": "dance", "ranges": [ [ 14, 19 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } }, { "text": "dancing", "ranges": [ [ 165, 172 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "interval": { "start_s": 0.0, "end_s": 36.0, "source_start_s": 58.5, "source_end_s": 94.5 }, "confidence": "high", "temporal_status": "localized" } ], "id": "action_group_021", "segment_count": 1, "caption_indices": [ 5 ] }, { "action": "mimes", "types": [ "simple" ], "segments": [ { "caption_index": 5, "loom_clip_id": "v_--veKG73Di4_1753_2832", "action_id": "action_3", "mentions": [ { "text": "mimes", "ranges": [ [ 61, 66 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "interval": { "start_s": 0.0, "end_s": 3.0, "source_start_s": 58.5, "source_end_s": 61.5 }, "confidence": "high", "temporal_status": "localized" } ], "id": "action_group_022", "segment_count": 1, "caption_indices": [ 5 ] }, { "action": "moves", "types": [ "simple" ], "segments": [ { "caption_index": 5, "loom_clip_id": "v_--veKG73Di4_1753_2832", "action_id": "action_2", "mentions": [ { "text": "moves", "ranges": [ [ 25, 30 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "interval": { "start_s": 0.0, "end_s": 2.0, "source_start_s": 58.5, "source_end_s": 60.5 }, "confidence": "high", "temporal_status": "localized" } ], "id": "action_group_023", "segment_count": 1, "caption_indices": [ 5 ] }, { "action": "playing", "types": [ "simple" ], "segments": [ { "caption_index": 5, "loom_clip_id": "v_--veKG73Di4_1753_2832", "action_id": "action_4", "mentions": [ { "text": "playing", "ranges": [ [ 67, 74 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "interval": { "start_s": 0.0, "end_s": 3.0, "source_start_s": 58.5, "source_end_s": 61.5 }, "confidence": "medium", "temporal_status": "localized" } ], "id": "action_group_024", "segment_count": 1, "caption_indices": [ 5 ] }, { "action": "steps", "types": [ "simple" ], "segments": [ { "caption_index": 5, "loom_clip_id": "v_--veKG73Di4_1753_2832", "action_id": "action_5", "mentions": [ { "text": "steps", "ranges": [ [ 105, 110 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "interval": { "start_s": 2.0, "end_s": 5.0, "source_start_s": 60.5, "source_end_s": 63.5 }, "confidence": "high", "temporal_status": "localized" } ], "id": "action_group_025", "segment_count": 1, "caption_indices": [ 5 ] }, { "action": "point", "types": [ "prepositional" ], "segments": [ { "caption_index": 6, "loom_clip_id": "v_--veKG73Di4_2832_3291", "action_id": "action_2", "mentions": [ { "text": "point", "ranges": [ [ 31, 36 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "interval": { "start_s": 2.0, "end_s": 3.0, "source_start_s": 96.5, "source_end_s": 97.5 }, "confidence": "high", "temporal_status": "supported" } ], "id": "action_group_026", "segment_count": 1, "caption_indices": [ 6 ] }, { "action": "stops", "types": [ "simple" ], "segments": [ { "caption_index": 6, "loom_clip_id": "v_--veKG73Di4_2832_3291", "action_id": "action_1", "mentions": [ { "text": "stops", "ranges": [ [ 12, 17 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "interval": { "start_s": 2.0, "end_s": 3.0, "source_start_s": 96.5, "source_end_s": 97.5 }, "confidence": "medium", "temporal_status": "supported" } ], "id": "action_group_027", "segment_count": 1, "caption_indices": [ 6 ] }, { "action": "dancing", "types": [ "simple" ], "segments": [ { "caption_index": 6, "loom_clip_id": "v_--veKG73Di4_2832_3291", "action_id": "action_3", "mentions": [ { "text": "dancing", "ranges": [ [ 99, 106 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "interval": { "start_s": 4.0, "end_s": 14.0, "source_start_s": 98.5, "source_end_s": 108.5 }, "confidence": "high", "temporal_status": "supported" } ], "id": "action_group_028", "segment_count": 1, "caption_indices": [ 6 ] }, { "action": "walking", "types": [ "simple" ], "segments": [ { "caption_index": 6, "loom_clip_id": "v_--veKG73Di4_2832_3291", "action_id": "action_5", "mentions": [ { "text": "walking", "ranges": [ [ 151, 158 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "interval": { "start_s": 14.0, "end_s": 15.310000000000002, "source_start_s": 108.5, "source_end_s": 109.81 }, "confidence": "high", "temporal_status": "supported" } ], "id": "action_group_029", "segment_count": 1, "caption_indices": [ 6 ] }, { "action": "speaking", "types": [ "simple" ], "segments": [ { "caption_index": 2, "loom_clip_id": "v_--veKG73Di4_300_764", "action_id": "A4", "mentions": [ { "text": "speaking", "ranges": [ [ 100, 108 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "interval": null, "confidence": "medium", "temporal_status": "uncertain" } ], "id": "action_group_030", "segment_count": 1, "caption_indices": [ 2 ] } ] }, { "source_video_id": "v_-0NfmccuL9Q", "grouping_rule": "Exact mention phrase set after case/whitespace normalization; action type is retained as metadata but does not split groups.", "actions": [ { "action": "enters", "types": [ "simple" ], "segments": [ { "caption_index": 1, "loom_clip_id": "v_-0NfmccuL9Q_185_545", "action_id": "action_2", "mentions": [ { "text": "enters", "ranges": [ [ 60, 66 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "interval": { "start_s": 1.0, "end_s": 6.0, "source_start_s": 7.17, "source_end_s": 12.17 }, "confidence": "high", "temporal_status": "supported" } ], "id": "action_group_001", "segment_count": 1, "caption_indices": [ 1 ] }, { "action": "wearing", "types": [ "simple" ], "segments": [ { "caption_index": 1, "loom_clip_id": "v_-0NfmccuL9Q_185_545", "action_id": "action_1", "mentions": [ { "text": "wearing", "ranges": [ [ 9, 16 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "interval": { "start_s": 4.0, "end_s": 12.000000000000002, "source_start_s": 10.17, "source_end_s": 18.17 }, "confidence": "high", "temporal_status": "supported" }, { "caption_index": 4, "loom_clip_id": "v_-0NfmccuL9Q_2641_2986", "action_id": "action_wearing_1", "mentions": [ { "text": "wearing", "ranges": [ [ 9, 16 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "interval": { "start_s": 1.0, "end_s": 11.5, "source_start_s": 89.12, "source_end_s": 99.62 }, "confidence": "high", "temporal_status": "localized_sparse" } ], "id": "action_group_002", "segment_count": 2, "caption_indices": [ 1, 4 ] }, { "action": "pick up", "types": [ "phrasal" ], "segments": [ { "caption_index": 1, "loom_clip_id": "v_-0NfmccuL9Q_185_545", "action_id": "action_4", "mentions": [ { "text": "pick up", "ranges": [ [ 128, 135 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "interval": { "start_s": 5.0, "end_s": 8.0, "source_start_s": 11.17, "source_end_s": 14.17 }, "confidence": "high", "temporal_status": "supported" }, { "caption_index": 8, "loom_clip_id": "v_-0NfmccuL9Q_6177_6444", "action_id": "action_2", "mentions": [ { "text": "pick up", "ranges": [ [ 80, 87 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "interval": { "start_s": 8.0, "end_s": 8.889999999999986, "source_start_s": 214.12, "source_end_s": 215.01 }, "confidence": "high", "temporal_status": "supported_at_clip_boundary" } ], "id": "action_group_003", "segment_count": 2, "caption_indices": [ 1, 8 ] }, { "action": "crouches", "types": [ "posture" ], "segments": [ { "caption_index": 1, "loom_clip_id": "v_-0NfmccuL9Q_185_545", "action_id": "action_3", "mentions": [ { "text": "crouches", "ranges": [ [ 86, 94 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "interval": { "start_s": 6.0, "end_s": 12.000000000000002, "source_start_s": 12.17, "source_end_s": 18.17 }, "confidence": "high", "temporal_status": "supported" }, { "caption_index": 7, "loom_clip_id": "v_-0NfmccuL9Q_5368_6177", "action_id": "action_2", "mentions": [ { "text": "crouches", "ranges": [ [ 58, 66 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "interval": { "start_s": 0.0, "end_s": 27.0, "source_start_s": 179.12, "source_end_s": 206.12 }, "confidence": "high", "temporal_status": "localized" } ], "id": "action_group_004", "segment_count": 2, "caption_indices": [ 1, 7 ] }, { "action": "reach", "types": [ "prepositional" ], "segments": [ { "caption_index": 1, "loom_clip_id": "v_-0NfmccuL9Q_185_545", "action_id": "action_5", "mentions": [ { "text": "reach", "ranges": [ [ 176, 181 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "interval": { "start_s": 7.0, "end_s": 9.0, "source_start_s": 13.17, "source_end_s": 15.17 }, "confidence": "medium", "temporal_status": "supported" } ], "id": "action_group_005", "segment_count": 1, "caption_indices": [ 1 ] }, { "action": "position", "types": [ "simple" ], "segments": [ { "caption_index": 1, "loom_clip_id": "v_-0NfmccuL9Q_185_545", "action_id": "action_6", "mentions": [ { "text": "position", "ranges": [ [ 201, 209 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "interval": { "start_s": 8.0, "end_s": 12.000000000000002, "source_start_s": 14.17, "source_end_s": 18.17 }, "confidence": "high", "temporal_status": "supported" } ], "id": "action_group_006", "segment_count": 1, "caption_indices": [ 1 ] }, { "action": "kneeling", "types": [ "posture" ], "segments": [ { "caption_index": 2, "loom_clip_id": "v_-0NfmccuL9Q_545_1099", "action_id": "action_1", "mentions": [ { "text": "kneeling", "ranges": [ [ 6, 14 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "interval": { "start_s": 0.0, "end_s": 18.5, "source_start_s": 18.17, "source_end_s": 36.67 }, "confidence": "high", "temporal_status": "supported" } ], "id": "action_group_007", "segment_count": 1, "caption_indices": [ 2 ] }, { "action": "operate", "types": [ "simple" ], "segments": [ { "caption_index": 2, "loom_clip_id": "v_-0NfmccuL9Q_545_1099", "action_id": "action_3", "mentions": [ { "text": "operate", "ranges": [ [ 50, 57 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "interval": { "start_s": 0.0, "end_s": 18.5, "source_start_s": 18.17, "source_end_s": 36.67 }, "confidence": "high", "temporal_status": "supported" } ], "id": "action_group_008", "segment_count": 1, "caption_indices": [ 2 ] }, { "action": "raising", "types": [ "simple" ], "segments": [ { "caption_index": 2, "loom_clip_id": "v_-0NfmccuL9Q_545_1099", "action_id": "action_4", "mentions": [ { "text": "raising", "ranges": [ [ 81, 88 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "interval": { "start_s": 0.0, "end_s": 18.5, "source_start_s": 18.17, "source_end_s": 36.67 }, "confidence": "high", "temporal_status": "supported" } ], "id": "action_group_009", "segment_count": 1, "caption_indices": [ 2 ] }, { "action": "uses", "types": [ "simple" ], "segments": [ { "caption_index": 2, "loom_clip_id": "v_-0NfmccuL9Q_545_1099", "action_id": "action_2", "mentions": [ { "text": "uses", "ranges": [ [ 27, 31 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "interval": { "start_s": 0.0, "end_s": 18.5, "source_start_s": 18.17, "source_end_s": 36.67 }, "confidence": "high", "temporal_status": "supported" }, { "caption_index": 3, "loom_clip_id": "v_-0NfmccuL9Q_1099_2385", "action_id": "action_2", "mentions": [ { "text": "uses", "ranges": [ [ 51, 55 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1, "ranges": [] } } ], "interval": { "start_s": 7.0, "end_s": 36.0, "source_start_s": 43.67, "source_end_s": 72.67 }, "confidence": "high", "temporal_status": "supported" } ], "id": "action_group_010", "segment_count": 2, "caption_indices": [ 2, 3 ] }, { "action": "attaches", "types": [ "simple" ], "segments": [ { "caption_index": 3, "loom_clip_id": "v_-0NfmccuL9Q_1099_2385", "action_id": "action_1", "mentions": [ { "text": "attaches", "ranges": [ [ 11, 19 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1, "ranges": [] } } ], "interval": { "start_s": 4.0, "end_s": 7.0, "source_start_s": 40.67, "source_end_s": 43.67 }, "confidence": "high", "temporal_status": "supported" }, { "caption_index": 6, "loom_clip_id": "v_-0NfmccuL9Q_4439_5368", "action_id": "action_2", "mentions": [ { "text": "attaches", "ranges": [ [ 57, 65 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "interval": { "start_s": 2.0, "end_s": 5.0, "source_start_s": 150.12, "source_end_s": 153.12 }, "confidence": "medium", "temporal_status": "supported" } ], "id": "action_group_011", "segment_count": 2, "caption_indices": [ 3, 6 ] }, { "action": "remove", "types": [ "simple" ], "segments": [ { "caption_index": 3, "loom_clip_id": "v_-0NfmccuL9Q_1099_2385", "action_id": "action_3", "mentions": [ { "text": "remove", "ranges": [ [ 77, 83 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1, "ranges": [] } } ], "interval": { "start_s": 7.0, "end_s": 36.0, "source_start_s": 43.67, "source_end_s": 72.67 }, "confidence": "high", "temporal_status": "supported" } ], "id": "action_group_012", "segment_count": 1, "caption_indices": [ 3 ] }, { "action": "placing", "types": [ "simple" ], "segments": [ { "caption_index": 3, "loom_clip_id": "v_-0NfmccuL9Q_1099_2385", "action_id": "action_4", "mentions": [ { "text": "placing", "ranges": [ [ 123, 130 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1, "ranges": [] } } ], "interval": { "start_s": 37.0, "end_s": 39.0, "source_start_s": 73.67, "source_end_s": 75.67 }, "confidence": "high", "temporal_status": "supported" }, { "caption_index": 7, "loom_clip_id": "v_-0NfmccuL9Q_5368_6177", "action_id": "action_1", "mentions": [ { "text": "placing", "ranges": [ [ 6, 13 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "interval": null, "confidence": "medium", "temporal_status": "uncertain" } ], "id": "action_group_013", "segment_count": 2, "caption_indices": [ 3, 7 ] }, { "action": "grip", "types": [ "simple" ], "segments": [ { "caption_index": 3, "loom_clip_id": "v_-0NfmccuL9Q_1099_2385", "action_id": "action_5", "mentions": [ { "text": "grip", "ranges": [ [ 173, 177 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1, "ranges": [] } } ], "interval": { "start_s": 39.0, "end_s": 42.0, "source_start_s": 75.67, "source_end_s": 78.67 }, "confidence": "high", "temporal_status": "supported" } ], "id": "action_group_014", "segment_count": 1, "caption_indices": [ 3 ] }, { "action": "pull", "types": [ "phrasal" ], "segments": [ { "caption_index": 3, "loom_clip_id": "v_-0NfmccuL9Q_1099_2385", "action_id": "action_6", "mentions": [ { "text": "pull", "ranges": [ [ 208, 212 ] ], "occurrence": 1, "particle": { "text": "off", "occurrence": 1, "ranges": [ [ 216, 219 ] ] } } ], "interval": { "start_s": 40.0, "end_s": 42.0, "source_start_s": 76.67, "source_end_s": 78.67 }, "confidence": "high", "temporal_status": "supported" } ], "id": "action_group_015", "segment_count": 1, "caption_indices": [ 3 ] }, { "action": "roll", "types": [ "simple" ], "segments": [ { "caption_index": 3, "loom_clip_id": "v_-0NfmccuL9Q_1099_2385", "action_id": "action_7", "mentions": [ { "text": "roll", "ranges": [ [ 239, 243 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1, "ranges": [] } } ], "interval": { "start_s": 42.0, "end_s": 42.91, "source_start_s": 78.67, "source_end_s": 79.58 }, "confidence": "medium", "temporal_status": "sparsely supported at clip boundary" } ], "id": "action_group_016", "segment_count": 1, "caption_indices": [ 3 ] }, { "action": "carries", "types": [ "simple" ], "segments": [ { "caption_index": 4, "loom_clip_id": "v_-0NfmccuL9Q_2641_2986", "action_id": "action_carries_1", "mentions": [ { "text": "carries", "ranges": [ [ 60, 67 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "interval": { "start_s": 1.0, "end_s": 3.0, "source_start_s": 89.12, "source_end_s": 91.12 }, "confidence": "high", "temporal_status": "localized_sparse" } ], "id": "action_group_017", "segment_count": 1, "caption_indices": [ 4 ] }, { "action": "align", "types": [ "simple" ], "segments": [ { "caption_index": 4, "loom_clip_id": "v_-0NfmccuL9Q_2641_2986", "action_id": "action_align_1", "mentions": [ { "text": "align", "ranges": [ [ 115, 120 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "interval": { "start_s": 3.0, "end_s": 11.5, "source_start_s": 91.12, "source_end_s": 99.62 }, "confidence": "high", "temporal_status": "localized_sparse" } ], "id": "action_group_018", "segment_count": 1, "caption_indices": [ 4 ] }, { "action": "lift", "types": [ "simple" ], "segments": [ { "caption_index": 4, "loom_clip_id": "v_-0NfmccuL9Q_2641_2986", "action_id": "action_lift_1", "mentions": [ { "text": "lift", "ranges": [ [ 164, 168 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "interval": { "start_s": 3.0, "end_s": 11.5, "source_start_s": 91.12, "source_end_s": 99.62 }, "confidence": "high", "temporal_status": "localized_sparse" } ], "id": "action_group_019", "segment_count": 1, "caption_indices": [ 4 ] }, { "action": "bend over", "types": [ "phrasal" ], "segments": [ { "caption_index": 4, "loom_clip_id": "v_-0NfmccuL9Q_2641_2986", "action_id": "action_bend_over_1", "mentions": [ { "text": "bend over", "ranges": [ [ 104, 113 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "interval": { "start_s": 4.0, "end_s": 11.5, "source_start_s": 92.12, "source_end_s": 99.62 }, "confidence": "high", "temporal_status": "localized_sparse" } ], "id": "action_group_020", "segment_count": 1, "caption_indices": [ 4 ] }, { "action": "screw · securing", "types": [ "simple" ], "segments": [ { "caption_index": 5, "loom_clip_id": "v_-0NfmccuL9Q_2986_4439", "action_id": "A3", "mentions": [ { "text": "screw", "ranges": [ [ 103, 108 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } }, { "text": "securing", "ranges": [ [ 199, 207 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "interval": { "start_s": 0.0, "end_s": 2.0, "source_start_s": 99.62, "source_end_s": 101.62 }, "confidence": "high", "temporal_status": "localized" }, { "caption_index": 5, "loom_clip_id": "v_-0NfmccuL9Q_2986_4439", "action_id": "A3", "mentions": [ { "text": "screw", "ranges": [ [ 103, 108 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } }, { "text": "securing", "ranges": [ [ 199, 207 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "interval": { "start_s": 4.0, "end_s": 15.0, "source_start_s": 103.62, "source_end_s": 114.62 }, "confidence": "high", "temporal_status": "localized" }, { "caption_index": 5, "loom_clip_id": "v_-0NfmccuL9Q_2986_4439", "action_id": "A3", "mentions": [ { "text": "screw", "ranges": [ [ 103, 108 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } }, { "text": "securing", "ranges": [ [ 199, 207 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "interval": { "start_s": 16.0, "end_s": 22.0, "source_start_s": 115.62, "source_end_s": 121.62 }, "confidence": "high", "temporal_status": "localized" }, { "caption_index": 5, "loom_clip_id": "v_-0NfmccuL9Q_2986_4439", "action_id": "A3", "mentions": [ { "text": "screw", "ranges": [ [ 103, 108 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } }, { "text": "securing", "ranges": [ [ 199, 207 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "interval": { "start_s": 24.0, "end_s": 30.0, "source_start_s": 123.62, "source_end_s": 129.62 }, "confidence": "high", "temporal_status": "localized" }, { "caption_index": 5, "loom_clip_id": "v_-0NfmccuL9Q_2986_4439", "action_id": "A3", "mentions": [ { "text": "screw", "ranges": [ [ 103, 108 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } }, { "text": "securing", "ranges": [ [ 199, 207 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "interval": { "start_s": 31.0, "end_s": 36.0, "source_start_s": 130.62, "source_end_s": 135.62 }, "confidence": "high", "temporal_status": "localized" }, { "caption_index": 5, "loom_clip_id": "v_-0NfmccuL9Q_2986_4439", "action_id": "A3", "mentions": [ { "text": "screw", "ranges": [ [ 103, 108 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } }, { "text": "securing", "ranges": [ [ 199, 207 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "interval": { "start_s": 37.0, "end_s": 45.0, "source_start_s": 136.62, "source_end_s": 144.62 }, "confidence": "high", "temporal_status": "localized" } ], "id": "action_group_021", "segment_count": 6, "caption_indices": [ 5 ] }, { "action": "Squatting", "types": [ "posture" ], "segments": [ { "caption_index": 5, "loom_clip_id": "v_-0NfmccuL9Q_2986_4439", "action_id": "A1", "mentions": [ { "text": "Squatting", "ranges": [ [ 0, 9 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "interval": { "start_s": 0.0, "end_s": 48.5, "source_start_s": 99.62, "source_end_s": 148.12 }, "confidence": "high", "temporal_status": "localized" } ], "id": "action_group_022", "segment_count": 1, "caption_indices": [ 5 ] }, { "action": "picks up", "types": [ "phrasal" ], "segments": [ { "caption_index": 5, "loom_clip_id": "v_-0NfmccuL9Q_2986_4439", "action_id": "A2", "mentions": [ { "text": "picks up", "ranges": [ [ 52, 60 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "interval": { "start_s": 2.0, "end_s": 5.0, "source_start_s": 101.62, "source_end_s": 104.62 }, "confidence": "high", "temporal_status": "localized" }, { "caption_index": 6, "loom_clip_id": "v_-0NfmccuL9Q_4439_5368", "action_id": "action_1", "mentions": [ { "text": "picks up", "ranges": [ [ 11, 19 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "interval": { "start_s": 2.0, "end_s": 3.0, "source_start_s": 150.12, "source_end_s": 151.12 }, "confidence": "medium", "temporal_status": "supported" } ], "id": "action_group_023", "segment_count": 2, "caption_indices": [ 5, 6 ] }, { "action": "using", "types": [ "simple" ], "segments": [ { "caption_index": 5, "loom_clip_id": "v_-0NfmccuL9Q_2986_4439", "action_id": "A4", "mentions": [ { "text": "using", "ranges": [ [ 155, 160 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "interval": { "start_s": 16.0, "end_s": 23.0, "source_start_s": 115.62, "source_end_s": 122.62 }, "confidence": "medium", "temporal_status": "localized" } ], "id": "action_group_024", "segment_count": 1, "caption_indices": [ 5 ] }, { "action": "wipe", "types": [ "simple" ], "segments": [ { "caption_index": 5, "loom_clip_id": "v_-0NfmccuL9Q_2986_4439", "action_id": "A5", "mentions": [ { "text": "wipe", "ranges": [ [ 178, 182 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "interval": { "start_s": 16.0, "end_s": 23.0, "source_start_s": 115.62, "source_end_s": 122.62 }, "confidence": "medium", "temporal_status": "localized" } ], "id": "action_group_025", "segment_count": 1, "caption_indices": [ 5 ] }, { "action": "removes", "types": [ "simple" ], "segments": [ { "caption_index": 5, "loom_clip_id": "v_-0NfmccuL9Q_2986_4439", "action_id": "A6", "mentions": [ { "text": "removes", "ranges": [ [ 233, 240 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "interval": { "start_s": 46.0, "end_s": 48.5, "source_start_s": 145.62, "source_end_s": 148.12 }, "confidence": "high", "temporal_status": "localized" } ], "id": "action_group_026", "segment_count": 1, "caption_indices": [ 5 ] }, { "action": "Crouching", "types": [ "posture" ], "segments": [ { "caption_index": 6, "loom_clip_id": "v_-0NfmccuL9Q_4439_5368", "action_id": "action_3", "mentions": [ { "text": "Crouching", "ranges": [ [ 82, 91 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "interval": { "start_s": 0.0, "end_s": 31.0, "source_start_s": 148.12, "source_end_s": 179.12 }, "confidence": "high", "temporal_status": "supported" } ], "id": "action_group_027", "segment_count": 1, "caption_indices": [ 6 ] }, { "action": "tighten", "types": [ "simple" ], "segments": [ { "caption_index": 6, "loom_clip_id": "v_-0NfmccuL9Q_4439_5368", "action_id": "action_5", "mentions": [ { "text": "tighten", "ranges": [ [ 140, 147 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "interval": { "start_s": 5.0, "end_s": 31.0, "source_start_s": 153.12, "source_end_s": 179.12 }, "confidence": "high", "temporal_status": "supported" } ], "id": "action_group_028", "segment_count": 1, "caption_indices": [ 6 ] }, { "action": "use", "types": [ "simple" ], "segments": [ { "caption_index": 6, "loom_clip_id": "v_-0NfmccuL9Q_4439_5368", "action_id": "action_4", "mentions": [ { "text": "use", "ranges": [ [ 103, 106 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "interval": { "start_s": 5.0, "end_s": 31.0, "source_start_s": 153.12, "source_end_s": 179.12 }, "confidence": "high", "temporal_status": "supported" } ], "id": "action_group_029", "segment_count": 1, "caption_indices": [ 6 ] }, { "action": "lower", "types": [ "simple" ], "segments": [ { "caption_index": 7, "loom_clip_id": "v_-0NfmccuL9Q_5368_6177", "action_id": "action_4", "mentions": [ { "text": "lower", "ranges": [ [ 132, 137 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "interval": { "start_s": 3.0, "end_s": 27.0, "source_start_s": 182.12, "source_end_s": 206.12 }, "confidence": "medium", "temporal_status": "localized" } ], "id": "action_group_030", "segment_count": 1, "caption_indices": [ 7 ] }, { "action": "operate · use", "types": [ "simple" ], "segments": [ { "caption_index": 7, "loom_clip_id": "v_-0NfmccuL9Q_5368_6177", "action_id": "action_3", "mentions": [ { "text": "operate", "ranges": [ [ 70, 77 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } }, { "text": "use", "ranges": [ [ 97, 100 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "interval": { "start_s": 3.0, "end_s": 27.0, "source_start_s": 182.12, "source_end_s": 206.12 }, "confidence": "high", "temporal_status": "localized" } ], "id": "action_group_031", "segment_count": 1, "caption_indices": [ 7 ] }, { "action": "pulls", "types": [ "phrasal" ], "segments": [ { "caption_index": 8, "loom_clip_id": "v_-0NfmccuL9Q_6177_6444", "action_id": "action_1", "mentions": [ { "text": "pulls", "ranges": [ [ 11, 16 ] ], "occurrence": 1, "particle": { "text": "out", "occurrence": 1, "range": [ 41, 44 ] } } ], "interval": { "start_s": 0.0, "end_s": 2.0, "source_start_s": 206.12, "source_end_s": 208.12 }, "confidence": "high", "temporal_status": "supported" }, { "caption_index": 8, "loom_clip_id": "v_-0NfmccuL9Q_6177_6444", "action_id": "action_1", "mentions": [ { "text": "pulls", "ranges": [ [ 11, 16 ] ], "occurrence": 1, "particle": { "text": "out", "occurrence": 1, "range": [ 41, 44 ] } } ], "interval": { "start_s": 5.0, "end_s": 7.0, "source_start_s": 211.12, "source_end_s": 213.12 }, "confidence": "high", "temporal_status": "supported" } ], "id": "action_group_032", "segment_count": 2, "caption_indices": [ 8 ] }, { "action": "stand up", "types": [ "phrasal" ], "segments": [ { "caption_index": 8, "loom_clip_id": "v_-0NfmccuL9Q_6177_6444", "action_id": "action_3", "mentions": [ { "text": "stand up", "ranges": [ [ 123, 131 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "interval": null, "confidence": "high", "temporal_status": "uncertain" } ], "id": "action_group_033", "segment_count": 1, "caption_indices": [ 8 ] }, { "action": "walk", "types": [ "simple" ], "segments": [ { "caption_index": 8, "loom_clip_id": "v_-0NfmccuL9Q_6177_6444", "action_id": "action_4", "mentions": [ { "text": "walk", "ranges": [ [ 137, 141 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "interval": null, "confidence": "high", "temporal_status": "uncertain" } ], "id": "action_group_034", "segment_count": 1, "caption_indices": [ 8 ] } ] }, { "source_video_id": "v_-3Ss9jSRbKk", "grouping_rule": "Exact mention phrase set after case/whitespace normalization; action type is retained as metadata but does not split groups.", "actions": [ { "action": "holds", "types": [ "simple", "posture" ], "segments": [ { "caption_index": 2, "loom_clip_id": "v_-3Ss9jSRbKk_283_807", "action_id": "A4", "mentions": [ { "text": "holds", "ranges": [ [ 201, 206 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "interval": { "start_s": 0.0, "end_s": 17.5, "source_start_s": 9.44, "source_end_s": 26.94 }, "confidence": "high", "temporal_status": "supported" }, { "caption_index": 3, "loom_clip_id": "v_-3Ss9jSRbKk_807_1107", "action_id": "action_3", "mentions": [ { "text": "holds", "ranges": [ [ 80, 85 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "interval": { "start_s": 1.0, "end_s": 7.0, "source_start_s": 27.94, "source_end_s": 33.94 }, "confidence": "high", "temporal_status": "supported" } ], "id": "action_group_001", "segment_count": 2, "caption_indices": [ 2, 3 ] }, { "action": "is lying", "types": [ "posture" ], "segments": [ { "caption_index": 2, "loom_clip_id": "v_-3Ss9jSRbKk_283_807", "action_id": "A1", "mentions": [ { "text": "is lying", "ranges": [ [ 41, 49 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "interval": { "start_s": 0.0, "end_s": 17.5, "source_start_s": 9.44, "source_end_s": 26.94 }, "confidence": "high", "temporal_status": "supported" } ], "id": "action_group_002", "segment_count": 1, "caption_indices": [ 2 ] }, { "action": "lifting", "types": [ "simple" ], "segments": [ { "caption_index": 3, "loom_clip_id": "v_-3Ss9jSRbKk_807_1107", "action_id": "action_2", "mentions": [ { "text": "lifting", "ranges": [ [ 32, 39 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "interval": { "start_s": 0.0, "end_s": 2.0, "source_start_s": 26.94, "source_end_s": 28.94 }, "confidence": "high", "temporal_status": "supported" }, { "caption_index": 4, "loom_clip_id": "v_-3Ss9jSRbKk_1107_1602", "action_id": "action_2", "mentions": [ { "text": "lifting", "ranges": [ [ 37, 44 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "interval": { "start_s": 2.0, "end_s": 4.0, "source_start_s": 38.94, "source_end_s": 40.94 }, "confidence": "high", "temporal_status": "localized_from_sparse_samples" }, { "caption_index": 5, "loom_clip_id": "v_-3Ss9jSRbKk_1602_2025", "action_id": "action_3", "mentions": [ { "text": "lifting", "ranges": [ [ 101, 108 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "interval": { "start_s": 3.0, "end_s": 7.0, "source_start_s": 56.44, "source_end_s": 60.44 }, "confidence": "high", "temporal_status": "localized" } ], "id": "action_group_003", "segment_count": 3, "caption_indices": [ 3, 4, 5 ] }, { "action": "performs a sit-up", "types": [ "light_verb" ], "segments": [ { "caption_index": 3, "loom_clip_id": "v_-3Ss9jSRbKk_807_1107", "action_id": "action_1", "mentions": [ { "text": "performs a sit-up", "ranges": [ [ 11, 28 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "interval": { "start_s": 0.0, "end_s": 8.0, "source_start_s": 26.94, "source_end_s": 34.94 }, "confidence": "high", "temporal_status": "supported" } ], "id": "action_group_004", "segment_count": 1, "caption_indices": [ 3 ] }, { "action": "lowers", "types": [ "simple" ], "segments": [ { "caption_index": 3, "loom_clip_id": "v_-3Ss9jSRbKk_807_1107", "action_id": "action_4", "mentions": [ { "text": "lowers", "ranges": [ [ 142, 148 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "interval": { "start_s": 6.0, "end_s": 8.0, "source_start_s": 32.94, "source_end_s": 34.94 }, "confidence": "high", "temporal_status": "supported" }, { "caption_index": 4, "loom_clip_id": "v_-3Ss9jSRbKk_1107_1602", "action_id": "action_4", "mentions": [ { "text": "lowers", "ranges": [ [ 117, 123 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "interval": { "start_s": 4.0, "end_s": 7.0, "source_start_s": 40.94, "source_end_s": 43.94 }, "confidence": "high", "temporal_status": "localized_from_sparse_samples" } ], "id": "action_group_005", "segment_count": 2, "caption_indices": [ 3, 4 ] }, { "action": "performs a second sit-up", "types": [ "light_verb" ], "segments": [ { "caption_index": 4, "loom_clip_id": "v_-3Ss9jSRbKk_1107_1602", "action_id": "action_1", "mentions": [ { "text": "performs a second sit-up", "ranges": [ [ 11, 35 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "interval": { "start_s": 2.0, "end_s": 7.0, "source_start_s": 38.94, "source_end_s": 43.94 }, "confidence": "high", "temporal_status": "localized_from_sparse_samples" } ], "id": "action_group_006", "segment_count": 1, "caption_indices": [ 4 ] }, { "action": "holding", "types": [ "posture" ], "segments": [ { "caption_index": 4, "loom_clip_id": "v_-3Ss9jSRbKk_1107_1602", "action_id": "action_3", "mentions": [ { "text": "holding", "ranges": [ [ 86, 93 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "interval": { "start_s": 3.0, "end_s": 5.0, "source_start_s": 39.94, "source_end_s": 41.94 }, "confidence": "high", "temporal_status": "localized_from_sparse_samples" } ], "id": "action_group_007", "segment_count": 1, "caption_indices": [ 4 ] }, { "action": "remains", "types": [ "posture" ], "segments": [ { "caption_index": 4, "loom_clip_id": "v_-3Ss9jSRbKk_1107_1602", "action_id": "action_5", "mentions": [ { "text": "remains", "ranges": [ [ 160, 167 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "interval": { "start_s": 6.0, "end_s": 14.0, "source_start_s": 42.94, "source_end_s": 50.94 }, "confidence": "high", "temporal_status": "localized_from_sparse_samples" } ], "id": "action_group_008", "segment_count": 1, "caption_indices": [ 4 ] }, { "action": "changes · crossing", "types": [ "simple" ], "segments": [ { "caption_index": 5, "loom_clip_id": "v_-3Ss9jSRbKk_1602_2025", "action_id": "action_1", "mentions": [ { "text": "changes", "ranges": [ [ 11, 18 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } }, { "text": "crossing", "ranges": [ [ 37, 45 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "interval": { "start_s": 2.0, "end_s": 4.0, "source_start_s": 55.44, "source_end_s": 57.44 }, "confidence": "high", "temporal_status": "localized" } ], "id": "action_group_009", "segment_count": 1, "caption_indices": [ 5 ] }, { "action": "performs another sit-up", "types": [ "light_verb" ], "segments": [ { "caption_index": 5, "loom_clip_id": "v_-3Ss9jSRbKk_1602_2025", "action_id": "action_2", "mentions": [ { "text": "performs another sit-up", "ranges": [ [ 76, 99 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "interval": { "start_s": 3.0, "end_s": 13.0, "source_start_s": 56.44, "source_end_s": 66.44 }, "confidence": "high", "temporal_status": "localized" } ], "id": "action_group_010", "segment_count": 1, "caption_indices": [ 5 ] }, { "action": "lowering", "types": [ "simple" ], "segments": [ { "caption_index": 5, "loom_clip_id": "v_-3Ss9jSRbKk_1602_2025", "action_id": "action_4", "mentions": [ { "text": "lowering", "ranges": [ [ 129, 137 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "interval": { "start_s": 7.0, "end_s": 13.0, "source_start_s": 60.44, "source_end_s": 66.44 }, "confidence": "high", "temporal_status": "localized" } ], "id": "action_group_011", "segment_count": 1, "caption_indices": [ 5 ] }, { "action": "uncrosses", "types": [ "simple" ], "segments": [ { "caption_index": 5, "loom_clip_id": "v_-3Ss9jSRbKk_1602_2025", "action_id": "action_5", "mentions": [ { "text": "uncrosses", "ranges": [ [ 165, 174 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "interval": { "start_s": 11.0, "end_s": 13.0, "source_start_s": 64.44, "source_end_s": 66.44 }, "confidence": "high", "temporal_status": "localized" } ], "id": "action_group_012", "segment_count": 1, "caption_indices": [ 5 ] }, { "action": "rests", "types": [ "posture" ], "segments": [ { "caption_index": 5, "loom_clip_id": "v_-3Ss9jSRbKk_1602_2025", "action_id": "action_6", "mentions": [ { "text": "rests", "ranges": [ [ 188, 193 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "interval": { "start_s": 12.0, "end_s": 14.129999999999995, "source_start_s": 65.44, "source_end_s": 67.57 }, "confidence": "high", "temporal_status": "localized" } ], "id": "action_group_013", "segment_count": 1, "caption_indices": [ 5 ] }, { "action": "clasp", "types": [ "simple" ], "segments": [ { "caption_index": 2, "loom_clip_id": "v_-3Ss9jSRbKk_283_807", "action_id": "A3", "mentions": [ { "text": "clasp", "ranges": [ [ 165, 170 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "interval": null, "confidence": "high", "temporal_status": "not_observed_in_sampled_frames" } ], "id": "action_group_014", "segment_count": 1, "caption_indices": [ 2 ] }, { "action": "moves", "types": [ "simple" ], "segments": [ { "caption_index": 2, "loom_clip_id": "v_-3Ss9jSRbKk_283_807", "action_id": "A2", "mentions": [ { "text": "moves", "ranges": [ [ 132, 137 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "interval": null, "confidence": "high", "temporal_status": "not_observed_in_sampled_frames" } ], "id": "action_group_015", "segment_count": 1, "caption_indices": [ 2 ] } ] }, { "source_video_id": "v_-4Q_zG9EChY", "grouping_rule": "Exact mention phrase set after case/whitespace normalization; action type is retained as metadata but does not split groups.", "actions": [ { "action": "holding", "types": [ "simple" ], "segments": [ { "caption_index": 1, "loom_clip_id": "v_-4Q_zG9EChY_180_1484", "action_id": "action_3", "mentions": [ { "text": "holding", "ranges": [ [ 68, 75 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "interval": { "start_s": 1.0, "end_s": 6.0, "source_start_s": 7.01, "source_end_s": 12.01 }, "confidence": "medium", "temporal_status": "localized" } ], "id": "action_group_001", "segment_count": 1, "caption_indices": [ 1 ] }, { "action": "wearing", "types": [ "simple" ], "segments": [ { "caption_index": 1, "loom_clip_id": "v_-4Q_zG9EChY_180_1484", "action_id": "action_1", "mentions": [ { "text": "wearing", "ranges": [ [ 19, 26 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "interval": { "start_s": 1.0, "end_s": 6.0, "source_start_s": 7.01, "source_end_s": 12.01 }, "confidence": "high", "temporal_status": "localized" }, { "caption_index": 1, "loom_clip_id": "v_-4Q_zG9EChY_180_1484", "action_id": "action_1", "mentions": [ { "text": "wearing", "ranges": [ [ 19, 26 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "interval": { "start_s": 7.0, "end_s": 43.5, "source_start_s": 13.01, "source_end_s": 49.51 }, "confidence": "high", "temporal_status": "localized" } ], "id": "action_group_002", "segment_count": 2, "caption_indices": [ 1 ] }, { "action": "balances", "types": [ "simple" ], "segments": [ { "caption_index": 1, "loom_clip_id": "v_-4Q_zG9EChY_180_1484", "action_id": "action_2", "mentions": [ { "text": "balances", "ranges": [ [ 38, 46 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "interval": { "start_s": 7.0, "end_s": 43.5, "source_start_s": 13.01, "source_end_s": 49.51 }, "confidence": "high", "temporal_status": "localized" } ], "id": "action_group_003", "segment_count": 1, "caption_indices": [ 1 ] }, { "action": "stabilize", "types": [ "simple" ], "segments": [ { "caption_index": 1, "loom_clip_id": "v_-4Q_zG9EChY_180_1484", "action_id": "action_6", "mentions": [ { "text": "stabilize", "ranges": [ [ 166, 175 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "interval": { "start_s": 7.0, "end_s": 43.5, "source_start_s": 13.01, "source_end_s": 49.51 }, "confidence": "high", "temporal_status": "localized" } ], "id": "action_group_004", "segment_count": 1, "caption_indices": [ 1 ] }, { "action": "using", "types": [ "simple" ], "segments": [ { "caption_index": 1, "loom_clip_id": "v_-4Q_zG9EChY_180_1484", "action_id": "action_5", "mentions": [ { "text": "using", "ranges": [ [ 146, 151 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "interval": { "start_s": 7.0, "end_s": 43.5, "source_start_s": 13.01, "source_end_s": 49.51 }, "confidence": "high", "temporal_status": "localized" }, { "caption_index": 3, "loom_clip_id": "v_-4Q_zG9EChY_3012_3432", "action_id": "A5", "mentions": [ { "text": "using", "ranges": [ [ 164, 169 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "interval": { "start_s": 0.0, "end_s": 14.0, "source_start_s": 100.51, "source_end_s": 114.51 }, "confidence": "high", "temporal_status": "localized" } ], "id": "action_group_005", "segment_count": 2, "caption_indices": [ 1, 3 ] }, { "action": "walking · take careful steps", "types": [ "simple" ], "segments": [ { "caption_index": 1, "loom_clip_id": "v_-4Q_zG9EChY_180_1484", "action_id": "action_4", "mentions": [ { "text": "walking", "ranges": [ [ 122, 129 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } }, { "text": "take careful steps", "ranges": [ [ 199, 217 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "interval": { "start_s": 7.0, "end_s": 43.5, "source_start_s": 13.01, "source_end_s": 49.51 }, "confidence": "high", "temporal_status": "localized" } ], "id": "action_group_006", "segment_count": 1, "caption_indices": [ 1 ] }, { "action": "extend", "types": [ "simple" ], "segments": [ { "caption_index": 2, "loom_clip_id": "v_-4Q_zG9EChY_1484_3012", "action_id": "action_5", "mentions": [ { "text": "extend", "ranges": [ [ 130, 136 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "interval": { "start_s": 0.0, "end_s": 2.0, "source_start_s": 49.51, "source_end_s": 51.51 }, "confidence": "high", "temporal_status": "localized" }, { "caption_index": 2, "loom_clip_id": "v_-4Q_zG9EChY_1484_3012", "action_id": "action_5", "mentions": [ { "text": "extend", "ranges": [ [ 130, 136 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "interval": { "start_s": 3.0, "end_s": 7.0, "source_start_s": 52.51, "source_end_s": 56.51 }, "confidence": "high", "temporal_status": "localized" }, { "caption_index": 2, "loom_clip_id": "v_-4Q_zG9EChY_1484_3012", "action_id": "action_5", "mentions": [ { "text": "extend", "ranges": [ [ 130, 136 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "interval": { "start_s": 10.0, "end_s": 18.0, "source_start_s": 59.51, "source_end_s": 67.50999999999999 }, "confidence": "high", "temporal_status": "localized" }, { "caption_index": 2, "loom_clip_id": "v_-4Q_zG9EChY_1484_3012", "action_id": "action_5", "mentions": [ { "text": "extend", "ranges": [ [ 130, 136 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "interval": { "start_s": 20.0, "end_s": 24.0, "source_start_s": 69.50999999999999, "source_end_s": 73.50999999999999 }, "confidence": "high", "temporal_status": "localized" }, { "caption_index": 2, "loom_clip_id": "v_-4Q_zG9EChY_1484_3012", "action_id": "action_5", "mentions": [ { "text": "extend", "ranges": [ [ 130, 136 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "interval": { "start_s": 26.0, "end_s": 31.0, "source_start_s": 75.50999999999999, "source_end_s": 80.50999999999999 }, "confidence": "high", "temporal_status": "localized" }, { "caption_index": 2, "loom_clip_id": "v_-4Q_zG9EChY_1484_3012", "action_id": "action_5", "mentions": [ { "text": "extend", "ranges": [ [ 130, 136 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "interval": { "start_s": 32.0, "end_s": 36.0, "source_start_s": 81.50999999999999, "source_end_s": 85.50999999999999 }, "confidence": "high", "temporal_status": "localized" }, { "caption_index": 2, "loom_clip_id": "v_-4Q_zG9EChY_1484_3012", "action_id": "action_5", "mentions": [ { "text": "extend", "ranges": [ [ 130, 136 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "interval": { "start_s": 40.0, "end_s": 46.0, "source_start_s": 89.50999999999999, "source_end_s": 95.50999999999999 }, "confidence": "high", "temporal_status": "localized" }, { "caption_index": 2, "loom_clip_id": "v_-4Q_zG9EChY_1484_3012", "action_id": "action_5", "mentions": [ { "text": "extend", "ranges": [ [ 130, 136 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "interval": { "start_s": 48.0, "end_s": 51.00000000000001, "source_start_s": 97.50999999999999, "source_end_s": 100.51 }, "confidence": "high", "temporal_status": "localized" } ], "id": "action_group_007", "segment_count": 8, "caption_indices": [ 2 ] }, { "action": "walking · move", "types": [ "simple" ], "segments": [ { "caption_index": 2, "loom_clip_id": "v_-4Q_zG9EChY_1484_3012", "action_id": "action_4", "mentions": [ { "text": "walking", "ranges": [ [ 108, 115 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } }, { "text": "move", "ranges": [ [ 173, 177 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "interval": { "start_s": 0.0, "end_s": 51.00000000000001, "source_start_s": 49.51, "source_end_s": 100.51 }, "confidence": "high", "temporal_status": "localized" } ], "id": "action_group_008", "segment_count": 1, "caption_indices": [ 2 ] }, { "action": "squats", "types": [ "posture" ], "segments": [ { "caption_index": 2, "loom_clip_id": "v_-4Q_zG9EChY_1484_3012", "action_id": "action_2", "mentions": [ { "text": "squats", "ranges": [ [ 63, 69 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "interval": { "start_s": 1.0, "end_s": 2.0, "source_start_s": 50.51, "source_end_s": 51.51 }, "confidence": "high", "temporal_status": "localized" } ], "id": "action_group_009", "segment_count": 1, "caption_indices": [ 2 ] }, { "action": "pivot", "types": [ "simple" ], "segments": [ { "caption_index": 3, "loom_clip_id": "v_-4Q_zG9EChY_3012_3432", "action_id": "A4", "mentions": [ { "text": "pivot", "ranges": [ [ 126, 131 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "interval": { "start_s": 0.0, "end_s": 14.0, "source_start_s": 100.51, "source_end_s": 114.51 }, "confidence": "high", "temporal_status": "localized" } ], "id": "action_group_010", "segment_count": 1, "caption_indices": [ 3 ] }, { "action": "facing", "types": [ "posture" ], "segments": [ { "caption_index": 4, "loom_clip_id": "v_-4Q_zG9EChY_3432_4108", "action_id": "action_1", "mentions": [ { "text": "facing", "ranges": [ [ 4, 10 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "interval": { "start_s": 0.0, "end_s": 21.0, "source_start_s": 114.51, "source_end_s": 135.51 }, "confidence": "medium", "temporal_status": "supported_sparse" } ], "id": "action_group_011", "segment_count": 1, "caption_indices": [ 4 ] }, { "action": "maintain their balance", "types": [ "other" ], "segments": [ { "caption_index": 4, "loom_clip_id": "v_-4Q_zG9EChY_3432_4108", "action_id": "action_3", "mentions": [ { "text": "maintain their balance", "ranges": [ [ 109, 131 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "interval": { "start_s": 0.0, "end_s": 21.0, "source_start_s": 114.51, "source_end_s": 135.51 }, "confidence": "high", "temporal_status": "supported_sparse" } ], "id": "action_group_012", "segment_count": 1, "caption_indices": [ 4 ] }, { "action": "walks", "types": [ "simple" ], "segments": [ { "caption_index": 4, "loom_clip_id": "v_-4Q_zG9EChY_3432_4108", "action_id": "action_2", "mentions": [ { "text": "walks", "ranges": [ [ 46, 51 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "interval": { "start_s": 0.0, "end_s": 20.0, "source_start_s": 114.51, "source_end_s": 134.51 }, "confidence": "high", "temporal_status": "supported_sparse" } ], "id": "action_group_013", "segment_count": 1, "caption_indices": [ 4 ] }, { "action": "reaching", "types": [ "simple" ], "segments": [ { "caption_index": 3, "loom_clip_id": "v_-4Q_zG9EChY_3012_3432", "action_id": "A1", "mentions": [ { "text": "reaching", "ranges": [ [ 5, 13 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "interval": null, "confidence": "high", "temporal_status": "uncertain" }, { "caption_index": 4, "loom_clip_id": "v_-4Q_zG9EChY_3432_4108", "action_id": "action_4", "mentions": [ { "text": "reaching", "ranges": [ [ 168, 176 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "interval": { "start_s": 19.0, "end_s": 21.0, "source_start_s": 133.51, "source_end_s": 135.51 }, "confidence": "high", "temporal_status": "supported_sparse" } ], "id": "action_group_014", "segment_count": 2, "caption_indices": [ 3, 4 ] }, { "action": "jump", "types": [ "simple" ], "segments": [ { "caption_index": 4, "loom_clip_id": "v_-4Q_zG9EChY_3432_4108", "action_id": "action_5", "mentions": [ { "text": "jump", "ranges": [ [ 191, 195 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "interval": { "start_s": 20.0, "end_s": 22.0, "source_start_s": 134.51, "source_end_s": 136.51 }, "confidence": "medium", "temporal_status": "inferred_from_sparse_transition" } ], "id": "action_group_015", "segment_count": 1, "caption_indices": [ 4 ] }, { "action": "face", "types": [ "posture" ], "segments": [ { "caption_index": 3, "loom_clip_id": "v_-4Q_zG9EChY_3012_3432", "action_id": "A6", "mentions": [ { "text": "face", "ranges": [ [ 203, 207 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "interval": null, "confidence": "high", "temporal_status": "uncertain" } ], "id": "action_group_016", "segment_count": 1, "caption_indices": [ 3 ] }, { "action": "plants", "types": [ "simple" ], "segments": [ { "caption_index": 3, "loom_clip_id": "v_-4Q_zG9EChY_3012_3432", "action_id": "A3", "mentions": [ { "text": "plants", "ranges": [ [ 92, 98 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "interval": null, "confidence": "high", "temporal_status": "uncertain" } ], "id": "action_group_017", "segment_count": 1, "caption_indices": [ 3 ] }, { "action": "regain balance", "types": [ "other" ], "segments": [ { "caption_index": 2, "loom_clip_id": "v_-4Q_zG9EChY_1484_3012", "action_id": "action_3", "mentions": [ { "text": "regain balance", "ranges": [ [ 73, 87 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "interval": null, "confidence": "high", "temporal_status": "uncertain" } ], "id": "action_group_018", "segment_count": 1, "caption_indices": [ 2 ] }, { "action": "tosses", "types": [ "simple" ], "segments": [ { "caption_index": 2, "loom_clip_id": "v_-4Q_zG9EChY_1484_3012", "action_id": "action_1", "mentions": [ { "text": "tosses", "ranges": [ [ 11, 17 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "interval": null, "confidence": "high", "temporal_status": "uncertain" } ], "id": "action_group_019", "segment_count": 1, "caption_indices": [ 2 ] }, { "action": "walking · came", "types": [ "simple" ], "segments": [ { "caption_index": 3, "loom_clip_id": "v_-4Q_zG9EChY_3012_3432", "action_id": "A2", "mentions": [ { "text": "walking", "ranges": [ [ 80, 87 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } }, { "text": "came", "ranges": [ [ 227, 231 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "interval": null, "confidence": "high", "temporal_status": "uncertain" } ], "id": "action_group_020", "segment_count": 1, "caption_indices": [ 3 ] } ] }, { "source_video_id": "v_-DaX_1bBNAo", "grouping_rule": "Exact mention phrase set after case/whitespace normalization; action type is retained as metadata but does not split groups.", "actions": [ { "action": "looking", "types": [ "prepositional", "simple" ], "segments": [ { "caption_index": 1, "loom_clip_id": "v_-DaX_1bBNAo_1_345", "action_id": "action_2", "mentions": [ { "text": "looking", "ranges": [ [ 105, 112 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "interval": { "start_s": 0.0, "end_s": 2.0, "source_start_s": 0.0, "source_end_s": 2.0 }, "confidence": "high", "temporal_status": "supported" }, { "caption_index": 2, "loom_clip_id": "v_-DaX_1bBNAo_345_590", "action_id": "action_2", "mentions": [ { "text": "looking", "ranges": [ [ 24, 31 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "interval": { "start_s": 0.0, "end_s": 2.0, "source_start_s": 11.5, "source_end_s": 13.5 }, "confidence": "medium", "temporal_status": "localized" } ], "id": "action_group_001", "segment_count": 2, "caption_indices": [ 1, 2 ] }, { "action": "smiles", "types": [ "simple" ], "segments": [ { "caption_index": 1, "loom_clip_id": "v_-DaX_1bBNAo_1_345", "action_id": "action_1", "mentions": [ { "text": "smiles", "ranges": [ [ 92, 98 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "interval": { "start_s": 0.0, "end_s": 3.0, "source_start_s": 0.0, "source_end_s": 3.0 }, "confidence": "high", "temporal_status": "supported" }, { "caption_index": 2, "loom_clip_id": "v_-DaX_1bBNAo_345_590", "action_id": "action_1", "mentions": [ { "text": "smiles", "ranges": [ [ 11, 17 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "interval": { "start_s": 0.0, "end_s": 8.190000000000001, "source_start_s": 11.5, "source_end_s": 19.69 }, "confidence": "high", "temporal_status": "localized" } ], "id": "action_group_002", "segment_count": 2, "caption_indices": [ 1, 2 ] }, { "action": "raise", "types": [ "simple" ], "segments": [ { "caption_index": 1, "loom_clip_id": "v_-DaX_1bBNAo_1_345", "action_id": "action_3", "mentions": [ { "text": "raise", "ranges": [ [ 133, 138 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "interval": { "start_s": 2.0, "end_s": 4.0, "source_start_s": 2.0, "source_end_s": 4.0 }, "confidence": "high", "temporal_status": "supported" }, { "caption_index": 1, "loom_clip_id": "v_-DaX_1bBNAo_1_345", "action_id": "action_6", "mentions": [ { "text": "raise", "ranges": [ [ 236, 241 ] ], "occurrence": 2, "particle": { "text": null, "occurrence": 1 } } ], "interval": { "start_s": 6.0, "end_s": 8.0, "source_start_s": 6.0, "source_end_s": 8.0 }, "confidence": "high", "temporal_status": "supported" } ], "id": "action_group_003", "segment_count": 2, "caption_indices": [ 1 ] }, { "action": "hold", "types": [ "simple" ], "segments": [ { "caption_index": 1, "loom_clip_id": "v_-DaX_1bBNAo_1_345", "action_id": "action_4", "mentions": [ { "text": "hold", "ranges": [ [ 184, 188 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "interval": { "start_s": 3.0, "end_s": 6.0, "source_start_s": 3.0, "source_end_s": 6.0 }, "confidence": "high", "temporal_status": "supported" } ], "id": "action_group_004", "segment_count": 1, "caption_indices": [ 1 ] }, { "action": "lowering", "types": [ "simple" ], "segments": [ { "caption_index": 1, "loom_clip_id": "v_-DaX_1bBNAo_1_345", "action_id": "action_5", "mentions": [ { "text": "lowering", "ranges": [ [ 218, 226 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "interval": { "start_s": 5.0, "end_s": 7.0, "source_start_s": 5.0, "source_end_s": 7.0 }, "confidence": "high", "temporal_status": "supported" } ], "id": "action_group_005", "segment_count": 1, "caption_indices": [ 1 ] }, { "action": "bringing", "types": [ "simple" ], "segments": [ { "caption_index": 1, "loom_clip_id": "v_-DaX_1bBNAo_1_345", "action_id": "action_7", "mentions": [ { "text": "bringing", "ranges": [ [ 291, 299 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "interval": { "start_s": 9.0, "end_s": 11.0, "source_start_s": 9.0, "source_end_s": 11.0 }, "confidence": "high", "temporal_status": "supported" }, { "caption_index": 2, "loom_clip_id": "v_-DaX_1bBNAo_345_590", "action_id": "action_3", "mentions": [ { "text": "bringing", "ranges": [ [ 54, 62 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "interval": { "start_s": 3.0, "end_s": 5.0, "source_start_s": 14.5, "source_end_s": 16.5 }, "confidence": "high", "temporal_status": "localized" } ], "id": "action_group_006", "segment_count": 2, "caption_indices": [ 1, 2 ] }, { "action": "placing", "types": [ "simple" ], "segments": [ { "caption_index": 2, "loom_clip_id": "v_-DaX_1bBNAo_345_590", "action_id": "action_5", "mentions": [ { "text": "placing", "ranges": [ [ 126, 133 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "interval": { "start_s": 6.0, "end_s": 8.0, "source_start_s": 17.5, "source_end_s": 19.5 }, "confidence": "high", "temporal_status": "localized" } ], "id": "action_group_007", "segment_count": 1, "caption_indices": [ 2 ] }, { "action": "pose", "types": [ "posture" ], "segments": [ { "caption_index": 2, "loom_clip_id": "v_-DaX_1bBNAo_345_590", "action_id": "action_4", "mentions": [ { "text": "pose", "ranges": [ [ 118, 122 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "interval": { "start_s": 7.0, "end_s": 8.0, "source_start_s": 18.5, "source_end_s": 19.5 }, "confidence": "high", "temporal_status": "localized" } ], "id": "action_group_008", "segment_count": 1, "caption_indices": [ 2 ] }, { "action": "leans", "types": [ "posture" ], "segments": [ { "caption_index": 2, "loom_clip_id": "v_-DaX_1bBNAo_345_590", "action_id": "action_6", "mentions": [ { "text": "leans", "ranges": [ [ 180, 185 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "interval": { "start_s": 8.0, "end_s": 8.190000000000001, "source_start_s": 19.5, "source_end_s": 19.69 }, "confidence": "medium", "temporal_status": "localized" } ], "id": "action_group_009", "segment_count": 1, "caption_indices": [ 2 ] }, { "action": "moves", "types": [ "simple" ], "segments": [ { "caption_index": 2, "loom_clip_id": "v_-DaX_1bBNAo_345_590", "action_id": "action_7", "mentions": [ { "text": "moves", "ranges": [ [ 198, 203 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "interval": { "start_s": 8.0, "end_s": 8.190000000000001, "source_start_s": 19.5, "source_end_s": 19.69 }, "confidence": "high", "temporal_status": "localized" } ], "id": "action_group_010", "segment_count": 1, "caption_indices": [ 2 ] } ] }, { "source_video_id": "v_-IkL9Kq9zfQ", "grouping_rule": "Exact mention phrase set after case/whitespace normalization; action type is retained as metadata but does not split groups.", "actions": [ { "action": "completes one full stroke", "types": [ "light_verb" ], "segments": [ { "caption_index": 1, "loom_clip_id": "v_-IkL9Kq9zfQ_1_182", "action_id": "A7", "mentions": [ { "text": "completes one full stroke", "ranges": [ [ 209, 234 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "interval": { "start_s": 0.0, "end_s": 4.0, "source_start_s": 0.0, "source_end_s": 4.0 }, "confidence": "medium", "temporal_status": "supported" } ], "id": "action_group_001", "segment_count": 1, "caption_indices": [ 1 ] }, { "action": "maintaining", "types": [ "simple" ], "segments": [ { "caption_index": 1, "loom_clip_id": "v_-IkL9Kq9zfQ_1_182", "action_id": "A8", "mentions": [ { "text": "maintaining", "ranges": [ [ 255, 266 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "interval": { "start_s": 0.0, "end_s": 7.0, "source_start_s": 0.0, "source_end_s": 7.0 }, "confidence": "medium", "temporal_status": "supported" } ], "id": "action_group_002", "segment_count": 1, "caption_indices": [ 1 ] }, { "action": "navigate", "types": [ "simple" ], "segments": [ { "caption_index": 1, "loom_clip_id": "v_-IkL9Kq9zfQ_1_182", "action_id": "A9", "mentions": [ { "text": "navigate", "ranges": [ [ 295, 303 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "interval": { "start_s": 0.0, "end_s": 7.28, "source_start_s": 0.0, "source_end_s": 7.28 }, "confidence": "high", "temporal_status": "supported" }, { "caption_index": 2, "loom_clip_id": "v_-IkL9Kq9zfQ_291_604", "action_id": "action_4", "mentions": [ { "text": "navigate", "ranges": [ [ 176, 184 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "interval": { "start_s": 1.0, "end_s": 12.5, "source_start_s": 12.64, "source_end_s": 24.14 }, "confidence": "high", "temporal_status": "localized" } ], "id": "action_group_003", "segment_count": 2, "caption_indices": [ 1, 2 ] }, { "action": "paddling", "types": [ "simple" ], "segments": [ { "caption_index": 1, "loom_clip_id": "v_-IkL9Kq9zfQ_1_182", "action_id": "A3", "mentions": [ { "text": "paddling", "ranges": [ [ 85, 93 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "interval": { "start_s": 0.0, "end_s": 7.0, "source_start_s": 0.0, "source_end_s": 7.0 }, "confidence": "high", "temporal_status": "supported" }, { "caption_index": 2, "loom_clip_id": "v_-IkL9Kq9zfQ_291_604", "action_id": "action_2", "mentions": [ { "text": "paddling", "ranges": [ [ 89, 97 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "interval": { "start_s": 1.0, "end_s": 12.5, "source_start_s": 12.64, "source_end_s": 24.14 }, "confidence": "high", "temporal_status": "localized" }, { "caption_index": 3, "loom_clip_id": "v_-IkL9Kq9zfQ_604_1034", "action_id": "action_1", "mentions": [ { "text": "paddling", "ranges": [ [ 25, 33 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "interval": { "start_s": 0.0, "end_s": 4.0, "source_start_s": 24.14, "source_end_s": 28.14 }, "confidence": "high", "temporal_status": "supported" }, { "caption_index": 3, "loom_clip_id": "v_-IkL9Kq9zfQ_604_1034", "action_id": "action_3", "mentions": [ { "text": "paddling", "ranges": [ [ 109, 117 ] ], "occurrence": 2, "particle": { "text": null, "occurrence": 1 } } ], "interval": { "start_s": 5.0, "end_s": 10.0, "source_start_s": 29.14, "source_end_s": 34.14 }, "confidence": "high", "temporal_status": "supported" }, { "caption_index": 3, "loom_clip_id": "v_-IkL9Kq9zfQ_604_1034", "action_id": "action_3", "mentions": [ { "text": "paddling", "ranges": [ [ 109, 117 ] ], "occurrence": 2, "particle": { "text": null, "occurrence": 1 } } ], "interval": { "start_s": 12.0, "end_s": 17.22, "source_start_s": 36.14, "source_end_s": 41.36 }, "confidence": "high", "temporal_status": "supported" } ], "id": "action_group_004", "segment_count": 5, "caption_indices": [ 1, 2, 3 ] }, { "action": "propel", "types": [ "simple" ], "segments": [ { "caption_index": 1, "loom_clip_id": "v_-IkL9Kq9zfQ_1_182", "action_id": "A6", "mentions": [ { "text": "propel", "ranges": [ [ 172, 178 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "interval": { "start_s": 0.0, "end_s": 7.28, "source_start_s": 0.0, "source_end_s": 7.28 }, "confidence": "high", "temporal_status": "supported" } ], "id": "action_group_005", "segment_count": 1, "caption_indices": [ 1 ] }, { "action": "pulling", "types": [ "simple" ], "segments": [ { "caption_index": 1, "loom_clip_id": "v_-IkL9Kq9zfQ_1_182", "action_id": "A5", "mentions": [ { "text": "pulling", "ranges": [ [ 140, 147 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "interval": { "start_s": 0.0, "end_s": 3.0, "source_start_s": 0.0, "source_end_s": 3.0 }, "confidence": "high", "temporal_status": "supported" }, { "caption_index": 1, "loom_clip_id": "v_-IkL9Kq9zfQ_1_182", "action_id": "A5", "mentions": [ { "text": "pulling", "ranges": [ [ 140, 147 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "interval": { "start_s": 4.0, "end_s": 7.0, "source_start_s": 4.0, "source_end_s": 7.0 }, "confidence": "high", "temporal_status": "supported" } ], "id": "action_group_006", "segment_count": 2, "caption_indices": [ 1 ] }, { "action": "sits", "types": [ "posture" ], "segments": [ { "caption_index": 1, "loom_clip_id": "v_-IkL9Kq9zfQ_1_182", "action_id": "A2", "mentions": [ { "text": "sits", "ranges": [ [ 64, 68 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "interval": { "start_s": 0.0, "end_s": 7.28, "source_start_s": 0.0, "source_end_s": 7.28 }, "confidence": "high", "temporal_status": "supported" } ], "id": "action_group_007", "segment_count": 1, "caption_indices": [ 1 ] }, { "action": "use", "types": [ "simple" ], "segments": [ { "caption_index": 1, "loom_clip_id": "v_-IkL9Kq9zfQ_1_182", "action_id": "A4", "mentions": [ { "text": "use", "ranges": [ [ 120, 123 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "interval": { "start_s": 0.0, "end_s": 7.0, "source_start_s": 0.0, "source_end_s": 7.0 }, "confidence": "high", "temporal_status": "supported" }, { "caption_index": 2, "loom_clip_id": "v_-IkL9Kq9zfQ_291_604", "action_id": "action_3", "mentions": [ { "text": "use", "ranges": [ [ 129, 132 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "interval": { "start_s": 1.0, "end_s": 12.5, "source_start_s": 12.64, "source_end_s": 24.14 }, "confidence": "high", "temporal_status": "localized" } ], "id": "action_group_008", "segment_count": 2, "caption_indices": [ 1, 2 ] }, { "action": "wearing", "types": [ "simple" ], "segments": [ { "caption_index": 1, "loom_clip_id": "v_-IkL9Kq9zfQ_1_182", "action_id": "A1", "mentions": [ { "text": "wearing", "ranges": [ [ 9, 16 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "interval": { "start_s": 0.0, "end_s": 7.28, "source_start_s": 0.0, "source_end_s": 7.28 }, "confidence": "high", "temporal_status": "supported" }, { "caption_index": 2, "loom_clip_id": "v_-IkL9Kq9zfQ_291_604", "action_id": "action_1", "mentions": [ { "text": "wearing", "ranges": [ [ 9, 16 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "interval": { "start_s": 1.0, "end_s": 12.5, "source_start_s": 12.64, "source_end_s": 24.14 }, "confidence": "high", "temporal_status": "localized" } ], "id": "action_group_009", "segment_count": 2, "caption_indices": [ 1, 2 ] }, { "action": "move", "types": [ "simple" ], "segments": [ { "caption_index": 2, "loom_clip_id": "v_-IkL9Kq9zfQ_291_604", "action_id": "action_5", "mentions": [ { "text": "move", "ranges": [ [ 231, 235 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "interval": { "start_s": 1.0, "end_s": 12.5, "source_start_s": 12.64, "source_end_s": 24.14 }, "confidence": "high", "temporal_status": "localized" } ], "id": "action_group_010", "segment_count": 1, "caption_indices": [ 2 ] }, { "action": "raises", "types": [ "simple" ], "segments": [ { "caption_index": 3, "loom_clip_id": "v_-IkL9Kq9zfQ_604_1034", "action_id": "action_2", "mentions": [ { "text": "raises", "ranges": [ [ 38, 44 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "interval": { "start_s": 4.0, "end_s": 5.0, "source_start_s": 28.14, "source_end_s": 29.14 }, "confidence": "medium", "temporal_status": "supported" } ], "id": "action_group_011", "segment_count": 1, "caption_indices": [ 3 ] } ] }, { "source_video_id": "v_-Sp-gOEGy2o", "grouping_rule": "Exact mention phrase set after case/whitespace normalization; action type is retained as metadata but does not split groups.", "actions": [ { "action": "keeps", "types": [ "simple" ], "segments": [ { "caption_index": 1, "loom_clip_id": "v_-Sp-gOEGy2o_192_387", "action_id": "action_6", "mentions": [ { "text": "keeps", "ranges": [ [ 151, 156 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "interval": { "start_s": 1.0, "end_s": 2.0, "source_start_s": 7.41, "source_end_s": 8.41 }, "confidence": "low", "temporal_status": "localized" } ], "id": "action_group_001", "segment_count": 1, "caption_indices": [ 1 ] }, { "action": "kneels", "types": [ "posture" ], "segments": [ { "caption_index": 1, "loom_clip_id": "v_-Sp-gOEGy2o_192_387", "action_id": "action_2", "mentions": [ { "text": "kneels", "ranges": [ [ 68, 74 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "interval": { "start_s": 1.0, "end_s": 6.5, "source_start_s": 7.41, "source_end_s": 12.91 }, "confidence": "high", "temporal_status": "localized" } ], "id": "action_group_002", "segment_count": 1, "caption_indices": [ 1 ] }, { "action": "rests", "types": [ "simple" ], "segments": [ { "caption_index": 1, "loom_clip_id": "v_-Sp-gOEGy2o_192_387", "action_id": "action_5", "mentions": [ { "text": "rests", "ranges": [ [ 114, 119 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "interval": { "start_s": 1.0, "end_s": 6.5, "source_start_s": 7.41, "source_end_s": 12.91 }, "confidence": "medium", "temporal_status": "localized" } ], "id": "action_group_003", "segment_count": 1, "caption_indices": [ 1 ] }, { "action": "speaking", "types": [ "simple" ], "segments": [ { "caption_index": 1, "loom_clip_id": "v_-Sp-gOEGy2o_192_387", "action_id": "action_4", "mentions": [ { "text": "speaking", "ranges": [ [ 101, 109 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "interval": { "start_s": 1.0, "end_s": 5.0, "source_start_s": 7.41, "source_end_s": 11.41 }, "confidence": "high", "temporal_status": "localized" }, { "caption_index": 1, "loom_clip_id": "v_-Sp-gOEGy2o_192_387", "action_id": "action_4", "mentions": [ { "text": "speaking", "ranges": [ [ 101, 109 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "interval": { "start_s": 6.0, "end_s": 6.5, "source_start_s": 12.41, "source_end_s": 12.91 }, "confidence": "high", "temporal_status": "localized" } ], "id": "action_group_004", "segment_count": 2, "caption_indices": [ 1 ] }, { "action": "wearing", "types": [ "simple" ], "segments": [ { "caption_index": 1, "loom_clip_id": "v_-Sp-gOEGy2o_192_387", "action_id": "action_1", "mentions": [ { "text": "wearing", "ranges": [ [ 35, 42 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "interval": { "start_s": 1.0, "end_s": 6.5, "source_start_s": 7.41, "source_end_s": 12.91 }, "confidence": "high", "temporal_status": "localized" }, { "caption_index": 11, "loom_clip_id": "v_-Sp-gOEGy2o_4995_5385", "action_id": "A1", "mentions": [ { "text": "wearing", "ranges": [ [ 9, 16 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "interval": { "start_s": 1.0, "end_s": 13.0, "source_start_s": 167.67, "source_end_s": 179.67 }, "confidence": "high", "temporal_status": "localized_from_sparse_samples" } ], "id": "action_group_005", "segment_count": 2, "caption_indices": [ 1, 11 ] }, { "action": "looking", "types": [ "simple", "prepositional" ], "segments": [ { "caption_index": 1, "loom_clip_id": "v_-Sp-gOEGy2o_192_387", "action_id": "action_3", "mentions": [ { "text": "looking", "ranges": [ [ 81, 88 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "interval": { "start_s": 3.0, "end_s": 6.5, "source_start_s": 9.41, "source_end_s": 12.91 }, "confidence": "medium", "temporal_status": "localized" }, { "caption_index": 2, "loom_clip_id": "v_-Sp-gOEGy2o_387_612", "action_id": "action_2", "mentions": [ { "text": "looking", "ranges": [ [ 23, 30 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "interval": { "start_s": 1.0, "end_s": 7.5, "source_start_s": 13.91, "source_end_s": 20.41 }, "confidence": "high", "temporal_status": "supported" }, { "caption_index": 3, "loom_clip_id": "v_-Sp-gOEGy2o_612_821", "action_id": "action_look_1", "mentions": [ { "text": "looking", "ranges": [ [ 122, 129 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "interval": { "start_s": 0.0, "end_s": 6.0, "source_start_s": 20.41, "source_end_s": 26.41 }, "confidence": "high", "temporal_status": "supported_sparse" } ], "id": "action_group_006", "segment_count": 3, "caption_indices": [ 1, 2, 3 ] }, { "action": "leans", "types": [ "posture" ], "segments": [ { "caption_index": 2, "loom_clip_id": "v_-Sp-gOEGy2o_387_612", "action_id": "action_1", "mentions": [ { "text": "leans", "ranges": [ [ 8, 13 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "interval": { "start_s": 1.0, "end_s": 7.5, "source_start_s": 13.91, "source_end_s": 20.41 }, "confidence": "high", "temporal_status": "supported" }, { "caption_index": 10, "loom_clip_id": "v_-Sp-gOEGy2o_4478_4615", "action_id": "action_1", "mentions": [ { "text": "leans", "ranges": [ [ 8, 13 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "interval": { "start_s": 0.0, "end_s": 3.0, "source_start_s": 149.41, "source_end_s": 152.41 }, "confidence": "high", "temporal_status": "supported_by_sampled_frames" } ], "id": "action_group_007", "segment_count": 2, "caption_indices": [ 2, 10 ] }, { "action": "reaches out", "types": [ "phrasal" ], "segments": [ { "caption_index": 2, "loom_clip_id": "v_-Sp-gOEGy2o_387_612", "action_id": "action_3", "mentions": [ { "text": "reaches out", "ranges": [ [ 53, 64 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "interval": { "start_s": 1.0, "end_s": 3.0, "source_start_s": 13.91, "source_end_s": 15.91 }, "confidence": "high", "temporal_status": "supported" } ], "id": "action_group_008", "segment_count": 1, "caption_indices": [ 2 ] }, { "action": "picks up", "types": [ "phrasal" ], "segments": [ { "caption_index": 2, "loom_clip_id": "v_-Sp-gOEGy2o_387_612", "action_id": "action_4", "mentions": [ { "text": "picks up", "ranges": [ [ 86, 94 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "interval": { "start_s": 6.0, "end_s": 7.5, "source_start_s": 18.91, "source_end_s": 20.41 }, "confidence": "high", "temporal_status": "supported" }, { "caption_index": 5, "loom_clip_id": "v_-Sp-gOEGy2o_1106_1391", "action_id": "action_1", "mentions": [ { "text": "picks up", "ranges": [ [ 3, 11 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "interval": { "start_s": 0.0, "end_s": 2.0, "source_start_s": 36.91, "source_end_s": 38.91 }, "confidence": "medium", "temporal_status": "sparsely localized" }, { "caption_index": 6, "loom_clip_id": "v_-Sp-gOEGy2o_1391_1945", "action_id": "action_1", "mentions": [ { "text": "picks up", "ranges": [ [ 3, 11 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "interval": { "start_s": 1.0, "end_s": 3.0, "source_start_s": 47.41, "source_end_s": 49.41 }, "confidence": "high", "temporal_status": "localized" }, { "caption_index": 8, "loom_clip_id": "v_-Sp-gOEGy2o_2560_3279", "action_id": "action_1", "mentions": [ { "text": "picks up", "ranges": [ [ 8, 16 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "interval": null, "confidence": "high", "temporal_status": "uncertain" }, { "caption_index": 12, "loom_clip_id": "v_-Sp-gOEGy2o_5385_6284", "action_id": "action_2", "mentions": [ { "text": "picks up", "ranges": [ [ 34, 42 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "interval": null, "confidence": "high", "temporal_status": "not_observed_in_sampled_frames" } ], "id": "action_group_009", "segment_count": 5, "caption_indices": [ 2, 5, 6, 8, 12 ] }, { "action": "holds", "types": [ "simple" ], "segments": [ { "caption_index": 3, "loom_clip_id": "v_-Sp-gOEGy2o_612_821", "action_id": "action_hold_1", "mentions": [ { "text": "holds", "ranges": [ [ 92, 97 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "interval": { "start_s": 1.0, "end_s": 5.0, "source_start_s": 21.41, "source_end_s": 25.41 }, "confidence": "high", "temporal_status": "supported_sparse" }, { "caption_index": 3, "loom_clip_id": "v_-Sp-gOEGy2o_612_821", "action_id": "action_hold_1", "mentions": [ { "text": "holds", "ranges": [ [ 92, 97 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "interval": { "start_s": 6.0, "end_s": 7.0, "source_start_s": 26.41, "source_end_s": 27.41 }, "confidence": "high", "temporal_status": "supported_sparse" }, { "caption_index": 12, "loom_clip_id": "v_-Sp-gOEGy2o_5385_6284", "action_id": "action_3", "mentions": [ { "text": "holds", "ranges": [ [ 117, 122 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "interval": { "start_s": 0.0, "end_s": 30.0, "source_start_s": 179.67, "source_end_s": 209.67 }, "confidence": "high", "temporal_status": "supported" } ], "id": "action_group_010", "segment_count": 3, "caption_indices": [ 3, 12 ] }, { "action": "places", "types": [ "simple" ], "segments": [ { "caption_index": 4, "loom_clip_id": "v_-Sp-gOEGy2o_821_1106", "action_id": "action_1", "mentions": [ { "text": "places", "ranges": [ [ 18, 24 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "interval": { "start_s": 0.0, "end_s": 8.0, "source_start_s": 27.41, "source_end_s": 35.41 }, "confidence": "medium", "temporal_status": "localized_from_sparse_samples" }, { "caption_index": 6, "loom_clip_id": "v_-Sp-gOEGy2o_1391_1945", "action_id": "action_2", "mentions": [ { "text": "places", "ranges": [ [ 59, 65 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "interval": { "start_s": 3.0, "end_s": 7.0, "source_start_s": 49.41, "source_end_s": 53.41 }, "confidence": "high", "temporal_status": "localized" }, { "caption_index": 7, "loom_clip_id": "v_-Sp-gOEGy2o_1945_2560", "action_id": "action_5", "mentions": [ { "text": "places", "ranges": [ [ 129, 135 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "interval": { "start_s": 19.0, "end_s": 20.5, "source_start_s": 83.91, "source_end_s": 85.41 }, "confidence": "high", "temporal_status": "localized" }, { "caption_index": 8, "loom_clip_id": "v_-Sp-gOEGy2o_2560_3279", "action_id": "action_3", "mentions": [ { "text": "places", "ranges": [ [ 85, 91 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "interval": { "start_s": 20.0, "end_s": 24.0, "source_start_s": 105.41, "source_end_s": 109.41 }, "confidence": "high", "temporal_status": "localized" } ], "id": "action_group_011", "segment_count": 4, "caption_indices": [ 4, 6, 7, 8 ] }, { "action": "igniting", "types": [ "simple" ], "segments": [ { "caption_index": 4, "loom_clip_id": "v_-Sp-gOEGy2o_821_1106", "action_id": "action_2", "mentions": [ { "text": "igniting", "ranges": [ [ 75, 83 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "interval": { "start_s": 8.0, "end_s": 9.0, "source_start_s": 35.41, "source_end_s": 36.41 }, "confidence": "high", "temporal_status": "localized_from_sparse_samples" } ], "id": "action_group_012", "segment_count": 1, "caption_indices": [ 4 ] }, { "action": "add", "types": [ "simple" ], "segments": [ { "caption_index": 4, "loom_clip_id": "v_-Sp-gOEGy2o_821_1106", "action_id": "action_4", "mentions": [ { "text": "add", "ranges": [ [ 140, 143 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "interval": { "start_s": 9.0, "end_s": 9.499999999999996, "source_start_s": 36.41, "source_end_s": 36.91 }, "confidence": "high", "temporal_status": "localized_from_sparse_samples" }, { "caption_index": 6, "loom_clip_id": "v_-Sp-gOEGy2o_1391_1945", "action_id": "action_3", "mentions": [ { "text": "add", "ranges": [ [ 110, 113 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "interval": { "start_s": 7.0, "end_s": 15.0, "source_start_s": 53.41, "source_end_s": 61.41 }, "confidence": "high", "temporal_status": "localized" } ], "id": "action_group_013", "segment_count": 2, "caption_indices": [ 4, 6 ] }, { "action": "uses", "types": [ "simple" ], "segments": [ { "caption_index": 4, "loom_clip_id": "v_-Sp-gOEGy2o_821_1106", "action_id": "action_3", "mentions": [ { "text": "uses", "ranges": [ [ 117, 121 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "interval": { "start_s": 9.0, "end_s": 9.499999999999996, "source_start_s": 36.41, "source_end_s": 36.91 }, "confidence": "high", "temporal_status": "localized_from_sparse_samples" }, { "caption_index": 5, "loom_clip_id": "v_-Sp-gOEGy2o_1106_1391", "action_id": "action_2", "mentions": [ { "text": "uses", "ranges": [ [ 64, 68 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "interval": { "start_s": 7.0, "end_s": 9.5, "source_start_s": 43.91, "source_end_s": 46.41 }, "confidence": "high", "temporal_status": "localized" }, { "caption_index": 7, "loom_clip_id": "v_-Sp-gOEGy2o_1945_2560", "action_id": "action_2", "mentions": [ { "text": "uses", "ranges": [ [ 51, 55 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "interval": { "start_s": 16.0, "end_s": 19.0, "source_start_s": 80.91, "source_end_s": 83.91 }, "confidence": "medium", "temporal_status": "localized" } ], "id": "action_group_014", "segment_count": 3, "caption_indices": [ 4, 5, 7 ] }, { "action": "pick up", "types": [ "phrasal" ], "segments": [ { "caption_index": 5, "loom_clip_id": "v_-Sp-gOEGy2o_1106_1391", "action_id": "action_3", "mentions": [ { "text": "pick up", "ranges": [ [ 87, 94 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "interval": { "start_s": 7.0, "end_s": 9.0, "source_start_s": 43.91, "source_end_s": 45.91 }, "confidence": "high", "temporal_status": "localized" }, { "caption_index": 7, "loom_clip_id": "v_-Sp-gOEGy2o_1945_2560", "action_id": "action_3", "mentions": [ { "text": "pick up", "ranges": [ [ 74, 81 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "interval": { "start_s": 16.0, "end_s": 19.0, "source_start_s": 80.91, "source_end_s": 83.91 }, "confidence": "high", "temporal_status": "localized" }, { "caption_index": 10, "loom_clip_id": "v_-Sp-gOEGy2o_4478_4615", "action_id": "action_3", "mentions": [ { "text": "pick up", "ranges": [ [ 69, 76 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "interval": { "start_s": 2.0, "end_s": 4.0, "source_start_s": 151.41, "source_end_s": 153.41 }, "confidence": "high", "temporal_status": "supported_by_sampled_frames" } ], "id": "action_group_015", "segment_count": 3, "caption_indices": [ 5, 7, 10 ] }, { "action": "place", "types": [ "simple" ], "segments": [ { "caption_index": 5, "loom_clip_id": "v_-Sp-gOEGy2o_1106_1391", "action_id": "action_4", "mentions": [ { "text": "place", "ranges": [ [ 127, 132 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "interval": { "start_s": 8.0, "end_s": 9.5, "source_start_s": 44.91, "source_end_s": 46.41 }, "confidence": "high", "temporal_status": "localized" }, { "caption_index": 11, "loom_clip_id": "v_-Sp-gOEGy2o_4995_5385", "action_id": "A5", "mentions": [ { "text": "place", "ranges": [ [ 163, 168 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "interval": { "start_s": 5.0, "end_s": 9.0, "source_start_s": 171.67, "source_end_s": 175.67 }, "confidence": "high", "temporal_status": "localized_from_sparse_samples" } ], "id": "action_group_016", "segment_count": 2, "caption_indices": [ 5, 11 ] }, { "action": "holding", "types": [ "simple" ], "segments": [ { "caption_index": 7, "loom_clip_id": "v_-Sp-gOEGy2o_1945_2560", "action_id": "action_1", "mentions": [ { "text": "holding", "ranges": [ [ 6, 13 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "interval": { "start_s": 0.0, "end_s": 20.5, "source_start_s": 64.91, "source_end_s": 85.41 }, "confidence": "high", "temporal_status": "localized" }, { "caption_index": 9, "loom_clip_id": "v_-Sp-gOEGy2o_3279_4478", "action_id": "action_2", "mentions": [ { "text": "holding", "ranges": [ [ 19, 26 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "interval": { "start_s": 34.0, "end_s": 35.0, "source_start_s": 143.41, "source_end_s": 144.41 }, "confidence": "medium", "temporal_status": "localized" }, { "caption_index": 11, "loom_clip_id": "v_-Sp-gOEGy2o_4995_5385", "action_id": "A3", "mentions": [ { "text": "holding", "ranges": [ [ 98, 105 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "interval": { "start_s": 1.0, "end_s": 13.0, "source_start_s": 167.67, "source_end_s": 179.67 }, "confidence": "high", "temporal_status": "localized_from_sparse_samples" }, { "caption_index": 13, "loom_clip_id": "v_-Sp-gOEGy2o_6284_6582", "action_id": "action_6", "mentions": [ { "text": "holding", "ranges": [ [ 219, 226 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "interval": { "start_s": 0.0, "end_s": 9.950000000000017, "source_start_s": 209.67, "source_end_s": 219.62 }, "confidence": "high", "temporal_status": "localized" } ], "id": "action_group_017", "segment_count": 4, "caption_indices": [ 7, 9, 11, 13 ] }, { "action": "positions", "types": [ "simple" ], "segments": [ { "caption_index": 7, "loom_clip_id": "v_-Sp-gOEGy2o_1945_2560", "action_id": "action_4", "mentions": [ { "text": "positions", "ranges": [ [ 115, 124 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "interval": { "start_s": 18.0, "end_s": 20.5, "source_start_s": 82.91, "source_end_s": 85.41 }, "confidence": "high", "temporal_status": "localized" } ], "id": "action_group_018", "segment_count": 1, "caption_indices": [ 7 ] }, { "action": "lowers", "types": [ "simple" ], "segments": [ { "caption_index": 8, "loom_clip_id": "v_-Sp-gOEGy2o_2560_3279", "action_id": "action_2", "mentions": [ { "text": "lowers", "ranges": [ [ 71, 77 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "interval": { "start_s": 8.0, "end_s": 22.0, "source_start_s": 93.41, "source_end_s": 107.41 }, "confidence": "high", "temporal_status": "localized" }, { "caption_index": 13, "loom_clip_id": "v_-Sp-gOEGy2o_6284_6582", "action_id": "action_2", "mentions": [ { "text": "lowers", "ranges": [ [ 39, 45 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "interval": { "start_s": 0.0, "end_s": 3.0, "source_start_s": 209.67, "source_end_s": 212.67 }, "confidence": "medium", "temporal_status": "localized" } ], "id": "action_group_019", "segment_count": 2, "caption_indices": [ 8, 13 ] }, { "action": "building up", "types": [ "phrasal" ], "segments": [ { "caption_index": 8, "loom_clip_id": "v_-Sp-gOEGy2o_2560_3279", "action_id": "action_4", "mentions": [ { "text": "building up", "ranges": [ [ 145, 156 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "interval": { "start_s": 20.0, "end_s": 24.0, "source_start_s": 105.41, "source_end_s": 109.41 }, "confidence": "high", "temporal_status": "localized" } ], "id": "action_group_020", "segment_count": 1, "caption_indices": [ 8 ] }, { "action": "resting", "types": [ "posture" ], "segments": [ { "caption_index": 9, "loom_clip_id": "v_-Sp-gOEGy2o_3279_4478", "action_id": "action_3", "mentions": [ { "text": "resting", "ranges": [ [ 67, 74 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "interval": { "start_s": 0.0, "end_s": 2.0, "source_start_s": 109.41, "source_end_s": 111.41 }, "confidence": "high", "temporal_status": "localized" }, { "caption_index": 9, "loom_clip_id": "v_-Sp-gOEGy2o_3279_4478", "action_id": "action_3", "mentions": [ { "text": "resting", "ranges": [ [ 67, 74 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "interval": { "start_s": 10.0, "end_s": 14.0, "source_start_s": 119.41, "source_end_s": 123.41 }, "confidence": "high", "temporal_status": "localized" }, { "caption_index": 9, "loom_clip_id": "v_-Sp-gOEGy2o_3279_4478", "action_id": "action_3", "mentions": [ { "text": "resting", "ranges": [ [ 67, 74 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "interval": { "start_s": 30.0, "end_s": 33.0, "source_start_s": 139.41, "source_end_s": 142.41 }, "confidence": "high", "temporal_status": "localized" } ], "id": "action_group_021", "segment_count": 3, "caption_indices": [ 9 ] }, { "action": "sits", "types": [ "posture" ], "segments": [ { "caption_index": 9, "loom_clip_id": "v_-Sp-gOEGy2o_3279_4478", "action_id": "action_1", "mentions": [ { "text": "sits", "ranges": [ [ 8, 12 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "interval": { "start_s": 1.0, "end_s": 14.0, "source_start_s": 110.41, "source_end_s": 123.41 }, "confidence": "high", "temporal_status": "localized" }, { "caption_index": 9, "loom_clip_id": "v_-Sp-gOEGy2o_3279_4478", "action_id": "action_1", "mentions": [ { "text": "sits", "ranges": [ [ 8, 12 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "interval": { "start_s": 32.0, "end_s": 40.0, "source_start_s": 141.41, "source_end_s": 149.41 }, "confidence": "high", "temporal_status": "localized" } ], "id": "action_group_022", "segment_count": 2, "caption_indices": [ 9 ] }, { "action": "looks", "types": [ "prepositional" ], "segments": [ { "caption_index": 9, "loom_clip_id": "v_-Sp-gOEGy2o_3279_4478", "action_id": "action_4", "mentions": [ { "text": "looks", "ranges": [ [ 113, 118 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "interval": { "start_s": 3.0, "end_s": 8.0, "source_start_s": 112.41, "source_end_s": 117.41 }, "confidence": "high", "temporal_status": "localized" }, { "caption_index": 9, "loom_clip_id": "v_-Sp-gOEGy2o_3279_4478", "action_id": "action_4", "mentions": [ { "text": "looks", "ranges": [ [ 113, 118 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "interval": { "start_s": 35.0, "end_s": 40.0, "source_start_s": 144.41, "source_end_s": 149.41 }, "confidence": "high", "temporal_status": "localized" } ], "id": "action_group_023", "segment_count": 2, "caption_indices": [ 9 ] }, { "action": "speaks", "types": [ "simple" ], "segments": [ { "caption_index": 9, "loom_clip_id": "v_-Sp-gOEGy2o_3279_4478", "action_id": "action_5", "mentions": [ { "text": "speaks", "ranges": [ [ 152, 158 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "interval": { "start_s": 3.0, "end_s": 7.0, "source_start_s": 112.41, "source_end_s": 116.41 }, "confidence": "medium", "temporal_status": "localized" }, { "caption_index": 9, "loom_clip_id": "v_-Sp-gOEGy2o_3279_4478", "action_id": "action_5", "mentions": [ { "text": "speaks", "ranges": [ [ 152, 158 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "interval": { "start_s": 35.0, "end_s": 38.0, "source_start_s": 144.41, "source_end_s": 147.41 }, "confidence": "medium", "temporal_status": "localized" } ], "id": "action_group_024", "segment_count": 2, "caption_indices": [ 9 ] }, { "action": "moving", "types": [ "simple" ], "segments": [ { "caption_index": 10, "loom_clip_id": "v_-Sp-gOEGy2o_4478_4615", "action_id": "action_2", "mentions": [ { "text": "moving", "ranges": [ [ 29, 35 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "interval": { "start_s": 1.0, "end_s": 4.0, "source_start_s": 150.41, "source_end_s": 153.41 }, "confidence": "high", "temporal_status": "supported_by_sampled_frames" } ], "id": "action_group_025", "segment_count": 1, "caption_indices": [ 10 ] }, { "action": "squats", "types": [ "posture" ], "segments": [ { "caption_index": 11, "loom_clip_id": "v_-Sp-gOEGy2o_4995_5385", "action_id": "A2", "mentions": [ { "text": "squats", "ranges": [ [ 70, 76 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "interval": { "start_s": 1.0, "end_s": 13.0, "source_start_s": 167.67, "source_end_s": 179.67 }, "confidence": "high", "temporal_status": "localized_from_sparse_samples" } ], "id": "action_group_026", "segment_count": 1, "caption_indices": [ 11 ] }, { "action": "use", "types": [ "simple" ], "segments": [ { "caption_index": 11, "loom_clip_id": "v_-Sp-gOEGy2o_4995_5385", "action_id": "A4", "mentions": [ { "text": "use", "ranges": [ [ 136, 139 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "interval": { "start_s": 5.0, "end_s": 9.0, "source_start_s": 171.67, "source_end_s": 175.67 }, "confidence": "high", "temporal_status": "localized_from_sparse_samples" } ], "id": "action_group_027", "segment_count": 1, "caption_indices": [ 11 ] }, { "action": "watching", "types": [ "simple" ], "segments": [ { "caption_index": 11, "loom_clip_id": "v_-Sp-gOEGy2o_4995_5385", "action_id": "A7", "mentions": [ { "text": "watching", "ranges": [ [ 238, 246 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "interval": { "start_s": 5.0, "end_s": 6.0, "source_start_s": 171.67, "source_end_s": 172.67 }, "confidence": "medium", "temporal_status": "localized_from_sparse_samples" }, { "caption_index": 11, "loom_clip_id": "v_-Sp-gOEGy2o_4995_5385", "action_id": "A7", "mentions": [ { "text": "watching", "ranges": [ [ 238, 246 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "interval": { "start_s": 7.0, "end_s": 13.0, "source_start_s": 173.67, "source_end_s": 179.67 }, "confidence": "medium", "temporal_status": "localized_from_sparse_samples" } ], "id": "action_group_028", "segment_count": 2, "caption_indices": [ 11 ] }, { "action": "hold", "types": [ "simple" ], "segments": [ { "caption_index": 11, "loom_clip_id": "v_-Sp-gOEGy2o_4995_5385", "action_id": "A6", "mentions": [ { "text": "hold", "ranges": [ [ 215, 219 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "interval": { "start_s": 9.0, "end_s": 13.0, "source_start_s": 175.67, "source_end_s": 179.67 }, "confidence": "high", "temporal_status": "localized_from_sparse_samples" } ], "id": "action_group_029", "segment_count": 1, "caption_indices": [ 11 ] }, { "action": "aim", "types": [ "prepositional" ], "segments": [ { "caption_index": 12, "loom_clip_id": "v_-Sp-gOEGy2o_5385_6284", "action_id": "action_4", "mentions": [ { "text": "aim", "ranges": [ [ 156, 159 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "interval": { "start_s": 0.0, "end_s": 30.0, "source_start_s": 179.67, "source_end_s": 209.67 }, "confidence": "high", "temporal_status": "supported" } ], "id": "action_group_030", "segment_count": 1, "caption_indices": [ 12 ] }, { "action": "pump · work", "types": [ "simple" ], "segments": [ { "caption_index": 12, "loom_clip_id": "v_-Sp-gOEGy2o_5385_6284", "action_id": "action_5", "mentions": [ { "text": "pump", "ranges": [ [ 211, 215 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } }, { "text": "work", "ranges": [ [ 282, 286 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "interval": { "start_s": 0.0, "end_s": 30.0, "source_start_s": 179.67, "source_end_s": 209.67 }, "confidence": "high", "temporal_status": "supported" } ], "id": "action_group_031", "segment_count": 1, "caption_indices": [ 12 ] }, { "action": "stoke", "types": [ "simple" ], "segments": [ { "caption_index": 12, "loom_clip_id": "v_-Sp-gOEGy2o_5385_6284", "action_id": "action_6", "mentions": [ { "text": "stoke", "ranges": [ [ 222, 227 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "interval": { "start_s": 0.0, "end_s": 30.0, "source_start_s": 179.67, "source_end_s": 209.67 }, "confidence": "medium", "temporal_status": "supported" } ], "id": "action_group_032", "segment_count": 1, "caption_indices": [ 12 ] }, { "action": "observing", "types": [ "simple" ], "segments": [ { "caption_index": 13, "loom_clip_id": "v_-Sp-gOEGy2o_6284_6582", "action_id": "action_3", "mentions": [ { "text": "observing", "ranges": [ [ 50, 59 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "interval": { "start_s": 0.0, "end_s": 5.0, "source_start_s": 209.67, "source_end_s": 214.67 }, "confidence": "high", "temporal_status": "localized" } ], "id": "action_group_033", "segment_count": 1, "caption_indices": [ 13 ] }, { "action": "seated", "types": [ "posture" ], "segments": [ { "caption_index": 13, "loom_clip_id": "v_-Sp-gOEGy2o_6284_6582", "action_id": "action_5", "mentions": [ { "text": "seated", "ranges": [ [ 195, 201 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "interval": { "start_s": 0.0, "end_s": 9.950000000000017, "source_start_s": 209.67, "source_end_s": 219.62 }, "confidence": "high", "temporal_status": "localized" } ], "id": "action_group_034", "segment_count": 1, "caption_indices": [ 13 ] }, { "action": "using", "types": [ "simple" ], "segments": [ { "caption_index": 13, "loom_clip_id": "v_-Sp-gOEGy2o_6284_6582", "action_id": "action_1", "mentions": [ { "text": "using", "ranges": [ [ 17, 22 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "interval": { "start_s": 0.0, "end_s": 2.0, "source_start_s": 209.67, "source_end_s": 211.67 }, "confidence": "medium", "temporal_status": "localized" } ], "id": "action_group_035", "segment_count": 1, "caption_indices": [ 13 ] }, { "action": "bring", "types": [ "simple" ], "segments": [ { "caption_index": 13, "loom_clip_id": "v_-Sp-gOEGy2o_6284_6582", "action_id": "action_4", "mentions": [ { "text": "bring", "ranges": [ [ 104, 109 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "interval": { "start_s": 3.0, "end_s": 5.0, "source_start_s": 212.67, "source_end_s": 214.67 }, "confidence": "high", "temporal_status": "localized" } ], "id": "action_group_036", "segment_count": 1, "caption_indices": [ 13 ] }, { "action": "lets go", "types": [ "phrasal" ], "segments": [ { "caption_index": 12, "loom_clip_id": "v_-Sp-gOEGy2o_5385_6284", "action_id": "action_1", "mentions": [ { "text": "lets go", "ranges": [ [ 11, 18 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "interval": null, "confidence": "high", "temporal_status": "not_observed_in_sampled_frames" } ], "id": "action_group_037", "segment_count": 1, "caption_indices": [ 12 ] }, { "action": "strike", "types": [ "simple" ], "segments": [ { "caption_index": 2, "loom_clip_id": "v_-Sp-gOEGy2o_387_612", "action_id": "action_5", "mentions": [ { "text": "strike", "ranges": [ [ 154, 160 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "interval": null, "confidence": "high", "temporal_status": "uncertain" } ], "id": "action_group_038", "segment_count": 1, "caption_indices": [ 2 ] }, { "action": "striking", "types": [ "simple" ], "segments": [ { "caption_index": 3, "loom_clip_id": "v_-Sp-gOEGy2o_612_821", "action_id": "action_strike_1", "mentions": [ { "text": "striking", "ranges": [ [ 6, 14 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "interval": null, "confidence": "medium", "temporal_status": "uncertain" } ], "id": "action_group_039", "segment_count": 1, "caption_indices": [ 3 ] }, { "action": "transfers", "types": [ "simple" ], "segments": [ { "caption_index": 3, "loom_clip_id": "v_-Sp-gOEGy2o_612_821", "action_id": "action_transfer_1", "mentions": [ { "text": "transfers", "ranges": [ [ 47, 56 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "interval": null, "confidence": "medium", "temporal_status": "uncertain" } ], "id": "action_group_040", "segment_count": 1, "caption_indices": [ 3 ] }, { "action": "used", "types": [ "simple" ], "segments": [ { "caption_index": 10, "loom_clip_id": "v_-Sp-gOEGy2o_4478_4615", "action_id": "action_4", "mentions": [ { "text": "used", "ranges": [ [ 93, 97 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "interval": null, "confidence": "medium", "temporal_status": "earlier_event_not_supported_in_current_clip" } ], "id": "action_group_041", "segment_count": 1, "caption_indices": [ 10 ] } ] }, { "source_video_id": "v_-YMpwZkNc2A", "grouping_rule": "Exact mention phrase set after case/whitespace normalization; action type is retained as metadata but does not split groups.", "actions": [ { "action": "looks", "types": [ "simple" ], "segments": [ { "caption_index": 1, "loom_clip_id": "v_-YMpwZkNc2A_62_497", "action_id": "action_2", "mentions": [ { "text": "looks", "ranges": [ [ 61, 66 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "interval": { "start_s": 1.0, "end_s": 3.0, "source_start_s": 3.07, "source_end_s": 5.07 }, "confidence": "high", "temporal_status": "localized_sparse" }, { "caption_index": 1, "loom_clip_id": "v_-YMpwZkNc2A_62_497", "action_id": "action_2", "mentions": [ { "text": "looks", "ranges": [ [ 61, 66 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "interval": { "start_s": 4.0, "end_s": 8.0, "source_start_s": 6.07, "source_end_s": 10.07 }, "confidence": "high", "temporal_status": "localized_sparse" }, { "caption_index": 1, "loom_clip_id": "v_-YMpwZkNc2A_62_497", "action_id": "action_2", "mentions": [ { "text": "looks", "ranges": [ [ 61, 66 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "interval": { "start_s": 9.0, "end_s": 11.0, "source_start_s": 11.07, "source_end_s": 13.07 }, "confidence": "high", "temporal_status": "localized_sparse" }, { "caption_index": 12, "loom_clip_id": "v_-YMpwZkNc2A_4458_4717", "action_id": "action_2", "mentions": [ { "text": "looks", "ranges": [ [ 21, 26 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "interval": { "start_s": 1.0, "end_s": 3.0, "source_start_s": 149.75, "source_end_s": 151.75 }, "confidence": "high", "temporal_status": "supported_by_sampled_frames" }, { "caption_index": 12, "loom_clip_id": "v_-YMpwZkNc2A_4458_4717", "action_id": "action_3", "mentions": [ { "text": "looks", "ranges": [ [ 50, 55 ] ], "occurrence": 2, "particle": { "text": null, "occurrence": 1 } } ], "interval": { "start_s": 2.0, "end_s": 4.0, "source_start_s": 150.75, "source_end_s": 152.75 }, "confidence": "high", "temporal_status": "supported_by_sampled_frames" } ], "id": "action_group_001", "segment_count": 5, "caption_indices": [ 1, 12 ] }, { "action": "run", "types": [ "simple" ], "segments": [ { "caption_index": 1, "loom_clip_id": "v_-YMpwZkNc2A_62_497", "action_id": "action_3", "mentions": [ { "text": "run", "ranges": [ [ 78, 81 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "interval": { "start_s": 1.0, "end_s": 2.0, "source_start_s": 3.07, "source_end_s": 4.07 }, "confidence": "high", "temporal_status": "localized_sparse" } ], "id": "action_group_002", "segment_count": 1, "caption_indices": [ 1 ] }, { "action": "wearing", "types": [ "simple" ], "segments": [ { "caption_index": 1, "loom_clip_id": "v_-YMpwZkNc2A_62_497", "action_id": "action_1", "mentions": [ { "text": "wearing", "ranges": [ [ 26, 33 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "interval": { "start_s": 1.0, "end_s": 14.5, "source_start_s": 3.07, "source_end_s": 16.57 }, "confidence": "high", "temporal_status": "localized_sparse" } ], "id": "action_group_003", "segment_count": 1, "caption_indices": [ 1 ] }, { "action": "looking", "types": [ "simple" ], "segments": [ { "caption_index": 1, "loom_clip_id": "v_-YMpwZkNc2A_62_497", "action_id": "action_4", "mentions": [ { "text": "looking", "ranges": [ [ 119, 126 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "interval": { "start_s": 3.0, "end_s": 4.0, "source_start_s": 5.07, "source_end_s": 6.07 }, "confidence": "high", "temporal_status": "localized_sparse" }, { "caption_index": 1, "loom_clip_id": "v_-YMpwZkNc2A_62_497", "action_id": "action_4", "mentions": [ { "text": "looking", "ranges": [ [ 119, 126 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "interval": { "start_s": 8.0, "end_s": 9.0, "source_start_s": 10.07, "source_end_s": 11.07 }, "confidence": "high", "temporal_status": "localized_sparse" }, { "caption_index": 10, "loom_clip_id": "v_-YMpwZkNc2A_3874_4188", "action_id": "action_4", "mentions": [ { "text": "looking", "ranges": [ [ 94, 101 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "interval": { "start_s": 2.0, "end_s": 4.0, "source_start_s": 131.25, "source_end_s": 133.25 }, "confidence": "high", "temporal_status": "supported" } ], "id": "action_group_004", "segment_count": 3, "caption_indices": [ 1, 10 ] }, { "action": "talking · speaking", "types": [ "simple" ], "segments": [ { "caption_index": 1, "loom_clip_id": "v_-YMpwZkNc2A_62_497", "action_id": "action_5", "mentions": [ { "text": "talking", "ranges": [ [ 139, 146 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } }, { "text": "speaking", "ranges": [ [ 193, 201 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "interval": { "start_s": 4.0, "end_s": 14.5, "source_start_s": 6.07, "source_end_s": 16.57 }, "confidence": "medium", "temporal_status": "localized_sparse" } ], "id": "action_group_005", "segment_count": 1, "caption_indices": [ 1 ] }, { "action": "gesture", "types": [ "simple" ], "segments": [ { "caption_index": 1, "loom_clip_id": "v_-YMpwZkNc2A_62_497", "action_id": "action_6", "mentions": [ { "text": "gesture", "ranges": [ [ 158, 165 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "interval": { "start_s": 11.0, "end_s": 14.5, "source_start_s": 13.07, "source_end_s": 16.57 }, "confidence": "high", "temporal_status": "localized_sparse" } ], "id": "action_group_006", "segment_count": 1, "caption_indices": [ 1 ] }, { "action": "picks up", "types": [ "phrasal" ], "segments": [ { "caption_index": 2, "loom_clip_id": "v_-YMpwZkNc2A_497_781", "action_id": "action_1", "mentions": [ { "text": "picks up", "ranges": [ [ 11, 19 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "interval": { "start_s": 0.0, "end_s": 2.0, "source_start_s": 16.57, "source_end_s": 18.57 }, "confidence": "high", "temporal_status": "supported_by_sparse_samples" }, { "caption_index": 11, "loom_clip_id": "v_-YMpwZkNc2A_4188_4458", "action_id": "action_1", "mentions": [ { "text": "picks up", "ranges": [ [ 8, 16 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "interval": { "start_s": 0.0, "end_s": 2.0, "source_start_s": 139.75, "source_end_s": 141.75 }, "confidence": "medium", "temporal_status": "localized" } ], "id": "action_group_007", "segment_count": 2, "caption_indices": [ 2, 11 ] }, { "action": "holding · displaying", "types": [ "phrasal" ], "segments": [ { "caption_index": 2, "loom_clip_id": "v_-YMpwZkNc2A_497_781", "action_id": "action_2", "mentions": [ { "text": "holding", "ranges": [ [ 90, 97 ] ], "occurrence": 1, "particle": { "text": "up", "range": [ 103, 105 ], "occurrence": 2 } }, { "text": "displaying", "ranges": [ [ 124, 134 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "interval": { "start_s": 2.0, "end_s": 6.0, "source_start_s": 18.57, "source_end_s": 22.57 }, "confidence": "high", "temporal_status": "supported_by_sparse_samples" } ], "id": "action_group_008", "segment_count": 1, "caption_indices": [ 2 ] }, { "action": "speak", "types": [ "simple" ], "segments": [ { "caption_index": 2, "loom_clip_id": "v_-YMpwZkNc2A_497_781", "action_id": "action_3", "mentions": [ { "text": "speak", "ranges": [ [ 112, 117 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "interval": { "start_s": 2.0, "end_s": 6.0, "source_start_s": 18.57, "source_end_s": 22.57 }, "confidence": "medium", "temporal_status": "supported_by_sparse_samples" } ], "id": "action_group_009", "segment_count": 1, "caption_indices": [ 2 ] }, { "action": "putting", "types": [ "phrasal" ], "segments": [ { "caption_index": 2, "loom_clip_id": "v_-YMpwZkNc2A_497_781", "action_id": "action_4", "mentions": [ { "text": "putting", "ranges": [ [ 152, 159 ] ], "occurrence": 1, "particle": { "text": "down", "range": [ 165, 169 ], "occurrence": 1 } } ], "interval": { "start_s": 6.0, "end_s": 7.0, "source_start_s": 22.57, "source_end_s": 23.57 }, "confidence": "medium", "temporal_status": "supported_by_sparse_transition" } ], "id": "action_group_010", "segment_count": 1, "caption_indices": [ 2 ] }, { "action": "reappears", "types": [ "simple" ], "segments": [ { "caption_index": 3, "loom_clip_id": "v_-YMpwZkNc2A_781_1036", "action_id": "A1", "mentions": [ { "text": "reappears", "ranges": [ [ 11, 20 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "interval": { "start_s": 1.0, "end_s": 3.0, "source_start_s": 27.07, "source_end_s": 29.07 }, "confidence": "medium", "temporal_status": "sparse_localized" } ], "id": "action_group_011", "segment_count": 1, "caption_indices": [ 3 ] }, { "action": "apply · spread", "types": [ "simple" ], "segments": [ { "caption_index": 3, "loom_clip_id": "v_-YMpwZkNc2A_781_1036", "action_id": "A2", "mentions": [ { "text": "apply", "ranges": [ [ 35, 40 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } }, { "text": "spread", "ranges": [ [ 81, 87 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "interval": { "start_s": 4.0, "end_s": 5.0, "source_start_s": 30.07, "source_end_s": 31.07 }, "confidence": "high", "temporal_status": "sparse_localized" }, { "caption_index": 3, "loom_clip_id": "v_-YMpwZkNc2A_781_1036", "action_id": "A2", "mentions": [ { "text": "apply", "ranges": [ [ 35, 40 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } }, { "text": "spread", "ranges": [ [ 81, 87 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "interval": { "start_s": 6.0, "end_s": 8.5, "source_start_s": 32.07, "source_end_s": 34.57 }, "confidence": "high", "temporal_status": "sparse_localized" } ], "id": "action_group_012", "segment_count": 2, "caption_indices": [ 3 ] }, { "action": "shaving", "types": [ "simple" ], "segments": [ { "caption_index": 4, "loom_clip_id": "v_-YMpwZkNc2A_1036_1576", "action_id": "action_shave_right_cheek", "mentions": [ { "text": "shaving", "ranges": [ [ 18, 25 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "interval": { "start_s": 0.0, "end_s": 1.0, "source_start_s": 34.57, "source_end_s": 35.57 }, "confidence": "medium", "temporal_status": "sparse_supported" }, { "caption_index": 4, "loom_clip_id": "v_-YMpwZkNc2A_1036_1576", "action_id": "action_shave_right_cheek", "mentions": [ { "text": "shaving", "ranges": [ [ 18, 25 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "interval": { "start_s": 3.0, "end_s": 10.0, "source_start_s": 37.57, "source_end_s": 44.57 }, "confidence": "medium", "temporal_status": "sparse_supported" }, { "caption_index": 4, "loom_clip_id": "v_-YMpwZkNc2A_1036_1576", "action_id": "action_shave_neck", "mentions": [ { "text": "shaving", "ranges": [ [ 104, 111 ] ], "occurrence": 2, "particle": { "text": null, "occurrence": 1 } } ], "interval": { "start_s": 11.0, "end_s": 12.0, "source_start_s": 45.57, "source_end_s": 46.57 }, "confidence": "medium", "temporal_status": "sparse_supported" }, { "caption_index": 6, "loom_clip_id": "v_-YMpwZkNc2A_2400_2834", "action_id": "action_1", "mentions": [ { "text": "shaving", "ranges": [ [ 21, 28 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "interval": { "start_s": 0.0, "end_s": 1.0, "source_start_s": 80.07, "source_end_s": 81.07 }, "confidence": "high", "temporal_status": "localized" }, { "caption_index": 6, "loom_clip_id": "v_-YMpwZkNc2A_2400_2834", "action_id": "action_1", "mentions": [ { "text": "shaving", "ranges": [ [ 21, 28 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "interval": { "start_s": 2.0, "end_s": 4.0, "source_start_s": 82.07, "source_end_s": 84.07 }, "confidence": "high", "temporal_status": "localized" }, { "caption_index": 6, "loom_clip_id": "v_-YMpwZkNc2A_2400_2834", "action_id": "action_1", "mentions": [ { "text": "shaving", "ranges": [ [ 21, 28 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "interval": { "start_s": 5.0, "end_s": 6.0, "source_start_s": 85.07, "source_end_s": 86.07 }, "confidence": "high", "temporal_status": "localized" }, { "caption_index": 7, "loom_clip_id": "v_-YMpwZkNc2A_2834_3224", "action_id": "action_shaving", "mentions": [ { "text": "shaving", "ranges": [ [ 20, 27 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "interval": { "start_s": 0.0, "end_s": 2.0, "source_start_s": 94.57, "source_end_s": 96.57 }, "confidence": "high", "temporal_status": "localized" } ], "id": "action_group_013", "segment_count": 7, "caption_indices": [ 4, 6, 7 ] }, { "action": "shave", "types": [ "simple" ], "segments": [ { "caption_index": 4, "loom_clip_id": "v_-YMpwZkNc2A_1036_1576", "action_id": "action_shave_left_cheek", "mentions": [ { "text": "shave", "ranges": [ [ 143, 148 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "interval": { "start_s": 12.0, "end_s": 15.0, "source_start_s": 46.57, "source_end_s": 49.57 }, "confidence": "medium", "temporal_status": "sparse_supported" }, { "caption_index": 4, "loom_clip_id": "v_-YMpwZkNc2A_1036_1576", "action_id": "action_shave_left_cheek", "mentions": [ { "text": "shave", "ranges": [ [ 143, 148 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "interval": { "start_s": 16.0, "end_s": 17.0, "source_start_s": 50.57, "source_end_s": 51.57 }, "confidence": "medium", "temporal_status": "sparse_supported" }, { "caption_index": 5, "loom_clip_id": "v_-YMpwZkNc2A_1576_2400", "action_id": "action_shave_1", "mentions": [ { "text": "shave", "ranges": [ [ 24, 29 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } }, { "text": "shave", "ranges": [ [ 108, 113 ] ], "occurrence": 2, "particle": { "text": null, "occurrence": 1 } } ], "interval": { "start_s": 0.0, "end_s": 2.0, "source_start_s": 52.57, "source_end_s": 54.57 }, "confidence": "high", "temporal_status": "supported" }, { "caption_index": 5, "loom_clip_id": "v_-YMpwZkNc2A_1576_2400", "action_id": "action_shave_1", "mentions": [ { "text": "shave", "ranges": [ [ 24, 29 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } }, { "text": "shave", "ranges": [ [ 108, 113 ] ], "occurrence": 2, "particle": { "text": null, "occurrence": 1 } } ], "interval": { "start_s": 3.0, "end_s": 5.0, "source_start_s": 55.57, "source_end_s": 57.57 }, "confidence": "high", "temporal_status": "supported" }, { "caption_index": 5, "loom_clip_id": "v_-YMpwZkNc2A_1576_2400", "action_id": "action_shave_1", "mentions": [ { "text": "shave", "ranges": [ [ 24, 29 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } }, { "text": "shave", "ranges": [ [ 108, 113 ] ], "occurrence": 2, "particle": { "text": null, "occurrence": 1 } } ], "interval": { "start_s": 6.0, "end_s": 8.0, "source_start_s": 58.57, "source_end_s": 60.57 }, "confidence": "high", "temporal_status": "supported" }, { "caption_index": 5, "loom_clip_id": "v_-YMpwZkNc2A_1576_2400", "action_id": "action_shave_1", "mentions": [ { "text": "shave", "ranges": [ [ 24, 29 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } }, { "text": "shave", "ranges": [ [ 108, 113 ] ], "occurrence": 2, "particle": { "text": null, "occurrence": 1 } } ], "interval": { "start_s": 9.0, "end_s": 10.0, "source_start_s": 61.57, "source_end_s": 62.57 }, "confidence": "high", "temporal_status": "supported" }, { "caption_index": 5, "loom_clip_id": "v_-YMpwZkNc2A_1576_2400", "action_id": "action_shave_1", "mentions": [ { "text": "shave", "ranges": [ [ 24, 29 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } }, { "text": "shave", "ranges": [ [ 108, 113 ] ], "occurrence": 2, "particle": { "text": null, "occurrence": 1 } } ], "interval": { "start_s": 12.0, "end_s": 14.0, "source_start_s": 64.57, "source_end_s": 66.57 }, "confidence": "high", "temporal_status": "supported" }, { "caption_index": 5, "loom_clip_id": "v_-YMpwZkNc2A_1576_2400", "action_id": "action_shave_1", "mentions": [ { "text": "shave", "ranges": [ [ 24, 29 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } }, { "text": "shave", "ranges": [ [ 108, 113 ] ], "occurrence": 2, "particle": { "text": null, "occurrence": 1 } } ], "interval": { "start_s": 16.0, "end_s": 17.0, "source_start_s": 68.57, "source_end_s": 69.57 }, "confidence": "high", "temporal_status": "supported" }, { "caption_index": 5, "loom_clip_id": "v_-YMpwZkNc2A_1576_2400", "action_id": "action_shave_1", "mentions": [ { "text": "shave", "ranges": [ [ 24, 29 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } }, { "text": "shave", "ranges": [ [ 108, 113 ] ], "occurrence": 2, "particle": { "text": null, "occurrence": 1 } } ], "interval": { "start_s": 19.0, "end_s": 21.0, "source_start_s": 71.57, "source_end_s": 73.57 }, "confidence": "high", "temporal_status": "supported" }, { "caption_index": 5, "loom_clip_id": "v_-YMpwZkNc2A_1576_2400", "action_id": "action_shave_1", "mentions": [ { "text": "shave", "ranges": [ [ 24, 29 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } }, { "text": "shave", "ranges": [ [ 108, 113 ] ], "occurrence": 2, "particle": { "text": null, "occurrence": 1 } } ], "interval": { "start_s": 22.0, "end_s": 25.0, "source_start_s": 74.57, "source_end_s": 77.57 }, "confidence": "high", "temporal_status": "supported" }, { "caption_index": 5, "loom_clip_id": "v_-YMpwZkNc2A_1576_2400", "action_id": "action_shave_1", "mentions": [ { "text": "shave", "ranges": [ [ 24, 29 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } }, { "text": "shave", "ranges": [ [ 108, 113 ] ], "occurrence": 2, "particle": { "text": null, "occurrence": 1 } } ], "interval": { "start_s": 26.0, "end_s": 27.499999999999993, "source_start_s": 78.57, "source_end_s": 80.07 }, "confidence": "high", "temporal_status": "supported" }, { "caption_index": 6, "loom_clip_id": "v_-YMpwZkNc2A_2400_2834", "action_id": "action_3", "mentions": [ { "text": "shave", "ranges": [ [ 187, 192 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "interval": { "start_s": 8.0, "end_s": 12.0, "source_start_s": 88.07, "source_end_s": 92.07 }, "confidence": "high", "temporal_status": "localized" }, { "caption_index": 6, "loom_clip_id": "v_-YMpwZkNc2A_2400_2834", "action_id": "action_3", "mentions": [ { "text": "shave", "ranges": [ [ 187, 192 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "interval": { "start_s": 14.0, "end_s": 14.5, "source_start_s": 94.07, "source_end_s": 94.57 }, "confidence": "high", "temporal_status": "localized" } ], "id": "action_group_014", "segment_count": 13, "caption_indices": [ 4, 5, 6 ] }, { "action": "tilt", "types": [ "simple" ], "segments": [ { "caption_index": 5, "loom_clip_id": "v_-YMpwZkNc2A_1576_2400", "action_id": "action_tilt_1", "mentions": [ { "text": "tilt", "ranges": [ [ 84, 88 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "interval": { "start_s": 1.0, "end_s": 2.0, "source_start_s": 53.57, "source_end_s": 54.57 }, "confidence": "high", "temporal_status": "supported" }, { "caption_index": 5, "loom_clip_id": "v_-YMpwZkNc2A_1576_2400", "action_id": "action_tilt_1", "mentions": [ { "text": "tilt", "ranges": [ [ 84, 88 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "interval": { "start_s": 3.0, "end_s": 5.0, "source_start_s": 55.57, "source_end_s": 57.57 }, "confidence": "high", "temporal_status": "supported" }, { "caption_index": 5, "loom_clip_id": "v_-YMpwZkNc2A_1576_2400", "action_id": "action_tilt_1", "mentions": [ { "text": "tilt", "ranges": [ [ 84, 88 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "interval": { "start_s": 6.0, "end_s": 10.0, "source_start_s": 58.57, "source_end_s": 62.57 }, "confidence": "high", "temporal_status": "supported" }, { "caption_index": 5, "loom_clip_id": "v_-YMpwZkNc2A_1576_2400", "action_id": "action_tilt_1", "mentions": [ { "text": "tilt", "ranges": [ [ 84, 88 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "interval": { "start_s": 12.0, "end_s": 14.0, "source_start_s": 64.57, "source_end_s": 66.57 }, "confidence": "high", "temporal_status": "supported" }, { "caption_index": 5, "loom_clip_id": "v_-YMpwZkNc2A_1576_2400", "action_id": "action_tilt_1", "mentions": [ { "text": "tilt", "ranges": [ [ 84, 88 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "interval": { "start_s": 16.0, "end_s": 18.0, "source_start_s": 68.57, "source_end_s": 70.57 }, "confidence": "high", "temporal_status": "supported" }, { "caption_index": 5, "loom_clip_id": "v_-YMpwZkNc2A_1576_2400", "action_id": "action_tilt_1", "mentions": [ { "text": "tilt", "ranges": [ [ 84, 88 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "interval": { "start_s": 19.0, "end_s": 21.0, "source_start_s": 71.57, "source_end_s": 73.57 }, "confidence": "high", "temporal_status": "supported" }, { "caption_index": 5, "loom_clip_id": "v_-YMpwZkNc2A_1576_2400", "action_id": "action_tilt_1", "mentions": [ { "text": "tilt", "ranges": [ [ 84, 88 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "interval": { "start_s": 22.0, "end_s": 25.0, "source_start_s": 74.57, "source_end_s": 77.57 }, "confidence": "high", "temporal_status": "supported" }, { "caption_index": 5, "loom_clip_id": "v_-YMpwZkNc2A_1576_2400", "action_id": "action_tilt_1", "mentions": [ { "text": "tilt", "ranges": [ [ 84, 88 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "interval": { "start_s": 26.0, "end_s": 27.499999999999993, "source_start_s": 78.57, "source_end_s": 80.07 }, "confidence": "high", "temporal_status": "supported" } ], "id": "action_group_015", "segment_count": 8, "caption_indices": [ 5 ] }, { "action": "bending", "types": [ "simple" ], "segments": [ { "caption_index": 5, "loom_clip_id": "v_-YMpwZkNc2A_1576_2400", "action_id": "action_bend_1", "mentions": [ { "text": "bending", "ranges": [ [ 143, 150 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "interval": { "start_s": 2.0, "end_s": 3.0, "source_start_s": 54.57, "source_end_s": 55.57 }, "confidence": "high", "temporal_status": "supported" }, { "caption_index": 5, "loom_clip_id": "v_-YMpwZkNc2A_1576_2400", "action_id": "action_bend_1", "mentions": [ { "text": "bending", "ranges": [ [ 143, 150 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "interval": { "start_s": 5.0, "end_s": 6.0, "source_start_s": 57.57, "source_end_s": 58.57 }, "confidence": "high", "temporal_status": "supported" }, { "caption_index": 5, "loom_clip_id": "v_-YMpwZkNc2A_1576_2400", "action_id": "action_bend_1", "mentions": [ { "text": "bending", "ranges": [ [ 143, 150 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "interval": { "start_s": 10.0, "end_s": 12.0, "source_start_s": 62.57, "source_end_s": 64.57 }, "confidence": "high", "temporal_status": "supported" }, { "caption_index": 5, "loom_clip_id": "v_-YMpwZkNc2A_1576_2400", "action_id": "action_bend_1", "mentions": [ { "text": "bending", "ranges": [ [ 143, 150 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "interval": { "start_s": 14.0, "end_s": 15.0, "source_start_s": 66.57, "source_end_s": 67.57 }, "confidence": "high", "temporal_status": "supported" }, { "caption_index": 5, "loom_clip_id": "v_-YMpwZkNc2A_1576_2400", "action_id": "action_bend_1", "mentions": [ { "text": "bending", "ranges": [ [ 143, 150 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "interval": { "start_s": 18.0, "end_s": 19.0, "source_start_s": 70.57, "source_end_s": 71.57 }, "confidence": "high", "temporal_status": "supported" }, { "caption_index": 5, "loom_clip_id": "v_-YMpwZkNc2A_1576_2400", "action_id": "action_bend_1", "mentions": [ { "text": "bending", "ranges": [ [ 143, 150 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "interval": { "start_s": 21.0, "end_s": 22.0, "source_start_s": 73.57, "source_end_s": 74.57 }, "confidence": "high", "temporal_status": "supported" }, { "caption_index": 5, "loom_clip_id": "v_-YMpwZkNc2A_1576_2400", "action_id": "action_bend_1", "mentions": [ { "text": "bending", "ranges": [ [ 143, 150 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "interval": { "start_s": 25.0, "end_s": 26.0, "source_start_s": 77.57, "source_end_s": 78.57 }, "confidence": "high", "temporal_status": "supported" } ], "id": "action_group_016", "segment_count": 7, "caption_indices": [ 5 ] }, { "action": "leans", "types": [ "simple" ], "segments": [ { "caption_index": 7, "loom_clip_id": "v_-YMpwZkNc2A_2834_3224", "action_id": "action_leaning", "mentions": [ { "text": "leans", "ranges": [ [ 32, 37 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "interval": { "start_s": 2.0, "end_s": 4.0, "source_start_s": 96.57, "source_end_s": 98.57 }, "confidence": "high", "temporal_status": "localized" } ], "id": "action_group_017", "segment_count": 1, "caption_indices": [ 7 ] }, { "action": "rinse", "types": [ "simple" ], "segments": [ { "caption_index": 5, "loom_clip_id": "v_-YMpwZkNc2A_1576_2400", "action_id": "action_rinse_1", "mentions": [ { "text": "rinse", "ranges": [ [ 159, 164 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "interval": null, "confidence": "medium", "temporal_status": "not_directly_visible" }, { "caption_index": 7, "loom_clip_id": "v_-YMpwZkNc2A_2834_3224", "action_id": "action_rinsing", "mentions": [ { "text": "rinse", "ranges": [ [ 49, 54 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "interval": { "start_s": 4.0, "end_s": 5.0, "source_start_s": 98.57, "source_end_s": 99.57 }, "confidence": "medium", "temporal_status": "localized" } ], "id": "action_group_018", "segment_count": 2, "caption_indices": [ 5, 7 ] }, { "action": "wipe away", "types": [ "phrasal" ], "segments": [ { "caption_index": 7, "loom_clip_id": "v_-YMpwZkNc2A_2834_3224", "action_id": "action_wiping_cream", "mentions": [ { "text": "wipe away", "ranges": [ [ 83, 92 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "interval": { "start_s": 6.0, "end_s": 7.0, "source_start_s": 100.57, "source_end_s": 101.57 }, "confidence": "high", "temporal_status": "localized" } ], "id": "action_group_019", "segment_count": 1, "caption_indices": [ 7 ] }, { "action": "check", "types": [ "simple" ], "segments": [ { "caption_index": 7, "loom_clip_id": "v_-YMpwZkNc2A_2834_3224", "action_id": "action_checking_reflection", "mentions": [ { "text": "check", "ranges": [ [ 152, 157 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "interval": { "start_s": 7.0, "end_s": 12.0, "source_start_s": 101.57, "source_end_s": 106.57 }, "confidence": "medium", "temporal_status": "localized" } ], "id": "action_group_020", "segment_count": 1, "caption_indices": [ 7 ] }, { "action": "wiping", "types": [ "simple" ], "segments": [ { "caption_index": 7, "loom_clip_id": "v_-YMpwZkNc2A_2834_3224", "action_id": "action_wiping_face_again", "mentions": [ { "text": "wiping", "ranges": [ [ 182, 188 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "interval": { "start_s": 12.0, "end_s": 13.0, "source_start_s": 106.57, "source_end_s": 107.57 }, "confidence": "high", "temporal_status": "localized" } ], "id": "action_group_021", "segment_count": 1, "caption_indices": [ 7 ] }, { "action": "trim", "types": [ "simple" ], "segments": [ { "caption_index": 8, "loom_clip_id": "v_-YMpwZkNc2A_3224_3578", "action_id": "action_2", "mentions": [ { "text": "trim", "ranges": [ [ 89, 93 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "interval": { "start_s": 0.0, "end_s": 1.0, "source_start_s": 107.57, "source_end_s": 108.57 }, "confidence": "high", "temporal_status": "localized" }, { "caption_index": 8, "loom_clip_id": "v_-YMpwZkNc2A_3224_3578", "action_id": "action_2", "mentions": [ { "text": "trim", "ranges": [ [ 89, 93 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "interval": { "start_s": 6.0, "end_s": 9.0, "source_start_s": 113.57, "source_end_s": 116.57 }, "confidence": "high", "temporal_status": "localized" } ], "id": "action_group_022", "segment_count": 2, "caption_indices": [ 8 ] }, { "action": "uses", "types": [ "simple" ], "segments": [ { "caption_index": 8, "loom_clip_id": "v_-YMpwZkNc2A_3224_3578", "action_id": "action_1", "mentions": [ { "text": "uses", "ranges": [ [ 61, 65 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "interval": { "start_s": 0.0, "end_s": 1.0, "source_start_s": 107.57, "source_end_s": 108.57 }, "confidence": "high", "temporal_status": "localized" }, { "caption_index": 8, "loom_clip_id": "v_-YMpwZkNc2A_3224_3578", "action_id": "action_1", "mentions": [ { "text": "uses", "ranges": [ [ 61, 65 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "interval": { "start_s": 2.0, "end_s": 5.0, "source_start_s": 109.57, "source_end_s": 112.57 }, "confidence": "high", "temporal_status": "localized" }, { "caption_index": 8, "loom_clip_id": "v_-YMpwZkNc2A_3224_3578", "action_id": "action_1", "mentions": [ { "text": "uses", "ranges": [ [ 61, 65 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "interval": { "start_s": 6.0, "end_s": 9.0, "source_start_s": 113.57, "source_end_s": 116.57 }, "confidence": "high", "temporal_status": "localized" } ], "id": "action_group_023", "segment_count": 3, "caption_indices": [ 8 ] }, { "action": "shape", "types": [ "simple" ], "segments": [ { "caption_index": 8, "loom_clip_id": "v_-YMpwZkNc2A_3224_3578", "action_id": "action_3", "mentions": [ { "text": "shape", "ranges": [ [ 113, 118 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "interval": { "start_s": 2.0, "end_s": 5.0, "source_start_s": 109.57, "source_end_s": 112.57 }, "confidence": "high", "temporal_status": "localized" } ], "id": "action_group_024", "segment_count": 1, "caption_indices": [ 8 ] }, { "action": "lean", "types": [ "posture" ], "segments": [ { "caption_index": 8, "loom_clip_id": "v_-YMpwZkNc2A_3224_3578", "action_id": "action_4", "mentions": [ { "text": "lean", "ranges": [ [ 143, 147 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "interval": { "start_s": 10.0, "end_s": 11.0, "source_start_s": 117.57, "source_end_s": 118.57 }, "confidence": "high", "temporal_status": "localized" } ], "id": "action_group_025", "segment_count": 1, "caption_indices": [ 8 ] }, { "action": "move", "types": [ "simple" ], "segments": [ { "caption_index": 2, "loom_clip_id": "v_-YMpwZkNc2A_497_781", "action_id": "action_5", "mentions": [ { "text": "move", "ranges": [ [ 181, 185 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "interval": null, "confidence": "high", "temporal_status": "not_observed_in_samples" }, { "caption_index": 8, "loom_clip_id": "v_-YMpwZkNc2A_3224_3578", "action_id": "action_5", "mentions": [ { "text": "move", "ranges": [ [ 160, 164 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "interval": { "start_s": 11.0, "end_s": 11.820000000000007, "source_start_s": 118.57, "source_end_s": 119.39 }, "confidence": "high", "temporal_status": "localized" } ], "id": "action_group_026", "segment_count": 2, "caption_indices": [ 2, 8 ] }, { "action": "enters", "types": [ "simple" ], "segments": [ { "caption_index": 9, "loom_clip_id": "v_-YMpwZkNc2A_3634_3874", "action_id": "action_enter", "mentions": [ { "text": "enters", "ranges": [ [ 37, 43 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "interval": { "start_s": 0.0, "end_s": 2.0, "source_start_s": 121.25, "source_end_s": 123.25 }, "confidence": "medium", "temporal_status": "localized" } ], "id": "action_group_027", "segment_count": 1, "caption_indices": [ 9 ] }, { "action": "lifts", "types": [ "simple" ], "segments": [ { "caption_index": 9, "loom_clip_id": "v_-YMpwZkNc2A_3634_3874", "action_id": "action_lift_towel", "mentions": [ { "text": "lifts", "ranges": [ [ 73, 78 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "interval": { "start_s": 1.0, "end_s": 3.0, "source_start_s": 122.25, "source_end_s": 124.25 }, "confidence": "medium", "temporal_status": "localized" } ], "id": "action_group_028", "segment_count": 1, "caption_indices": [ 9 ] }, { "action": "dry", "types": [ "simple" ], "segments": [ { "caption_index": 9, "loom_clip_id": "v_-YMpwZkNc2A_3634_3874", "action_id": "action_dry_face", "mentions": [ { "text": "dry", "ranges": [ [ 123, 126 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "interval": { "start_s": 2.0, "end_s": 4.0, "source_start_s": 123.25, "source_end_s": 125.25 }, "confidence": "high", "temporal_status": "localized" }, { "caption_index": 9, "loom_clip_id": "v_-YMpwZkNc2A_3634_3874", "action_id": "action_dry_neck", "mentions": [ { "text": "dry", "ranges": [ [ 201, 204 ] ], "occurrence": 2, "particle": { "text": null, "occurrence": 1 } } ], "interval": { "start_s": 4.0, "end_s": 6.0, "source_start_s": 125.25, "source_end_s": 127.25 }, "confidence": "high", "temporal_status": "localized" } ], "id": "action_group_029", "segment_count": 2, "caption_indices": [ 9 ] }, { "action": "rubs", "types": [ "simple" ], "segments": [ { "caption_index": 9, "loom_clip_id": "v_-YMpwZkNc2A_3634_3874", "action_id": "action_rub_face", "mentions": [ { "text": "rubs", "ranges": [ [ 140, 144 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "interval": { "start_s": 2.0, "end_s": 4.0, "source_start_s": 123.25, "source_end_s": 125.25 }, "confidence": "high", "temporal_status": "localized" } ], "id": "action_group_030", "segment_count": 1, "caption_indices": [ 9 ] }, { "action": "moving", "types": [ "simple" ], "segments": [ { "caption_index": 9, "loom_clip_id": "v_-YMpwZkNc2A_3634_3874", "action_id": "action_move_towel_down", "mentions": [ { "text": "moving", "ranges": [ [ 183, 189 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "interval": { "start_s": 3.0, "end_s": 5.0, "source_start_s": 124.25, "source_end_s": 126.25 }, "confidence": "high", "temporal_status": "localized" } ], "id": "action_group_031", "segment_count": 1, "caption_indices": [ 9 ] }, { "action": "runs", "types": [ "prepositional" ], "segments": [ { "caption_index": 10, "loom_clip_id": "v_-YMpwZkNc2A_3874_4188", "action_id": "action_2", "mentions": [ { "text": "runs", "ranges": [ [ 41, 45 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "interval": { "start_s": 1.0, "end_s": 2.0, "source_start_s": 130.25, "source_end_s": 131.25 }, "confidence": "high", "temporal_status": "supported" } ], "id": "action_group_032", "segment_count": 1, "caption_indices": [ 10 ] }, { "action": "turns", "types": [ "simple" ], "segments": [ { "caption_index": 10, "loom_clip_id": "v_-YMpwZkNc2A_3874_4188", "action_id": "action_5", "mentions": [ { "text": "turns", "ranges": [ [ 111, 116 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "interval": { "start_s": 3.0, "end_s": 5.0, "source_start_s": 132.25, "source_end_s": 134.25 }, "confidence": "high", "temporal_status": "supported" }, { "caption_index": 12, "loom_clip_id": "v_-YMpwZkNc2A_4458_4717", "action_id": "action_5", "mentions": [ { "text": "turns", "ranges": [ [ 105, 110 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "interval": { "start_s": 8.0, "end_s": 8.639999999999986, "source_start_s": 156.75, "source_end_s": 157.39 }, "confidence": "medium", "temporal_status": "partially_observed_at_clip_boundary" } ], "id": "action_group_033", "segment_count": 2, "caption_indices": [ 10, 12 ] }, { "action": "touches", "types": [ "simple" ], "segments": [ { "caption_index": 10, "loom_clip_id": "v_-YMpwZkNc2A_3874_4188", "action_id": "action_7", "mentions": [ { "text": "touches", "ranges": [ [ 158, 165 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "interval": { "start_s": 4.0, "end_s": 5.0, "source_start_s": 133.25, "source_end_s": 134.25 }, "confidence": "high", "temporal_status": "supported" } ], "id": "action_group_034", "segment_count": 1, "caption_indices": [ 10 ] }, { "action": "talks · speak", "types": [ "simple" ], "segments": [ { "caption_index": 10, "loom_clip_id": "v_-YMpwZkNc2A_3874_4188", "action_id": "action_6", "mentions": [ { "text": "talks", "ranges": [ [ 140, 145 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } }, { "text": "speak", "ranges": [ [ 210, 215 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "interval": { "start_s": 6.0, "end_s": 10.5, "source_start_s": 135.25, "source_end_s": 139.75 }, "confidence": "high", "temporal_status": "supported" } ], "id": "action_group_035", "segment_count": 1, "caption_indices": [ 10 ] }, { "action": "talking", "types": [ "simple" ], "segments": [ { "caption_index": 11, "loom_clip_id": "v_-YMpwZkNc2A_4188_4458", "action_id": "action_3", "mentions": [ { "text": "talking", "ranges": [ [ 71, 78 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "interval": { "start_s": 0.0, "end_s": 4.0, "source_start_s": 139.75, "source_end_s": 143.75 }, "confidence": "medium", "temporal_status": "localized" }, { "caption_index": 11, "loom_clip_id": "v_-YMpwZkNc2A_4188_4458", "action_id": "action_3", "mentions": [ { "text": "talking", "ranges": [ [ 71, 78 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "interval": { "start_s": 5.0, "end_s": 8.0, "source_start_s": 144.75, "source_end_s": 147.75 }, "confidence": "medium", "temporal_status": "localized" } ], "id": "action_group_036", "segment_count": 2, "caption_indices": [ 11 ] }, { "action": "bringing", "types": [ "simple" ], "segments": [ { "caption_index": 11, "loom_clip_id": "v_-YMpwZkNc2A_4188_4458", "action_id": "action_5", "mentions": [ { "text": "bringing", "ranges": [ [ 114, 122 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "interval": { "start_s": 1.0, "end_s": 3.0, "source_start_s": 140.75, "source_end_s": 142.75 }, "confidence": "high", "temporal_status": "localized" }, { "caption_index": 11, "loom_clip_id": "v_-YMpwZkNc2A_4188_4458", "action_id": "action_5", "mentions": [ { "text": "bringing", "ranges": [ [ 114, 122 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "interval": { "start_s": 7.0, "end_s": 9.0, "source_start_s": 146.75, "source_end_s": 148.75 }, "confidence": "high", "temporal_status": "localized" } ], "id": "action_group_037", "segment_count": 2, "caption_indices": [ 11 ] }, { "action": "gestures", "types": [ "prepositional" ], "segments": [ { "caption_index": 11, "loom_clip_id": "v_-YMpwZkNc2A_4188_4458", "action_id": "action_4", "mentions": [ { "text": "gestures", "ranges": [ [ 83, 91 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "interval": { "start_s": 1.0, "end_s": 2.0, "source_start_s": 140.75, "source_end_s": 141.75 }, "confidence": "high", "temporal_status": "localized" }, { "caption_index": 11, "loom_clip_id": "v_-YMpwZkNc2A_4188_4458", "action_id": "action_4", "mentions": [ { "text": "gestures", "ranges": [ [ 83, 91 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "interval": { "start_s": 5.0, "end_s": 8.0, "source_start_s": 144.75, "source_end_s": 147.75 }, "confidence": "high", "temporal_status": "localized" } ], "id": "action_group_038", "segment_count": 2, "caption_indices": [ 11 ] }, { "action": "holds", "types": [ "phrasal" ], "segments": [ { "caption_index": 11, "loom_clip_id": "v_-YMpwZkNc2A_4188_4458", "action_id": "action_2", "mentions": [ { "text": "holds", "ranges": [ [ 53, 58 ] ], "occurrence": 1, "particle": { "text": "up", "occurrence": 2 } } ], "interval": { "start_s": 1.0, "end_s": 9.0, "source_start_s": 140.75, "source_end_s": 148.75 }, "confidence": "high", "temporal_status": "localized" } ], "id": "action_group_039", "segment_count": 1, "caption_indices": [ 11 ] }, { "action": "moves", "types": [ "simple" ], "segments": [ { "caption_index": 11, "loom_clip_id": "v_-YMpwZkNc2A_4188_4458", "action_id": "action_6", "mentions": [ { "text": "moves", "ranges": [ [ 151, 156 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "interval": { "start_s": 2.0, "end_s": 5.0, "source_start_s": 141.75, "source_end_s": 144.75 }, "confidence": "high", "temporal_status": "localized" }, { "caption_index": 11, "loom_clip_id": "v_-YMpwZkNc2A_4188_4458", "action_id": "action_6", "mentions": [ { "text": "moves", "ranges": [ [ 151, 156 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "interval": { "start_s": 7.0, "end_s": 9.0, "source_start_s": 146.75, "source_end_s": 148.75 }, "confidence": "high", "temporal_status": "localized" }, { "caption_index": 12, "loom_clip_id": "v_-YMpwZkNc2A_4458_4717", "action_id": "action_6", "mentions": [ { "text": "moves", "ranges": [ [ 137, 142 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "interval": { "start_s": 8.0, "end_s": 8.639999999999986, "source_start_s": 156.75, "source_end_s": 157.39 }, "confidence": "medium", "temporal_status": "partially_observed_at_clip_boundary" } ], "id": "action_group_040", "segment_count": 3, "caption_indices": [ 11, 12 ] }, { "action": "lowers", "types": [ "simple" ], "segments": [ { "caption_index": 10, "loom_clip_id": "v_-YMpwZkNc2A_3874_4188", "action_id": "action_1", "mentions": [ { "text": "lowers", "ranges": [ [ 8, 14 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "interval": null, "confidence": "high", "temporal_status": "unsupported_by_sampled_frames" }, { "caption_index": 12, "loom_clip_id": "v_-YMpwZkNc2A_4458_4717", "action_id": "action_1", "mentions": [ { "text": "lowers", "ranges": [ [ 3, 9 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "interval": { "start_s": 0.0, "end_s": 2.0, "source_start_s": 148.75, "source_end_s": 150.75 }, "confidence": "high", "temporal_status": "supported_by_sampled_frames" } ], "id": "action_group_041", "segment_count": 2, "caption_indices": [ 10, 12 ] }, { "action": "speaking", "types": [ "simple" ], "segments": [ { "caption_index": 12, "loom_clip_id": "v_-YMpwZkNc2A_4458_4717", "action_id": "action_4", "mentions": [ { "text": "speaking", "ranges": [ [ 71, 79 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "interval": { "start_s": 3.0, "end_s": 7.0, "source_start_s": 151.75, "source_end_s": 155.75 }, "confidence": "medium", "temporal_status": "supported_by_sampled_frames" } ], "id": "action_group_042", "segment_count": 1, "caption_indices": [ 12 ] }, { "action": "inspect", "types": [ "simple" ], "segments": [ { "caption_index": 6, "loom_clip_id": "v_-YMpwZkNc2A_2400_2834", "action_id": "action_2", "mentions": [ { "text": "inspect", "ranges": [ [ 146, 153 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "interval": null, "confidence": "high", "temporal_status": "uncertain" } ], "id": "action_group_043", "segment_count": 1, "caption_indices": [ 6 ] }, { "action": "stands", "types": [ "posture" ], "segments": [ { "caption_index": 10, "loom_clip_id": "v_-YMpwZkNc2A_3874_4188", "action_id": "action_3", "mentions": [ { "text": "stands", "ranges": [ [ 87, 93 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "interval": null, "confidence": "high", "temporal_status": "uncertain" } ], "id": "action_group_044", "segment_count": 1, "caption_indices": [ 10 ] } ] }, { "source_video_id": "v_-ceM4HHukhE", "grouping_rule": "Exact mention phrase set after case/whitespace normalization; action type is retained as metadata but does not split groups.", "actions": [ { "action": "walks", "types": [ "simple" ], "segments": [ { "caption_index": 1, "loom_clip_id": "v_-ceM4HHukhE_1_180", "action_id": "action_2", "mentions": [ { "text": "walks", "ranges": [ [ 38, 43 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "interval": { "start_s": 0.0, "end_s": 2.0, "source_start_s": 0.0, "source_end_s": 2.0 }, "confidence": "high", "temporal_status": "supported" } ], "id": "action_group_001", "segment_count": 1, "caption_indices": [ 1 ] }, { "action": "wearing", "types": [ "simple" ], "segments": [ { "caption_index": 1, "loom_clip_id": "v_-ceM4HHukhE_1_180", "action_id": "action_1", "mentions": [ { "text": "wearing", "ranges": [ [ 6, 13 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "interval": { "start_s": 0.0, "end_s": 6.0, "source_start_s": 0.0, "source_end_s": 6.0 }, "confidence": "high", "temporal_status": "supported" } ], "id": "action_group_002", "segment_count": 1, "caption_indices": [ 1 ] }, { "action": "sits down", "types": [ "posture" ], "segments": [ { "caption_index": 1, "loom_clip_id": "v_-ceM4HHukhE_1_180", "action_id": "action_3", "mentions": [ { "text": "sits down", "ranges": [ [ 68, 77 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "interval": { "start_s": 2.0, "end_s": 5.0, "source_start_s": 2.0, "source_end_s": 5.0 }, "confidence": "high", "temporal_status": "supported" } ], "id": "action_group_003", "segment_count": 1, "caption_indices": [ 1 ] }, { "action": "facing", "types": [ "posture" ], "segments": [ { "caption_index": 1, "loom_clip_id": "v_-ceM4HHukhE_1_180", "action_id": "action_4", "mentions": [ { "text": "facing", "ranges": [ [ 91, 97 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "interval": { "start_s": 4.0, "end_s": 6.0, "source_start_s": 4.0, "source_end_s": 6.0 }, "confidence": "high", "temporal_status": "supported" } ], "id": "action_group_004", "segment_count": 1, "caption_indices": [ 1 ] }, { "action": "sits", "types": [ "posture" ], "segments": [ { "caption_index": 2, "loom_clip_id": "v_-ceM4HHukhE_180_584", "action_id": "action_1", "mentions": [ { "text": "sits", "ranges": [ [ 3, 7 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "interval": { "start_s": 0.0, "end_s": 13.5, "source_start_s": 6.0, "source_end_s": 19.5 }, "confidence": "high", "temporal_status": "localized" } ], "id": "action_group_005", "segment_count": 1, "caption_indices": [ 2 ] }, { "action": "looks", "types": [ "simple" ], "segments": [ { "caption_index": 2, "loom_clip_id": "v_-ceM4HHukhE_180_584", "action_id": "action_4", "mentions": [ { "text": "looks", "ranges": [ [ 98, 103 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "interval": { "start_s": 3.0, "end_s": 4.0, "source_start_s": 9.0, "source_end_s": 10.0 }, "confidence": "medium", "temporal_status": "localized" }, { "caption_index": 2, "loom_clip_id": "v_-ceM4HHukhE_180_584", "action_id": "action_4", "mentions": [ { "text": "looks", "ranges": [ [ 98, 103 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "interval": { "start_s": 6.0, "end_s": 7.0, "source_start_s": 12.0, "source_end_s": 13.0 }, "confidence": "medium", "temporal_status": "localized" }, { "caption_index": 2, "loom_clip_id": "v_-ceM4HHukhE_180_584", "action_id": "action_4", "mentions": [ { "text": "looks", "ranges": [ [ 98, 103 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "interval": { "start_s": 9.0, "end_s": 11.0, "source_start_s": 15.0, "source_end_s": 17.0 }, "confidence": "medium", "temporal_status": "localized" } ], "id": "action_group_006", "segment_count": 3, "caption_indices": [ 2 ] }, { "action": "gesturing", "types": [ "simple" ], "segments": [ { "caption_index": 2, "loom_clip_id": "v_-ceM4HHukhE_180_584", "action_id": "action_3", "mentions": [ { "text": "gesturing", "ranges": [ [ 82, 91 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "interval": { "start_s": 8.0, "end_s": 13.5, "source_start_s": 14.0, "source_end_s": 19.5 }, "confidence": "high", "temporal_status": "localized" } ], "id": "action_group_007", "segment_count": 1, "caption_indices": [ 2 ] }, { "action": "leans", "types": [ "posture", "simple" ], "segments": [ { "caption_index": 3, "loom_clip_id": "v_-ceM4HHukhE_584_1064", "action_id": "action_1", "mentions": [ { "text": "leans", "ranges": [ [ 8, 13 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "interval": { "start_s": 1.0, "end_s": 9.0, "source_start_s": 20.5, "source_end_s": 28.5 }, "confidence": "high", "temporal_status": "localized" }, { "caption_index": 5, "loom_clip_id": "v_-ceM4HHukhE_2278_3641", "action_id": "action_1", "mentions": [ { "text": "leans", "ranges": [ [ 3, 8 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "interval": { "start_s": 0.0, "end_s": 6.0, "source_start_s": 76.0, "source_end_s": 82.0 }, "confidence": "high", "temporal_status": "supported" }, { "caption_index": 6, "loom_clip_id": "v_-ceM4HHukhE_3641_4540", "action_id": "action_3", "mentions": [ { "text": "leans", "ranges": [ [ 59, 64 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "interval": { "start_s": 14.0, "end_s": 18.0, "source_start_s": 135.5, "source_end_s": 139.5 }, "confidence": "high", "temporal_status": "localized" }, { "caption_index": 6, "loom_clip_id": "v_-ceM4HHukhE_3641_4540", "action_id": "action_3", "mentions": [ { "text": "leans", "ranges": [ [ 59, 64 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "interval": { "start_s": 20.0, "end_s": 23.0, "source_start_s": 141.5, "source_end_s": 144.5 }, "confidence": "high", "temporal_status": "localized" }, { "caption_index": 9, "loom_clip_id": "v_-ceM4HHukhE_6054_6319", "action_id": "action_3", "mentions": [ { "text": "leans", "ranges": [ [ 77, 82 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "interval": { "start_s": 7.0, "end_s": 8.840000000000003, "source_start_s": 209.0, "source_end_s": 210.84 }, "confidence": "high", "temporal_status": "sparse_localized" } ], "id": "action_group_008", "segment_count": 5, "caption_indices": [ 3, 5, 6, 9 ] }, { "action": "using", "types": [ "simple" ], "segments": [ { "caption_index": 3, "loom_clip_id": "v_-ceM4HHukhE_584_1064", "action_id": "action_3", "mentions": [ { "text": "using", "ranges": [ [ 61, 66 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "interval": { "start_s": 1.0, "end_s": 9.0, "source_start_s": 20.5, "source_end_s": 28.5 }, "confidence": "high", "temporal_status": "localized" } ], "id": "action_group_009", "segment_count": 1, "caption_indices": [ 3 ] }, { "action": "adjust", "types": [ "simple" ], "segments": [ { "caption_index": 3, "loom_clip_id": "v_-ceM4HHukhE_584_1064", "action_id": "action_2", "mentions": [ { "text": "adjust", "ranges": [ [ 25, 31 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "interval": { "start_s": 2.0, "end_s": 9.0, "source_start_s": 21.5, "source_end_s": 28.5 }, "confidence": "high", "temporal_status": "localized" }, { "caption_index": 3, "loom_clip_id": "v_-ceM4HHukhE_584_1064", "action_id": "action_2", "mentions": [ { "text": "adjust", "ranges": [ [ 25, 31 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "interval": { "start_s": 12.0, "end_s": 13.0, "source_start_s": 31.5, "source_end_s": 32.5 }, "confidence": "high", "temporal_status": "localized" } ], "id": "action_group_010", "segment_count": 2, "caption_indices": [ 3 ] }, { "action": "secure", "types": [ "simple" ], "segments": [ { "caption_index": 3, "loom_clip_id": "v_-ceM4HHukhE_584_1064", "action_id": "action_4", "mentions": [ { "text": "secure", "ranges": [ [ 81, 87 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "interval": { "start_s": 2.0, "end_s": 9.0, "source_start_s": 21.5, "source_end_s": 28.5 }, "confidence": "medium", "temporal_status": "localized" }, { "caption_index": 3, "loom_clip_id": "v_-ceM4HHukhE_584_1064", "action_id": "action_4", "mentions": [ { "text": "secure", "ranges": [ [ 81, 87 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "interval": { "start_s": 10.0, "end_s": 13.0, "source_start_s": 29.5, "source_end_s": 32.5 }, "confidence": "medium", "temporal_status": "localized" } ], "id": "action_group_011", "segment_count": 2, "caption_indices": [ 3 ] }, { "action": "Sitting", "types": [ "posture" ], "segments": [ { "caption_index": 4, "loom_clip_id": "v_-ceM4HHukhE_1064_2278", "action_id": "action_1", "mentions": [ { "text": "Sitting", "ranges": [ [ 0, 7 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "interval": { "start_s": 0.0, "end_s": 40.5, "source_start_s": 35.5, "source_end_s": 76.0 }, "confidence": "high", "temporal_status": "supported" } ], "id": "action_group_012", "segment_count": 1, "caption_indices": [ 4 ] }, { "action": "talk", "types": [ "simple" ], "segments": [ { "caption_index": 4, "loom_clip_id": "v_-ceM4HHukhE_1064_2278", "action_id": "action_2", "mentions": [ { "text": "talk", "ranges": [ [ 59, 63 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "interval": { "start_s": 0.0, "end_s": 27.0, "source_start_s": 35.5, "source_end_s": 62.5 }, "confidence": "medium", "temporal_status": "supported" }, { "caption_index": 4, "loom_clip_id": "v_-ceM4HHukhE_1064_2278", "action_id": "action_2", "mentions": [ { "text": "talk", "ranges": [ [ 59, 63 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "interval": { "start_s": 34.0, "end_s": 36.0, "source_start_s": 69.5, "source_end_s": 71.5 }, "confidence": "medium", "temporal_status": "supported" }, { "caption_index": 4, "loom_clip_id": "v_-ceM4HHukhE_1064_2278", "action_id": "action_2", "mentions": [ { "text": "talk", "ranges": [ [ 59, 63 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "interval": { "start_s": 39.0, "end_s": 40.0, "source_start_s": 74.5, "source_end_s": 75.5 }, "confidence": "medium", "temporal_status": "supported" } ], "id": "action_group_013", "segment_count": 3, "caption_indices": [ 4 ] }, { "action": "clasping", "types": [ "simple" ], "segments": [ { "caption_index": 4, "loom_clip_id": "v_-ceM4HHukhE_1064_2278", "action_id": "action_4", "mentions": [ { "text": "clasping", "ranges": [ [ 120, 128 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "interval": { "start_s": 2.0, "end_s": 10.0, "source_start_s": 37.5, "source_end_s": 45.5 }, "confidence": "high", "temporal_status": "supported" } ], "id": "action_group_014", "segment_count": 1, "caption_indices": [ 4 ] }, { "action": "gesture · pointing", "types": [ "simple" ], "segments": [ { "caption_index": 4, "loom_clip_id": "v_-ceM4HHukhE_1064_2278", "action_id": "action_3", "mentions": [ { "text": "gesture", "ranges": [ [ 68, 75 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } }, { "text": "pointing", "ranges": [ [ 77, 85 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "interval": { "start_s": 11.0, "end_s": 18.0, "source_start_s": 46.5, "source_end_s": 53.5 }, "confidence": "high", "temporal_status": "supported" }, { "caption_index": 4, "loom_clip_id": "v_-ceM4HHukhE_1064_2278", "action_id": "action_3", "mentions": [ { "text": "gesture", "ranges": [ [ 68, 75 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } }, { "text": "pointing", "ranges": [ [ 77, 85 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "interval": { "start_s": 22.0, "end_s": 25.0, "source_start_s": 57.5, "source_end_s": 60.5 }, "confidence": "high", "temporal_status": "supported" }, { "caption_index": 4, "loom_clip_id": "v_-ceM4HHukhE_1064_2278", "action_id": "action_3", "mentions": [ { "text": "gesture", "ranges": [ [ 68, 75 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } }, { "text": "pointing", "ranges": [ [ 77, 85 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "interval": { "start_s": 31.0, "end_s": 32.0, "source_start_s": 66.5, "source_end_s": 67.5 }, "confidence": "high", "temporal_status": "supported" }, { "caption_index": 4, "loom_clip_id": "v_-ceM4HHukhE_1064_2278", "action_id": "action_3", "mentions": [ { "text": "gesture", "ranges": [ [ 68, 75 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } }, { "text": "pointing", "ranges": [ [ 77, 85 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "interval": { "start_s": 35.0, "end_s": 36.0, "source_start_s": 70.5, "source_end_s": 71.5 }, "confidence": "high", "temporal_status": "supported" }, { "caption_index": 4, "loom_clip_id": "v_-ceM4HHukhE_1064_2278", "action_id": "action_3", "mentions": [ { "text": "gesture", "ranges": [ [ 68, 75 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } }, { "text": "pointing", "ranges": [ [ 77, 85 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "interval": { "start_s": 39.0, "end_s": 40.5, "source_start_s": 74.5, "source_end_s": 76.0 }, "confidence": "high", "temporal_status": "supported" } ], "id": "action_group_015", "segment_count": 5, "caption_indices": [ 4 ] }, { "action": "explaining", "types": [ "simple" ], "segments": [ { "caption_index": 5, "loom_clip_id": "v_-ceM4HHukhE_2278_3641", "action_id": "action_3", "mentions": [ { "text": "explaining", "ranges": [ [ 81, 91 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "interval": { "start_s": 0.0, "end_s": 42.0, "source_start_s": 76.0, "source_end_s": 118.0 }, "confidence": "medium", "temporal_status": "supported" } ], "id": "action_group_016", "segment_count": 1, "caption_indices": [ 5 ] }, { "action": "points", "types": [ "prepositional" ], "segments": [ { "caption_index": 5, "loom_clip_id": "v_-ceM4HHukhE_2278_3641", "action_id": "action_2", "mentions": [ { "text": "points", "ranges": [ [ 45, 51 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "interval": { "start_s": 1.0, "end_s": 5.0, "source_start_s": 77.0, "source_end_s": 81.0 }, "confidence": "high", "temporal_status": "supported" }, { "caption_index": 5, "loom_clip_id": "v_-ceM4HHukhE_2278_3641", "action_id": "action_2", "mentions": [ { "text": "points", "ranges": [ [ 45, 51 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "interval": { "start_s": 21.0, "end_s": 24.0, "source_start_s": 97.0, "source_end_s": 100.0 }, "confidence": "high", "temporal_status": "supported" }, { "caption_index": 5, "loom_clip_id": "v_-ceM4HHukhE_2278_3641", "action_id": "action_2", "mentions": [ { "text": "points", "ranges": [ [ 45, 51 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "interval": { "start_s": 32.0, "end_s": 35.0, "source_start_s": 108.0, "source_end_s": 111.0 }, "confidence": "high", "temporal_status": "supported" } ], "id": "action_group_017", "segment_count": 3, "caption_indices": [ 5 ] }, { "action": "straightens up", "types": [ "phrasal" ], "segments": [ { "caption_index": 5, "loom_clip_id": "v_-ceM4HHukhE_2278_3641", "action_id": "action_4", "mentions": [ { "text": "straightens up", "ranges": [ [ 111, 125 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "interval": { "start_s": 5.0, "end_s": 7.0, "source_start_s": 81.0, "source_end_s": 83.0 }, "confidence": "medium", "temporal_status": "supported" } ], "id": "action_group_018", "segment_count": 1, "caption_indices": [ 5 ] }, { "action": "demonstrate", "types": [ "simple" ], "segments": [ { "caption_index": 6, "loom_clip_id": "v_-ceM4HHukhE_3641_4540", "action_id": "action_4", "mentions": [ { "text": "demonstrate", "ranges": [ [ 86, 97 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "interval": { "start_s": 0.0, "end_s": 2.0, "source_start_s": 121.5, "source_end_s": 123.5 }, "confidence": "high", "temporal_status": "localized" }, { "caption_index": 6, "loom_clip_id": "v_-ceM4HHukhE_3641_4540", "action_id": "action_4", "mentions": [ { "text": "demonstrate", "ranges": [ [ 86, 97 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "interval": { "start_s": 8.0, "end_s": 10.0, "source_start_s": 129.5, "source_end_s": 131.5 }, "confidence": "high", "temporal_status": "localized" }, { "caption_index": 6, "loom_clip_id": "v_-ceM4HHukhE_3641_4540", "action_id": "action_4", "mentions": [ { "text": "demonstrate", "ranges": [ [ 86, 97 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "interval": { "start_s": 16.0, "end_s": 20.0, "source_start_s": 137.5, "source_end_s": 141.5 }, "confidence": "high", "temporal_status": "localized" }, { "caption_index": 6, "loom_clip_id": "v_-ceM4HHukhE_3641_4540", "action_id": "action_4", "mentions": [ { "text": "demonstrate", "ranges": [ [ 86, 97 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "interval": { "start_s": 22.0, "end_s": 30.0, "source_start_s": 143.5, "source_end_s": 151.5 }, "confidence": "high", "temporal_status": "localized" } ], "id": "action_group_019", "segment_count": 4, "caption_indices": [ 6 ] }, { "action": "extends", "types": [ "simple" ], "segments": [ { "caption_index": 6, "loom_clip_id": "v_-ceM4HHukhE_3641_4540", "action_id": "action_2", "mentions": [ { "text": "extends", "ranges": [ [ 37, 44 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "interval": { "start_s": 0.0, "end_s": 3.0, "source_start_s": 121.5, "source_end_s": 124.5 }, "confidence": "high", "temporal_status": "localized" } ], "id": "action_group_020", "segment_count": 1, "caption_indices": [ 6 ] }, { "action": "grabs", "types": [ "simple" ], "segments": [ { "caption_index": 6, "loom_clip_id": "v_-ceM4HHukhE_3641_4540", "action_id": "action_1", "mentions": [ { "text": "grabs", "ranges": [ [ 3, 8 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "interval": { "start_s": 0.0, "end_s": 2.0, "source_start_s": 121.5, "source_end_s": 123.5 }, "confidence": "high", "temporal_status": "localized" } ], "id": "action_group_021", "segment_count": 1, "caption_indices": [ 6 ] }, { "action": "demonstrates", "types": [ "simple" ], "segments": [ { "caption_index": 7, "loom_clip_id": "v_-ceM4HHukhE_4540_5245", "action_id": "action_2", "mentions": [ { "text": "demonstrates", "ranges": [ [ 29, 41 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "interval": { "start_s": 0.0, "end_s": 5.0, "source_start_s": 151.5, "source_end_s": 156.5 }, "confidence": "high", "temporal_status": "localized" } ], "id": "action_group_022", "segment_count": 1, "caption_indices": [ 7 ] }, { "action": "holding", "types": [ "simple" ], "segments": [ { "caption_index": 7, "loom_clip_id": "v_-ceM4HHukhE_4540_5245", "action_id": "action_1", "mentions": [ { "text": "holding", "ranges": [ [ 6, 13 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "interval": { "start_s": 0.0, "end_s": 23.5, "source_start_s": 151.5, "source_end_s": 175.0 }, "confidence": "high", "temporal_status": "localized" } ], "id": "action_group_023", "segment_count": 1, "caption_indices": [ 7 ] }, { "action": "keeping", "types": [ "posture" ], "segments": [ { "caption_index": 7, "loom_clip_id": "v_-ceM4HHukhE_4540_5245", "action_id": "action_5", "mentions": [ { "text": "keeping", "ranges": [ [ 130, 137 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "interval": { "start_s": 0.0, "end_s": 5.0, "source_start_s": 151.5, "source_end_s": 156.5 }, "confidence": "high", "temporal_status": "localized" } ], "id": "action_group_024", "segment_count": 1, "caption_indices": [ 7 ] }, { "action": "pushing", "types": [ "simple" ], "segments": [ { "caption_index": 7, "loom_clip_id": "v_-ceM4HHukhE_4540_5245", "action_id": "action_3", "mentions": [ { "text": "pushing", "ranges": [ [ 84, 91 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "interval": { "start_s": 0.0, "end_s": 5.0, "source_start_s": 151.5, "source_end_s": 156.5 }, "confidence": "high", "temporal_status": "localized" }, { "caption_index": 8, "loom_clip_id": "v_-ceM4HHukhE_5245_6054", "action_id": "action_push_1", "mentions": [ { "text": "pushing", "ranges": [ [ 60, 67 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "interval": { "start_s": 3.0, "end_s": 7.0, "source_start_s": 178.0, "source_end_s": 182.0 }, "confidence": "high", "temporal_status": "localized_sparse" }, { "caption_index": 8, "loom_clip_id": "v_-ceM4HHukhE_5245_6054", "action_id": "action_push_1", "mentions": [ { "text": "pushing", "ranges": [ [ 60, 67 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "interval": { "start_s": 9.0, "end_s": 11.0, "source_start_s": 184.0, "source_end_s": 186.0 }, "confidence": "high", "temporal_status": "localized_sparse" }, { "caption_index": 8, "loom_clip_id": "v_-ceM4HHukhE_5245_6054", "action_id": "action_push_1", "mentions": [ { "text": "pushing", "ranges": [ [ 60, 67 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "interval": { "start_s": 11.0, "end_s": 13.0, "source_start_s": 186.0, "source_end_s": 188.0 }, "confidence": "high", "temporal_status": "localized_sparse" }, { "caption_index": 8, "loom_clip_id": "v_-ceM4HHukhE_5245_6054", "action_id": "action_push_1", "mentions": [ { "text": "pushing", "ranges": [ [ 60, 67 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "interval": { "start_s": 13.0, "end_s": 15.0, "source_start_s": 188.0, "source_end_s": 190.0 }, "confidence": "high", "temporal_status": "localized_sparse" }, { "caption_index": 8, "loom_clip_id": "v_-ceM4HHukhE_5245_6054", "action_id": "action_push_1", "mentions": [ { "text": "pushing", "ranges": [ [ 60, 67 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "interval": { "start_s": 15.0, "end_s": 17.0, "source_start_s": 190.0, "source_end_s": 192.0 }, "confidence": "high", "temporal_status": "localized_sparse" }, { "caption_index": 8, "loom_clip_id": "v_-ceM4HHukhE_5245_6054", "action_id": "action_push_1", "mentions": [ { "text": "pushing", "ranges": [ [ 60, 67 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "interval": { "start_s": 17.0, "end_s": 19.0, "source_start_s": 192.0, "source_end_s": 194.0 }, "confidence": "high", "temporal_status": "localized_sparse" }, { "caption_index": 8, "loom_clip_id": "v_-ceM4HHukhE_5245_6054", "action_id": "action_push_1", "mentions": [ { "text": "pushing", "ranges": [ [ 60, 67 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "interval": { "start_s": 19.0, "end_s": 21.0, "source_start_s": 194.0, "source_end_s": 196.0 }, "confidence": "high", "temporal_status": "localized_sparse" }, { "caption_index": 8, "loom_clip_id": "v_-ceM4HHukhE_5245_6054", "action_id": "action_push_1", "mentions": [ { "text": "pushing", "ranges": [ [ 60, 67 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "interval": { "start_s": 22.0, "end_s": 24.0, "source_start_s": 197.0, "source_end_s": 199.0 }, "confidence": "high", "temporal_status": "localized_sparse" }, { "caption_index": 8, "loom_clip_id": "v_-ceM4HHukhE_5245_6054", "action_id": "action_push_1", "mentions": [ { "text": "pushing", "ranges": [ [ 60, 67 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "interval": { "start_s": 24.0, "end_s": 27.0, "source_start_s": 199.0, "source_end_s": 202.0 }, "confidence": "high", "temporal_status": "localized_sparse" } ], "id": "action_group_025", "segment_count": 10, "caption_indices": [ 7, 8 ] }, { "action": "slide", "types": [ "simple" ], "segments": [ { "caption_index": 7, "loom_clip_id": "v_-ceM4HHukhE_4540_5245", "action_id": "action_4", "mentions": [ { "text": "slide", "ranges": [ [ 109, 114 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "interval": { "start_s": 0.0, "end_s": 5.0, "source_start_s": 151.5, "source_end_s": 156.5 }, "confidence": "high", "temporal_status": "localized" } ], "id": "action_group_026", "segment_count": 1, "caption_indices": [ 7 ] }, { "action": "performs several complete and continuous rowing strokes", "types": [ "light_verb" ], "segments": [ { "caption_index": 8, "loom_clip_id": "v_-ceM4HHukhE_5245_6054", "action_id": "action_rowing_1", "mentions": [ { "text": "performs several complete and continuous rowing strokes", "ranges": [ [ 3, 58 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "interval": { "start_s": 0.0, "end_s": 27.0, "source_start_s": 175.0, "source_end_s": 202.0 }, "confidence": "high", "temporal_status": "localized_sparse" } ], "id": "action_group_027", "segment_count": 1, "caption_indices": [ 8 ] }, { "action": "pulling", "types": [ "simple" ], "segments": [ { "caption_index": 8, "loom_clip_id": "v_-ceM4HHukhE_5245_6054", "action_id": "action_pull_1", "mentions": [ { "text": "pulling", "ranges": [ [ 101, 108 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "interval": { "start_s": 4.0, "end_s": 7.0, "source_start_s": 179.0, "source_end_s": 182.0 }, "confidence": "high", "temporal_status": "localized_sparse" }, { "caption_index": 8, "loom_clip_id": "v_-ceM4HHukhE_5245_6054", "action_id": "action_pull_1", "mentions": [ { "text": "pulling", "ranges": [ [ 101, 108 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "interval": { "start_s": 9.0, "end_s": 11.0, "source_start_s": 184.0, "source_end_s": 186.0 }, "confidence": "high", "temporal_status": "localized_sparse" }, { "caption_index": 8, "loom_clip_id": "v_-ceM4HHukhE_5245_6054", "action_id": "action_pull_1", "mentions": [ { "text": "pulling", "ranges": [ [ 101, 108 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "interval": { "start_s": 11.0, "end_s": 13.0, "source_start_s": 186.0, "source_end_s": 188.0 }, "confidence": "high", "temporal_status": "localized_sparse" }, { "caption_index": 8, "loom_clip_id": "v_-ceM4HHukhE_5245_6054", "action_id": "action_pull_1", "mentions": [ { "text": "pulling", "ranges": [ [ 101, 108 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "interval": { "start_s": 13.0, "end_s": 15.0, "source_start_s": 188.0, "source_end_s": 190.0 }, "confidence": "high", "temporal_status": "localized_sparse" }, { "caption_index": 8, "loom_clip_id": "v_-ceM4HHukhE_5245_6054", "action_id": "action_pull_1", "mentions": [ { "text": "pulling", "ranges": [ [ 101, 108 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "interval": { "start_s": 15.0, "end_s": 17.0, "source_start_s": 190.0, "source_end_s": 192.0 }, "confidence": "high", "temporal_status": "localized_sparse" }, { "caption_index": 8, "loom_clip_id": "v_-ceM4HHukhE_5245_6054", "action_id": "action_pull_1", "mentions": [ { "text": "pulling", "ranges": [ [ 101, 108 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "interval": { "start_s": 17.0, "end_s": 19.0, "source_start_s": 192.0, "source_end_s": 194.0 }, "confidence": "high", "temporal_status": "localized_sparse" }, { "caption_index": 8, "loom_clip_id": "v_-ceM4HHukhE_5245_6054", "action_id": "action_pull_1", "mentions": [ { "text": "pulling", "ranges": [ [ 101, 108 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "interval": { "start_s": 19.0, "end_s": 21.0, "source_start_s": 194.0, "source_end_s": 196.0 }, "confidence": "high", "temporal_status": "localized_sparse" }, { "caption_index": 8, "loom_clip_id": "v_-ceM4HHukhE_5245_6054", "action_id": "action_pull_1", "mentions": [ { "text": "pulling", "ranges": [ [ 101, 108 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "interval": { "start_s": 22.0, "end_s": 24.0, "source_start_s": 197.0, "source_end_s": 199.0 }, "confidence": "high", "temporal_status": "localized_sparse" }, { "caption_index": 8, "loom_clip_id": "v_-ceM4HHukhE_5245_6054", "action_id": "action_pull_1", "mentions": [ { "text": "pulling", "ranges": [ [ 101, 108 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "interval": { "start_s": 24.0, "end_s": 27.0, "source_start_s": 199.0, "source_end_s": 202.0 }, "confidence": "high", "temporal_status": "localized_sparse" } ], "id": "action_group_028", "segment_count": 9, "caption_indices": [ 8 ] }, { "action": "leaning", "types": [ "simple" ], "segments": [ { "caption_index": 8, "loom_clip_id": "v_-ceM4HHukhE_5245_6054", "action_id": "action_lean_1", "mentions": [ { "text": "leaning", "ranges": [ [ 83, 90 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "interval": { "start_s": 5.0, "end_s": 7.0, "source_start_s": 180.0, "source_end_s": 182.0 }, "confidence": "high", "temporal_status": "localized_sparse" }, { "caption_index": 8, "loom_clip_id": "v_-ceM4HHukhE_5245_6054", "action_id": "action_lean_1", "mentions": [ { "text": "leaning", "ranges": [ [ 83, 90 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "interval": { "start_s": 10.0, "end_s": 11.0, "source_start_s": 185.0, "source_end_s": 186.0 }, "confidence": "high", "temporal_status": "localized_sparse" }, { "caption_index": 8, "loom_clip_id": "v_-ceM4HHukhE_5245_6054", "action_id": "action_lean_1", "mentions": [ { "text": "leaning", "ranges": [ [ 83, 90 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "interval": { "start_s": 12.0, "end_s": 13.0, "source_start_s": 187.0, "source_end_s": 188.0 }, "confidence": "high", "temporal_status": "localized_sparse" }, { "caption_index": 8, "loom_clip_id": "v_-ceM4HHukhE_5245_6054", "action_id": "action_lean_1", "mentions": [ { "text": "leaning", "ranges": [ [ 83, 90 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "interval": { "start_s": 14.0, "end_s": 15.0, "source_start_s": 189.0, "source_end_s": 190.0 }, "confidence": "high", "temporal_status": "localized_sparse" }, { "caption_index": 8, "loom_clip_id": "v_-ceM4HHukhE_5245_6054", "action_id": "action_lean_1", "mentions": [ { "text": "leaning", "ranges": [ [ 83, 90 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "interval": { "start_s": 16.0, "end_s": 17.0, "source_start_s": 191.0, "source_end_s": 192.0 }, "confidence": "high", "temporal_status": "localized_sparse" }, { "caption_index": 8, "loom_clip_id": "v_-ceM4HHukhE_5245_6054", "action_id": "action_lean_1", "mentions": [ { "text": "leaning", "ranges": [ [ 83, 90 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "interval": { "start_s": 18.0, "end_s": 19.0, "source_start_s": 193.0, "source_end_s": 194.0 }, "confidence": "high", "temporal_status": "localized_sparse" }, { "caption_index": 8, "loom_clip_id": "v_-ceM4HHukhE_5245_6054", "action_id": "action_lean_1", "mentions": [ { "text": "leaning", "ranges": [ [ 83, 90 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "interval": { "start_s": 20.0, "end_s": 21.0, "source_start_s": 195.0, "source_end_s": 196.0 }, "confidence": "high", "temporal_status": "localized_sparse" }, { "caption_index": 8, "loom_clip_id": "v_-ceM4HHukhE_5245_6054", "action_id": "action_lean_1", "mentions": [ { "text": "leaning", "ranges": [ [ 83, 90 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "interval": { "start_s": 23.0, "end_s": 24.0, "source_start_s": 198.0, "source_end_s": 199.0 }, "confidence": "high", "temporal_status": "localized_sparse" }, { "caption_index": 8, "loom_clip_id": "v_-ceM4HHukhE_5245_6054", "action_id": "action_lean_1", "mentions": [ { "text": "leaning", "ranges": [ [ 83, 90 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "interval": { "start_s": 25.0, "end_s": 27.0, "source_start_s": 200.0, "source_end_s": 202.0 }, "confidence": "high", "temporal_status": "localized_sparse" } ], "id": "action_group_029", "segment_count": 9, "caption_indices": [ 8 ] }, { "action": "returning", "types": [ "simple" ], "segments": [ { "caption_index": 8, "loom_clip_id": "v_-ceM4HHukhE_5245_6054", "action_id": "action_return_1", "mentions": [ { "text": "returning", "ranges": [ [ 140, 149 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "interval": { "start_s": 7.0, "end_s": 10.0, "source_start_s": 182.0, "source_end_s": 185.0 }, "confidence": "high", "temporal_status": "localized_sparse" }, { "caption_index": 8, "loom_clip_id": "v_-ceM4HHukhE_5245_6054", "action_id": "action_return_1", "mentions": [ { "text": "returning", "ranges": [ [ 140, 149 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "interval": { "start_s": 11.0, "end_s": 12.0, "source_start_s": 186.0, "source_end_s": 187.0 }, "confidence": "high", "temporal_status": "localized_sparse" }, { "caption_index": 8, "loom_clip_id": "v_-ceM4HHukhE_5245_6054", "action_id": "action_return_1", "mentions": [ { "text": "returning", "ranges": [ [ 140, 149 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "interval": { "start_s": 13.0, "end_s": 14.0, "source_start_s": 188.0, "source_end_s": 189.0 }, "confidence": "high", "temporal_status": "localized_sparse" }, { "caption_index": 8, "loom_clip_id": "v_-ceM4HHukhE_5245_6054", "action_id": "action_return_1", "mentions": [ { "text": "returning", "ranges": [ [ 140, 149 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "interval": { "start_s": 15.0, "end_s": 16.0, "source_start_s": 190.0, "source_end_s": 191.0 }, "confidence": "high", "temporal_status": "localized_sparse" }, { "caption_index": 8, "loom_clip_id": "v_-ceM4HHukhE_5245_6054", "action_id": "action_return_1", "mentions": [ { "text": "returning", "ranges": [ [ 140, 149 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "interval": { "start_s": 17.0, "end_s": 18.0, "source_start_s": 192.0, "source_end_s": 193.0 }, "confidence": "high", "temporal_status": "localized_sparse" }, { "caption_index": 8, "loom_clip_id": "v_-ceM4HHukhE_5245_6054", "action_id": "action_return_1", "mentions": [ { "text": "returning", "ranges": [ [ 140, 149 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "interval": { "start_s": 19.0, "end_s": 20.0, "source_start_s": 194.0, "source_end_s": 195.0 }, "confidence": "high", "temporal_status": "localized_sparse" }, { "caption_index": 8, "loom_clip_id": "v_-ceM4HHukhE_5245_6054", "action_id": "action_return_1", "mentions": [ { "text": "returning", "ranges": [ [ 140, 149 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "interval": { "start_s": 21.0, "end_s": 23.0, "source_start_s": 196.0, "source_end_s": 198.0 }, "confidence": "high", "temporal_status": "localized_sparse" }, { "caption_index": 8, "loom_clip_id": "v_-ceM4HHukhE_5245_6054", "action_id": "action_return_1", "mentions": [ { "text": "returning", "ranges": [ [ 140, 149 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "interval": { "start_s": 24.0, "end_s": 25.0, "source_start_s": 199.0, "source_end_s": 200.0 }, "confidence": "high", "temporal_status": "localized_sparse" } ], "id": "action_group_030", "segment_count": 8, "caption_indices": [ 8 ] }, { "action": "completing a final stroke", "types": [ "light_verb" ], "segments": [ { "caption_index": 9, "loom_clip_id": "v_-ceM4HHukhE_6054_6319", "action_id": "action_1", "mentions": [ { "text": "completing a final stroke", "ranges": [ [ 6, 31 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "interval": { "start_s": 0.0, "end_s": 2.0, "source_start_s": 202.0, "source_end_s": 204.0 }, "confidence": "high", "temporal_status": "sparse_localized" } ], "id": "action_group_031", "segment_count": 1, "caption_indices": [ 9 ] }, { "action": "places", "types": [ "simple" ], "segments": [ { "caption_index": 9, "loom_clip_id": "v_-ceM4HHukhE_6054_6319", "action_id": "action_2", "mentions": [ { "text": "places", "ranges": [ [ 36, 42 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "interval": { "start_s": 6.0, "end_s": 8.0, "source_start_s": 208.0, "source_end_s": 210.0 }, "confidence": "medium", "temporal_status": "sparse_localized" } ], "id": "action_group_032", "segment_count": 1, "caption_indices": [ 9 ] }, { "action": "unstrap", "types": [ "simple" ], "segments": [ { "caption_index": 9, "loom_clip_id": "v_-ceM4HHukhE_6054_6319", "action_id": "action_4", "mentions": [ { "text": "unstrap", "ranges": [ [ 94, 101 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "interval": { "start_s": 8.0, "end_s": 8.840000000000003, "source_start_s": 210.0, "source_end_s": 210.84 }, "confidence": "medium", "temporal_status": "sparse_localized" } ], "id": "action_group_033", "segment_count": 1, "caption_indices": [ 9 ] }, { "action": "talking", "types": [ "simple" ], "segments": [ { "caption_index": 2, "loom_clip_id": "v_-ceM4HHukhE_180_584", "action_id": "action_2", "mentions": [ { "text": "talking", "ranges": [ [ 70, 77 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "interval": null, "confidence": "medium", "temporal_status": "uncertain" } ], "id": "action_group_034", "segment_count": 1, "caption_indices": [ 2 ] } ] }, { "source_video_id": "v_-ypKYhjrlXw", "grouping_rule": "Exact mention phrase set after case/whitespace normalization; action type is retained as metadata but does not split groups.", "actions": [ { "action": "sits", "types": [ "posture" ], "segments": [ { "caption_index": 1, "loom_clip_id": "v_-ypKYhjrlXw_1_1064", "action_id": "action_1", "mentions": [ { "text": "sits", "ranges": [ [ 38, 42 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "interval": { "start_s": 0.0, "end_s": 35.5, "source_start_s": 0.0, "source_end_s": 35.5 }, "confidence": "high", "temporal_status": "localized" } ], "id": "action_group_001", "segment_count": 1, "caption_indices": [ 1 ] }, { "action": "speaks", "types": [ "simple" ], "segments": [ { "caption_index": 1, "loom_clip_id": "v_-ypKYhjrlXw_1_1064", "action_id": "action_2", "mentions": [ { "text": "speaks", "ranges": [ [ 58, 64 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "interval": { "start_s": 1.0, "end_s": 35.5, "source_start_s": 1.0, "source_end_s": 35.5 }, "confidence": "high", "temporal_status": "localized" } ], "id": "action_group_002", "segment_count": 1, "caption_indices": [ 1 ] }, { "action": "gesturing", "types": [ "simple" ], "segments": [ { "caption_index": 1, "loom_clip_id": "v_-ypKYhjrlXw_1_1064", "action_id": "action_3", "mentions": [ { "text": "gesturing", "ranges": [ [ 79, 88 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "interval": { "start_s": 29.0, "end_s": 30.0, "source_start_s": 29.0, "source_end_s": 30.0 }, "confidence": "high", "temporal_status": "localized" }, { "caption_index": 3, "loom_clip_id": "v_-ypKYhjrlXw_1454_1769", "action_id": "action_3", "mentions": [ { "text": "gesturing", "ranges": [ [ 84, 93 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "interval": { "start_s": 2.0, "end_s": 3.0, "source_start_s": 50.5, "source_end_s": 51.5 }, "confidence": "high", "temporal_status": "supported" }, { "caption_index": 3, "loom_clip_id": "v_-ypKYhjrlXw_1454_1769", "action_id": "action_3", "mentions": [ { "text": "gesturing", "ranges": [ [ 84, 93 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "interval": { "start_s": 4.0, "end_s": 5.0, "source_start_s": 52.5, "source_end_s": 53.5 }, "confidence": "high", "temporal_status": "supported" }, { "caption_index": 3, "loom_clip_id": "v_-ypKYhjrlXw_1454_1769", "action_id": "action_3", "mentions": [ { "text": "gesturing", "ranges": [ [ 84, 93 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "interval": { "start_s": 6.0, "end_s": 10.5, "source_start_s": 54.5, "source_end_s": 59.0 }, "confidence": "high", "temporal_status": "supported" } ], "id": "action_group_003", "segment_count": 4, "caption_indices": [ 1, 3 ] }, { "action": "gestures · pointing", "types": [ "prepositional" ], "segments": [ { "caption_index": 2, "loom_clip_id": "v_-ypKYhjrlXw_1064_1454", "action_id": "action_1", "mentions": [ { "text": "gestures", "ranges": [ [ 8, 16 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } }, { "text": "pointing", "ranges": [ [ 60, 68 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "interval": { "start_s": 0.0, "end_s": 11.0, "source_start_s": 35.5, "source_end_s": 46.5 }, "confidence": "high", "temporal_status": "localized" } ], "id": "action_group_004", "segment_count": 1, "caption_indices": [ 2 ] }, { "action": "picks up", "types": [ "phrasal" ], "segments": [ { "caption_index": 2, "loom_clip_id": "v_-ypKYhjrlXw_1064_1454", "action_id": "action_2", "mentions": [ { "text": "picks up", "ranges": [ [ 99, 107 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "interval": { "start_s": 11.0, "end_s": 12.0, "source_start_s": 46.5, "source_end_s": 47.5 }, "confidence": "high", "temporal_status": "localized" }, { "caption_index": 3, "loom_clip_id": "v_-ypKYhjrlXw_1454_1769", "action_id": "action_1", "mentions": [ { "text": "picks up", "ranges": [ [ 8, 16 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "interval": { "start_s": 1.0, "end_s": 3.0, "source_start_s": 49.5, "source_end_s": 51.5 }, "confidence": "high", "temporal_status": "supported" }, { "caption_index": 5, "loom_clip_id": "v_-ypKYhjrlXw_2488_2878", "action_id": "action_pick_up", "mentions": [ { "text": "picks up", "ranges": [ [ 3, 11 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "interval": null, "confidence": "high", "temporal_status": "not_observed_in_sampled_frames" }, { "caption_index": 6, "loom_clip_id": "v_-ypKYhjrlXw_2878_3387", "action_id": "action_1", "mentions": [ { "text": "picks up", "ranges": [ [ 8, 16 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "interval": { "start_s": 0.0, "end_s": 2.0, "source_start_s": 96.0, "source_end_s": 98.0 }, "confidence": "high", "temporal_status": "localized_sparse" }, { "caption_index": 6, "loom_clip_id": "v_-ypKYhjrlXw_2878_3387", "action_id": "action_3", "mentions": [ { "text": "picks up", "ranges": [ [ 84, 92 ] ], "occurrence": 2, "particle": { "text": null, "occurrence": 1 } } ], "interval": { "start_s": 3.0, "end_s": 5.0, "source_start_s": 99.0, "source_end_s": 101.0 }, "confidence": "medium", "temporal_status": "localized_sparse" }, { "caption_index": 8, "loom_clip_id": "v_-ypKYhjrlXw_4211_5306", "action_id": "A1", "mentions": [ { "text": "picks up", "ranges": [ [ 8, 16 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "interval": { "start_s": 11.0, "end_s": 12.0, "source_start_s": 151.5, "source_end_s": 152.5 }, "confidence": "medium", "temporal_status": "supported_sparse" }, { "caption_index": 9, "loom_clip_id": "v_-ypKYhjrlXw_5306_5650", "action_id": "action_1", "mentions": [ { "text": "picks up", "ranges": [ [ 8, 16 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "interval": null, "confidence": "high", "temporal_status": "uncertain" } ], "id": "action_group_005", "segment_count": 7, "caption_indices": [ 2, 3, 5, 6, 8, 9 ] }, { "action": "show", "types": [ "simple" ], "segments": [ { "caption_index": 2, "loom_clip_id": "v_-ypKYhjrlXw_1064_1454", "action_id": "action_4", "mentions": [ { "text": "show", "ranges": [ [ 137, 141 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "interval": { "start_s": 12.0, "end_s": 13.0, "source_start_s": 47.5, "source_end_s": 48.5 }, "confidence": "high", "temporal_status": "localized" } ], "id": "action_group_006", "segment_count": 1, "caption_indices": [ 2 ] }, { "action": "speaking · explaining", "types": [ "simple" ], "segments": [ { "caption_index": 3, "loom_clip_id": "v_-ypKYhjrlXw_1454_1769", "action_id": "action_4", "mentions": [ { "text": "speaking", "ranges": [ [ 128, 136 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } }, { "text": "explaining", "ranges": [ [ 138, 148 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "interval": { "start_s": 0.0, "end_s": 10.5, "source_start_s": 48.5, "source_end_s": 59.0 }, "confidence": "medium", "temporal_status": "supported" } ], "id": "action_group_007", "segment_count": 1, "caption_indices": [ 3 ] }, { "action": "holds · holding", "types": [ "phrasal" ], "segments": [ { "caption_index": 3, "loom_clip_id": "v_-ypKYhjrlXw_1454_1769", "action_id": "action_2", "mentions": [ { "text": "holds", "ranges": [ [ 66, 71 ] ], "occurrence": 1, "particle": { "text": "up", "occurrence": 2 } }, { "text": "holding", "ranges": [ [ 167, 174 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "interval": { "start_s": 2.0, "end_s": 10.5, "source_start_s": 50.5, "source_end_s": 59.0 }, "confidence": "high", "temporal_status": "supported" } ], "id": "action_group_008", "segment_count": 1, "caption_indices": [ 3 ] }, { "action": "leans", "types": [ "posture" ], "segments": [ { "caption_index": 4, "loom_clip_id": "v_-ypKYhjrlXw_1769_2488", "action_id": "action_lean", "mentions": [ { "text": "leans", "ranges": [ [ 8, 13 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "interval": { "start_s": 1.0, "end_s": 4.0, "source_start_s": 60.0, "source_end_s": 63.0 }, "confidence": "high", "temporal_status": "localized_sparse" } ], "id": "action_group_009", "segment_count": 1, "caption_indices": [ 4 ] }, { "action": "sprays", "types": [ "simple" ], "segments": [ { "caption_index": 4, "loom_clip_id": "v_-ypKYhjrlXw_1769_2488", "action_id": "action_spray", "mentions": [ { "text": "sprays", "ranges": [ [ 26, 32 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "interval": { "start_s": 1.0, "end_s": 4.0, "source_start_s": 60.0, "source_end_s": 63.0 }, "confidence": "high", "temporal_status": "localized_sparse" } ], "id": "action_group_010", "segment_count": 1, "caption_indices": [ 4 ] }, { "action": "spread", "types": [ "simple" ], "segments": [ { "caption_index": 4, "loom_clip_id": "v_-ypKYhjrlXw_1769_2488", "action_id": "action_spread", "mentions": [ { "text": "spread", "ranges": [ [ 132, 138 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "interval": { "start_s": 4.0, "end_s": 12.0, "source_start_s": 63.0, "source_end_s": 71.0 }, "confidence": "high", "temporal_status": "localized_sparse" }, { "caption_index": 5, "loom_clip_id": "v_-ypKYhjrlXw_2488_2878", "action_id": "action_spread", "mentions": [ { "text": "spread", "ranges": [ [ 123, 129 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "interval": { "start_s": 2.0, "end_s": 10.0, "source_start_s": 85.0, "source_end_s": 93.0 }, "confidence": "high", "temporal_status": "localized" }, { "caption_index": 7, "loom_clip_id": "v_-ypKYhjrlXw_3387_4211", "action_id": "A4", "mentions": [ { "text": "spread", "ranges": [ [ 159, 165 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "interval": { "start_s": 25.0, "end_s": 27.5, "source_start_s": 138.0, "source_end_s": 140.5 }, "confidence": "high", "temporal_status": "localized_sparse" }, { "caption_index": 8, "loom_clip_id": "v_-ypKYhjrlXw_4211_5306", "action_id": "A3", "mentions": [ { "text": "spread", "ranges": [ [ 127, 133 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "interval": { "start_s": 25.0, "end_s": 36.5, "source_start_s": 165.5, "source_end_s": 177.0 }, "confidence": "high", "temporal_status": "supported_sparse" } ], "id": "action_group_011", "segment_count": 4, "caption_indices": [ 4, 5, 7, 8 ] }, { "action": "uses", "types": [ "simple" ], "segments": [ { "caption_index": 4, "loom_clip_id": "v_-ypKYhjrlXw_1769_2488", "action_id": "action_use_finger", "mentions": [ { "text": "uses", "ranges": [ [ 101, 105 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "interval": { "start_s": 4.0, "end_s": 12.0, "source_start_s": 63.0, "source_end_s": 71.0 }, "confidence": "high", "temporal_status": "localized_sparse" }, { "caption_index": 5, "loom_clip_id": "v_-ypKYhjrlXw_2488_2878", "action_id": "action_use_finger", "mentions": [ { "text": "uses", "ranges": [ [ 98, 102 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "interval": { "start_s": 2.0, "end_s": 10.0, "source_start_s": 85.0, "source_end_s": 93.0 }, "confidence": "high", "temporal_status": "localized" }, { "caption_index": 7, "loom_clip_id": "v_-ypKYhjrlXw_3387_4211", "action_id": "A3", "mentions": [ { "text": "uses", "ranges": [ [ 127, 131 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "interval": { "start_s": 25.0, "end_s": 27.5, "source_start_s": 138.0, "source_end_s": 140.5 }, "confidence": "medium", "temporal_status": "localized_sparse" } ], "id": "action_group_012", "segment_count": 3, "caption_indices": [ 4, 5, 7 ] }, { "action": "dispenses", "types": [ "simple" ], "segments": [ { "caption_index": 5, "loom_clip_id": "v_-ypKYhjrlXw_2488_2878", "action_id": "action_dispense", "mentions": [ { "text": "dispenses", "ranges": [ [ 42, 51 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "interval": { "start_s": 0.0, "end_s": 1.0, "source_start_s": 83.0, "source_end_s": 84.0 }, "confidence": "high", "temporal_status": "localized" } ], "id": "action_group_013", "segment_count": 1, "caption_indices": [ 5 ] }, { "action": "talking", "types": [ "simple" ], "segments": [ { "caption_index": 6, "loom_clip_id": "v_-ypKYhjrlXw_2878_3387", "action_id": "action_5", "mentions": [ { "text": "talking", "ranges": [ [ 147, 154 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "interval": { "start_s": 0.0, "end_s": 1.0, "source_start_s": 96.0, "source_end_s": 97.0 }, "confidence": "medium", "temporal_status": "localized_sparse" }, { "caption_index": 6, "loom_clip_id": "v_-ypKYhjrlXw_2878_3387", "action_id": "action_5", "mentions": [ { "text": "talking", "ranges": [ [ 147, 154 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "interval": { "start_s": 15.0, "end_s": 17.0, "source_start_s": 111.0, "source_end_s": 113.0 }, "confidence": "medium", "temporal_status": "localized_sparse" } ], "id": "action_group_014", "segment_count": 2, "caption_indices": [ 6 ] }, { "action": "wipes", "types": [ "prepositional" ], "segments": [ { "caption_index": 6, "loom_clip_id": "v_-ypKYhjrlXw_2878_3387", "action_id": "action_2", "mentions": [ { "text": "wipes", "ranges": [ [ 36, 41 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "interval": { "start_s": 2.0, "end_s": 4.0, "source_start_s": 98.0, "source_end_s": 100.0 }, "confidence": "high", "temporal_status": "localized_sparse" } ], "id": "action_group_015", "segment_count": 1, "caption_indices": [ 6 ] }, { "action": "holds", "types": [ "simple" ], "segments": [ { "caption_index": 6, "loom_clip_id": "v_-ypKYhjrlXw_2878_3387", "action_id": "action_4", "mentions": [ { "text": "holds", "ranges": [ [ 132, 137 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "interval": { "start_s": 4.0, "end_s": 13.0, "source_start_s": 100.0, "source_end_s": 109.0 }, "confidence": "high", "temporal_status": "localized_sparse" }, { "caption_index": 9, "loom_clip_id": "v_-ypKYhjrlXw_5306_5650", "action_id": "action_2", "mentions": [ { "text": "holds", "ranges": [ [ 42, 47 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "interval": { "start_s": 2.0, "end_s": 8.0, "source_start_s": 179.0, "source_end_s": 185.0 }, "confidence": "high", "temporal_status": "localized" } ], "id": "action_group_016", "segment_count": 2, "caption_indices": [ 6, 9 ] }, { "action": "speak", "types": [ "simple" ], "segments": [ { "caption_index": 7, "loom_clip_id": "v_-ypKYhjrlXw_3387_4211", "action_id": "A1", "mentions": [ { "text": "speak", "ranges": [ [ 20, 25 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "interval": { "start_s": 0.0, "end_s": 21.0, "source_start_s": 113.0, "source_end_s": 134.0 }, "confidence": "medium", "temporal_status": "localized_sparse" } ], "id": "action_group_017", "segment_count": 1, "caption_indices": [ 7 ] }, { "action": "applies", "types": [ "simple" ], "segments": [ { "caption_index": 7, "loom_clip_id": "v_-ypKYhjrlXw_3387_4211", "action_id": "A2", "mentions": [ { "text": "applies", "ranges": [ [ 35, 42 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "interval": { "start_s": 20.0, "end_s": 25.0, "source_start_s": 133.0, "source_end_s": 138.0 }, "confidence": "high", "temporal_status": "localized_sparse" } ], "id": "action_group_018", "segment_count": 1, "caption_indices": [ 7 ] }, { "action": "squeezes", "types": [ "simple" ], "segments": [ { "caption_index": 8, "loom_clip_id": "v_-ypKYhjrlXw_4211_5306", "action_id": "A2", "mentions": [ { "text": "squeezes", "ranges": [ [ 38, 46 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "interval": { "start_s": 20.0, "end_s": 25.0, "source_start_s": 160.5, "source_end_s": 165.5 }, "confidence": "high", "temporal_status": "supported_sparse" } ], "id": "action_group_019", "segment_count": 1, "caption_indices": [ 8 ] }, { "action": "gestures", "types": [ "prepositional" ], "segments": [ { "caption_index": 9, "loom_clip_id": "v_-ypKYhjrlXw_5306_5650", "action_id": "action_3", "mentions": [ { "text": "gestures", "ranges": [ [ 77, 85 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "interval": { "start_s": 2.0, "end_s": 3.0, "source_start_s": 179.0, "source_end_s": 180.0 }, "confidence": "high", "temporal_status": "localized" } ], "id": "action_group_020", "segment_count": 1, "caption_indices": [ 9 ] }, { "action": "speaking", "types": [ "simple" ], "segments": [ { "caption_index": 9, "loom_clip_id": "v_-ypKYhjrlXw_5306_5650", "action_id": "action_4", "mentions": [ { "text": "speaking", "ranges": [ [ 103, 111 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "interval": { "start_s": 3.0, "end_s": 11.5, "source_start_s": 180.0, "source_end_s": 188.5 }, "confidence": "high", "temporal_status": "localized" }, { "caption_index": 10, "loom_clip_id": "v_-ypKYhjrlXw_5650_6375", "action_id": "A2", "mentions": [ { "text": "speaking", "ranges": [ [ 64, 72 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "interval": { "start_s": 2.0, "end_s": 23.0, "source_start_s": 190.5, "source_end_s": 211.5 }, "confidence": "high", "temporal_status": "localized" } ], "id": "action_group_021", "segment_count": 2, "caption_indices": [ 9, 10 ] }, { "action": "closes", "types": [ "simple" ], "segments": [ { "caption_index": 9, "loom_clip_id": "v_-ypKYhjrlXw_5306_5650", "action_id": "action_5", "mentions": [ { "text": "closes", "ranges": [ [ 121, 127 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "interval": { "start_s": 8.0, "end_s": 9.0, "source_start_s": 185.0, "source_end_s": 186.0 }, "confidence": "high", "temporal_status": "localized" } ], "id": "action_group_022", "segment_count": 1, "caption_indices": [ 9 ] }, { "action": "places", "types": [ "simple", "prepositional" ], "segments": [ { "caption_index": 2, "loom_clip_id": "v_-ypKYhjrlXw_1064_1454", "action_id": "action_5", "mentions": [ { "text": "places", "ranges": [ [ 176, 182 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "interval": null, "confidence": "high", "temporal_status": "not_localized" }, { "caption_index": 9, "loom_clip_id": "v_-ypKYhjrlXw_5306_5650", "action_id": "action_6", "mentions": [ { "text": "places", "ranges": [ [ 144, 150 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "interval": { "start_s": 9.0, "end_s": 11.0, "source_start_s": 186.0, "source_end_s": 188.0 }, "confidence": "high", "temporal_status": "localized" }, { "caption_index": 10, "loom_clip_id": "v_-ypKYhjrlXw_5650_6375", "action_id": "A1", "mentions": [ { "text": "places", "ranges": [ [ 8, 14 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "interval": { "start_s": 0.0, "end_s": 2.0, "source_start_s": 188.5, "source_end_s": 190.5 }, "confidence": "high", "temporal_status": "localized" } ], "id": "action_group_023", "segment_count": 3, "caption_indices": [ 2, 9, 10 ] }, { "action": "makes various gestures", "types": [ "light_verb" ], "segments": [ { "caption_index": 10, "loom_clip_id": "v_-ypKYhjrlXw_5650_6375", "action_id": "A3", "mentions": [ { "text": "makes various gestures", "ranges": [ [ 77, 99 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "interval": { "start_s": 4.0, "end_s": 19.0, "source_start_s": 192.5, "source_end_s": 207.5 }, "confidence": "high", "temporal_status": "localized" } ], "id": "action_group_024", "segment_count": 1, "caption_indices": [ 10 ] }, { "action": "resting", "types": [ "simple" ], "segments": [ { "caption_index": 10, "loom_clip_id": "v_-ypKYhjrlXw_5650_6375", "action_id": "A4", "mentions": [ { "text": "resting", "ranges": [ [ 146, 153 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "interval": { "start_s": 19.0, "end_s": 24.180000000000007, "source_start_s": 207.5, "source_end_s": 212.68 }, "confidence": "medium", "temporal_status": "localized" } ], "id": "action_group_025", "segment_count": 1, "caption_indices": [ 10 ] }, { "action": "opens", "types": [ "simple" ], "segments": [ { "caption_index": 2, "loom_clip_id": "v_-ypKYhjrlXw_1064_1454", "action_id": "action_3", "mentions": [ { "text": "opens", "ranges": [ [ 125, 130 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "interval": null, "confidence": "high", "temporal_status": "not_localized" } ], "id": "action_group_026", "segment_count": 1, "caption_indices": [ 2 ] } ] }, { "source_video_id": "v_00Dk03Jr70M", "grouping_rule": "Exact mention phrase set after case/whitespace normalization; action type is retained as metadata but does not split groups.", "actions": [ { "action": "bends over", "types": [ "phrasal" ], "segments": [ { "caption_index": 1, "loom_clip_id": "v_00Dk03Jr70M_105_292", "action_id": "action_1", "mentions": [ { "text": "bends over", "ranges": [ [ 54, 64 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "interval": { "start_s": 1.0, "end_s": 3.0, "source_start_s": 5.2, "source_end_s": 7.2 }, "confidence": "high", "temporal_status": "localized_sparse_support" }, { "caption_index": 3, "loom_clip_id": "v_00Dk03Jr70M_1118_1368", "action_id": "action_3", "mentions": [ { "text": "bends over", "ranges": [ [ 74, 84 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "interval": { "start_s": 4.0, "end_s": 10.0, "source_start_s": 48.7, "source_end_s": 54.7 }, "confidence": "high", "temporal_status": "localized" } ], "id": "action_group_001", "segment_count": 2, "caption_indices": [ 1, 3 ] }, { "action": "scoop", "types": [ "simple" ], "segments": [ { "caption_index": 1, "loom_clip_id": "v_00Dk03Jr70M_105_292", "action_id": "action_3", "mentions": [ { "text": "scoop", "ranges": [ [ 98, 103 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "interval": { "start_s": 1.0, "end_s": 6.0, "source_start_s": 5.2, "source_end_s": 10.2 }, "confidence": "medium", "temporal_status": "localized_sparse_support" }, { "caption_index": 5, "loom_clip_id": "v_00Dk03Jr70M_2505_2830", "action_id": "action_4", "mentions": [ { "text": "scoop", "ranges": [ [ 90, 95 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "interval": { "start_s": 9.0, "end_s": 13.0, "source_start_s": 109.2, "source_end_s": 113.2 }, "confidence": "high", "temporal_status": "localized" } ], "id": "action_group_002", "segment_count": 2, "caption_indices": [ 1, 5 ] }, { "action": "uses", "types": [ "simple" ], "segments": [ { "caption_index": 1, "loom_clip_id": "v_00Dk03Jr70M_105_292", "action_id": "action_2", "mentions": [ { "text": "uses", "ranges": [ [ 81, 85 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "interval": { "start_s": 1.0, "end_s": 6.0, "source_start_s": 5.2, "source_end_s": 10.2 }, "confidence": "medium", "temporal_status": "localized_sparse_support" } ], "id": "action_group_003", "segment_count": 1, "caption_indices": [ 1 ] }, { "action": "straightening up", "types": [ "phrasal" ], "segments": [ { "caption_index": 1, "loom_clip_id": "v_00Dk03Jr70M_105_292", "action_id": "action_4", "mentions": [ { "text": "straightening up", "ranges": [ [ 181, 197 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "interval": { "start_s": 5.0, "end_s": 7.0, "source_start_s": 9.2, "source_end_s": 11.2 }, "confidence": "medium", "temporal_status": "localized_sparse_support" } ], "id": "action_group_004", "segment_count": 1, "caption_indices": [ 1 ] }, { "action": "turns", "types": [ "simple" ], "segments": [ { "caption_index": 2, "loom_clip_id": "v_00Dk03Jr70M_292_1118", "action_id": "action_1", "mentions": [ { "text": "turns", "ranges": [ [ 3, 8 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "interval": { "start_s": 0.0, "end_s": 2.0, "source_start_s": 11.7, "source_end_s": 13.7 }, "confidence": "medium", "temporal_status": "localized" }, { "caption_index": 3, "loom_clip_id": "v_00Dk03Jr70M_1118_1368", "action_id": "action_2", "mentions": [ { "text": "turns", "ranges": [ [ 40, 45 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "interval": { "start_s": 2.0, "end_s": 4.0, "source_start_s": 46.7, "source_end_s": 48.7 }, "confidence": "high", "temporal_status": "localized" }, { "caption_index": 4, "loom_clip_id": "v_00Dk03Jr70M_1368_2505", "action_id": "action_turn_1", "mentions": [ { "text": "turns", "ranges": [ [ 3, 8 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "interval": { "start_s": 0.0, "end_s": 2.0, "source_start_s": 54.7, "source_end_s": 56.7 }, "confidence": "high", "temporal_status": "localized" }, { "caption_index": 5, "loom_clip_id": "v_00Dk03Jr70M_2505_2830", "action_id": "action_1", "mentions": [ { "text": "turns", "ranges": [ [ 3, 8 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "interval": { "start_s": 1.0, "end_s": 3.0, "source_start_s": 101.2, "source_end_s": 103.2 }, "confidence": "high", "temporal_status": "localized" } ], "id": "action_group_005", "segment_count": 4, "caption_indices": [ 2, 3, 4, 5 ] }, { "action": "applying · Working · spreads", "types": [ "simple" ], "segments": [ { "caption_index": 2, "loom_clip_id": "v_00Dk03Jr70M_292_1118", "action_id": "action_2", "mentions": [ { "text": "applying", "ranges": [ [ 32, 40 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } }, { "text": "Working", "ranges": [ [ 66, 73 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } }, { "text": "spreads", "ranges": [ [ 97, 104 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "interval": { "start_s": 2.0, "end_s": 33.0, "source_start_s": 13.7, "source_end_s": 44.7 }, "confidence": "high", "temporal_status": "localized" } ], "id": "action_group_006", "segment_count": 1, "caption_indices": [ 2 ] }, { "action": "creates", "types": [ "simple" ], "segments": [ { "caption_index": 2, "loom_clip_id": "v_00Dk03Jr70M_292_1118", "action_id": "action_3", "mentions": [ { "text": "creates", "ranges": [ [ 202, 209 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "interval": { "start_s": 2.0, "end_s": 33.0, "source_start_s": 13.7, "source_end_s": 44.7 }, "confidence": "high", "temporal_status": "localized" } ], "id": "action_group_007", "segment_count": 1, "caption_indices": [ 2 ] }, { "action": "moves", "types": [ "simple" ], "segments": [ { "caption_index": 2, "loom_clip_id": "v_00Dk03Jr70M_292_1118", "action_id": "action_4", "mentions": [ { "text": "moves", "ranges": [ [ 236, 241 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "interval": { "start_s": 2.0, "end_s": 31.0, "source_start_s": 13.7, "source_end_s": 42.7 }, "confidence": "medium", "temporal_status": "localized" }, { "caption_index": 5, "loom_clip_id": "v_00Dk03Jr70M_2505_2830", "action_id": "action_2", "mentions": [ { "text": "moves", "ranges": [ [ 32, 37 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "interval": { "start_s": 2.0, "end_s": 7.0, "source_start_s": 102.2, "source_end_s": 107.2 }, "confidence": "high", "temporal_status": "localized" } ], "id": "action_group_008", "segment_count": 2, "caption_indices": [ 2, 5 ] }, { "action": "using", "types": [ "simple" ], "segments": [ { "caption_index": 3, "loom_clip_id": "v_00Dk03Jr70M_1118_1368", "action_id": "action_1", "mentions": [ { "text": "using", "ranges": [ [ 6, 11 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "interval": { "start_s": 0.0, "end_s": 3.0, "source_start_s": 44.7, "source_end_s": 47.7 }, "confidence": "medium", "temporal_status": "localized" }, { "caption_index": 3, "loom_clip_id": "v_00Dk03Jr70M_1118_1368", "action_id": "action_4", "mentions": [ { "text": "using", "ranges": [ [ 86, 91 ] ], "occurrence": 2, "particle": { "text": null, "occurrence": 1 } } ], "interval": { "start_s": 3.0, "end_s": 10.0, "source_start_s": 47.7, "source_end_s": 54.7 }, "confidence": "medium", "temporal_status": "localized" }, { "caption_index": 4, "loom_clip_id": "v_00Dk03Jr70M_1368_2505", "action_id": "action_use_1", "mentions": [ { "text": "using", "ranges": [ [ 134, 139 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "interval": { "start_s": 2.0, "end_s": 35.0, "source_start_s": 56.7, "source_end_s": 89.7 }, "confidence": "high", "temporal_status": "localized" }, { "caption_index": 4, "loom_clip_id": "v_00Dk03Jr70M_1368_2505", "action_id": "action_use_1", "mentions": [ { "text": "using", "ranges": [ [ 134, 139 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "interval": { "start_s": 37.0, "end_s": 45.5, "source_start_s": 91.7, "source_end_s": 100.2 }, "confidence": "high", "temporal_status": "localized" } ], "id": "action_group_009", "segment_count": 4, "caption_indices": [ 3, 4 ] }, { "action": "reload", "types": [ "simple" ], "segments": [ { "caption_index": 3, "loom_clip_id": "v_00Dk03Jr70M_1118_1368", "action_id": "action_5", "mentions": [ { "text": "reload", "ranges": [ [ 106, 112 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "interval": { "start_s": 3.0, "end_s": 10.0, "source_start_s": 47.7, "source_end_s": 54.7 }, "confidence": "medium", "temporal_status": "localized" } ], "id": "action_group_010", "segment_count": 1, "caption_indices": [ 3 ] }, { "action": "applies", "types": [ "simple" ], "segments": [ { "caption_index": 4, "loom_clip_id": "v_00Dk03Jr70M_1368_2505", "action_id": "action_apply_1", "mentions": [ { "text": "applies", "ranges": [ [ 88, 95 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "interval": { "start_s": 2.0, "end_s": 35.0, "source_start_s": 56.7, "source_end_s": 89.7 }, "confidence": "high", "temporal_status": "localized" }, { "caption_index": 4, "loom_clip_id": "v_00Dk03Jr70M_1368_2505", "action_id": "action_apply_1", "mentions": [ { "text": "applies", "ranges": [ [ 88, 95 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "interval": { "start_s": 37.0, "end_s": 45.5, "source_start_s": 91.7, "source_end_s": 100.2 }, "confidence": "high", "temporal_status": "localized" } ], "id": "action_group_011", "segment_count": 2, "caption_indices": [ 4 ] }, { "action": "moving", "types": [ "simple" ], "segments": [ { "caption_index": 4, "loom_clip_id": "v_00Dk03Jr70M_1368_2505", "action_id": "action_move_1", "mentions": [ { "text": "moving", "ranges": [ [ 52, 58 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "interval": { "start_s": 2.0, "end_s": 45.5, "source_start_s": 56.7, "source_end_s": 100.2 }, "confidence": "high", "temporal_status": "localized" } ], "id": "action_group_012", "segment_count": 1, "caption_indices": [ 4 ] }, { "action": "plastering", "types": [ "simple" ], "segments": [ { "caption_index": 4, "loom_clip_id": "v_00Dk03Jr70M_1368_2505", "action_id": "action_plaster_1", "mentions": [ { "text": "plastering", "ranges": [ [ 40, 50 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "interval": { "start_s": 2.0, "end_s": 45.5, "source_start_s": 56.7, "source_end_s": 100.2 }, "confidence": "high", "temporal_status": "localized" } ], "id": "action_group_013", "segment_count": 1, "caption_indices": [ 4 ] }, { "action": "smooth", "types": [ "simple" ], "segments": [ { "caption_index": 4, "loom_clip_id": "v_00Dk03Jr70M_1368_2505", "action_id": "action_smooth_1", "mentions": [ { "text": "smooth", "ranges": [ [ 154, 160 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "interval": { "start_s": 2.0, "end_s": 35.0, "source_start_s": 56.7, "source_end_s": 89.7 }, "confidence": "high", "temporal_status": "localized" }, { "caption_index": 4, "loom_clip_id": "v_00Dk03Jr70M_1368_2505", "action_id": "action_smooth_1", "mentions": [ { "text": "smooth", "ranges": [ [ 154, 160 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "interval": { "start_s": 37.0, "end_s": 45.5, "source_start_s": 91.7, "source_end_s": 100.2 }, "confidence": "high", "temporal_status": "localized" } ], "id": "action_group_014", "segment_count": 2, "caption_indices": [ 4 ] }, { "action": "bends", "types": [ "simple" ], "segments": [ { "caption_index": 5, "loom_clip_id": "v_00Dk03Jr70M_2505_2830", "action_id": "action_3", "mentions": [ { "text": "bends", "ranges": [ [ 66, 71 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "interval": { "start_s": 7.0, "end_s": 12.0, "source_start_s": 107.2, "source_end_s": 112.2 }, "confidence": "high", "temporal_status": "localized" } ], "id": "action_group_015", "segment_count": 1, "caption_indices": [ 5 ] }, { "action": "Facing", "types": [ "posture" ], "segments": [ { "caption_index": 6, "loom_clip_id": "v_00Dk03Jr70M_2830_3480", "action_id": "action_1", "mentions": [ { "text": "Facing", "ranges": [ [ 0, 6 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "interval": { "start_s": 1.0, "end_s": 25.999999999999986, "source_start_s": 114.2, "source_end_s": 139.2 }, "confidence": "high", "temporal_status": "localized" } ], "id": "action_group_016", "segment_count": 1, "caption_indices": [ 6 ] }, { "action": "applies · spread", "types": [ "simple" ], "segments": [ { "caption_index": 6, "loom_clip_id": "v_00Dk03Jr70M_2830_3480", "action_id": "action_2", "mentions": [ { "text": "applies", "ranges": [ [ 26, 33 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } }, { "text": "spread", "ranges": [ [ 128, 134 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "interval": { "start_s": 4.0, "end_s": 21.0, "source_start_s": 117.2, "source_end_s": 134.2 }, "confidence": "high", "temporal_status": "localized" } ], "id": "action_group_017", "segment_count": 1, "caption_indices": [ 6 ] }, { "action": "uses long, sweeping strokes", "types": [ "light_verb" ], "segments": [ { "caption_index": 6, "loom_clip_id": "v_00Dk03Jr70M_2830_3480", "action_id": "action_3", "mentions": [ { "text": "uses long, sweeping strokes", "ranges": [ [ 97, 124 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "interval": { "start_s": 4.0, "end_s": 21.0, "source_start_s": 117.2, "source_end_s": 134.2 }, "confidence": "high", "temporal_status": "localized" } ], "id": "action_group_018", "segment_count": 1, "caption_indices": [ 6 ] }, { "action": "smooths", "types": [ "simple" ], "segments": [ { "caption_index": 6, "loom_clip_id": "v_00Dk03Jr70M_2830_3480", "action_id": "action_4", "mentions": [ { "text": "smooths", "ranges": [ [ 162, 169 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "interval": { "start_s": 21.0, "end_s": 25.999999999999986, "source_start_s": 134.2, "source_end_s": 139.2 }, "confidence": "high", "temporal_status": "localized" } ], "id": "action_group_019", "segment_count": 1, "caption_indices": [ 6 ] }, { "action": "makes a final smoothing motion", "types": [ "light_verb" ], "segments": [ { "caption_index": 7, "loom_clip_id": "v_00Dk03Jr70M_3480_3720", "action_id": "action_1", "mentions": [ { "text": "makes a final smoothing motion", "ranges": [ [ 3, 33 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "interval": { "start_s": 0.0, "end_s": 1.0, "source_start_s": 139.2, "source_end_s": 140.2 }, "confidence": "high", "temporal_status": "localized" } ], "id": "action_group_020", "segment_count": 1, "caption_indices": [ 7 ] }, { "action": "scrapes", "types": [ "simple" ], "segments": [ { "caption_index": 7, "loom_clip_id": "v_00Dk03Jr70M_3480_3720", "action_id": "action_2", "mentions": [ { "text": "scrapes", "ranges": [ [ 35, 42 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "interval": { "start_s": 1.0, "end_s": 3.0, "source_start_s": 140.2, "source_end_s": 142.2 }, "confidence": "high", "temporal_status": "localized" }, { "caption_index": 7, "loom_clip_id": "v_00Dk03Jr70M_3480_3720", "action_id": "action_2", "mentions": [ { "text": "scrapes", "ranges": [ [ 35, 42 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "interval": { "start_s": 4.0, "end_s": 7.0, "source_start_s": 143.2, "source_end_s": 146.2 }, "confidence": "high", "temporal_status": "localized" } ], "id": "action_group_021", "segment_count": 2, "caption_indices": [ 7 ] }, { "action": "walks out", "types": [ "phrasal" ], "segments": [ { "caption_index": 7, "loom_clip_id": "v_00Dk03Jr70M_3480_3720", "action_id": "action_3", "mentions": [ { "text": "walks out", "ranges": [ [ 82, 91 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "interval": { "start_s": 7.0, "end_s": 9.600000000000023, "source_start_s": 146.2, "source_end_s": 148.8 }, "confidence": "high", "temporal_status": "localized" } ], "id": "action_group_022", "segment_count": 1, "caption_indices": [ 7 ] } ] }, { "source_video_id": "v_0J55EGFLglQ", "grouping_rule": "Exact mention phrase set after case/whitespace normalization; action type is retained as metadata but does not split groups.", "actions": [ { "action": "runs · run", "types": [ "simple" ], "segments": [ { "caption_index": 1, "loom_clip_id": "v_0J55EGFLglQ_243_591", "action_id": "action_1", "mentions": [ { "text": "runs", "ranges": [ [ 58, 62 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } }, { "text": "run", "ranges": [ [ 140, 143 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "interval": { "start_s": 1.0, "end_s": 9.0, "source_start_s": 9.11, "source_end_s": 17.11 }, "confidence": "high", "temporal_status": "localized" } ], "id": "action_group_001", "segment_count": 1, "caption_indices": [ 1 ] }, { "action": "approach", "types": [ "simple" ], "segments": [ { "caption_index": 1, "loom_clip_id": "v_0J55EGFLglQ_243_591", "action_id": "action_3", "mentions": [ { "text": "approach", "ranges": [ [ 202, 210 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "interval": { "start_s": 4.0, "end_s": 10.0, "source_start_s": 12.11, "source_end_s": 18.11 }, "confidence": "high", "temporal_status": "localized" } ], "id": "action_group_002", "segment_count": 1, "caption_indices": [ 1 ] }, { "action": "slows", "types": [ "simple" ], "segments": [ { "caption_index": 1, "loom_clip_id": "v_0J55EGFLglQ_243_591", "action_id": "action_2", "mentions": [ { "text": "slows", "ranges": [ [ 188, 193 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "interval": { "start_s": 8.0, "end_s": 11.0, "source_start_s": 16.11, "source_end_s": 19.11 }, "confidence": "medium", "temporal_status": "localized" } ], "id": "action_group_003", "segment_count": 1, "caption_indices": [ 1 ] }, { "action": "come to a complete stop", "types": [ "light_verb" ], "segments": [ { "caption_index": 1, "loom_clip_id": "v_0J55EGFLglQ_243_591", "action_id": "action_4", "mentions": [ { "text": "come to a complete stop", "ranges": [ [ 241, 264 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "interval": { "start_s": 9.0, "end_s": 11.0, "source_start_s": 17.11, "source_end_s": 19.11 }, "confidence": "high", "temporal_status": "localized" } ], "id": "action_group_004", "segment_count": 1, "caption_indices": [ 1 ] }, { "action": "standing", "types": [ "posture" ], "segments": [ { "caption_index": 1, "loom_clip_id": "v_0J55EGFLglQ_243_591", "action_id": "action_5", "mentions": [ { "text": "standing", "ranges": [ [ 266, 274 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "interval": { "start_s": 10.0, "end_s": 11.61, "source_start_s": 18.11, "source_end_s": 19.72 }, "confidence": "high", "temporal_status": "localized" } ], "id": "action_group_005", "segment_count": 1, "caption_indices": [ 1 ] } ] }, { "source_video_id": "v_50MmJszF2XM", "grouping_rule": "Exact mention phrase set after case/whitespace normalization; action type is retained as metadata but does not split groups.", "actions": [ { "action": "stands", "types": [ "posture" ], "segments": [ { "caption_index": 1, "loom_clip_id": "v_50MmJszF2XM_156_636", "action_id": "action_standing", "mentions": [ { "text": "stands", "ranges": [ [ 85, 91 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "interval": { "start_s": 1.0, "end_s": 16.0, "source_start_s": 6.21, "source_end_s": 21.21 }, "confidence": "high", "temporal_status": "supported" }, { "caption_index": 5, "loom_clip_id": "v_50MmJszF2XM_1595_2089", "action_id": "action_5", "mentions": [ { "text": "stands", "ranges": [ [ 112, 118 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "interval": { "start_s": 4.0, "end_s": 16.499999999999993, "source_start_s": 57.21, "source_end_s": 69.71 }, "confidence": "high", "temporal_status": "localized" }, { "caption_index": 10, "loom_clip_id": "v_50MmJszF2XM_3857_4097", "action_id": "action_5", "mentions": [ { "text": "stands", "ranges": [ [ 115, 121 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "interval": { "start_s": 6.0, "end_s": 8.0, "source_start_s": 134.71, "source_end_s": 136.71 }, "confidence": "high", "temporal_status": "supported" } ], "id": "action_group_001", "segment_count": 3, "caption_indices": [ 1, 5, 10 ] }, { "action": "wearing", "types": [ "simple" ], "segments": [ { "caption_index": 1, "loom_clip_id": "v_50MmJszF2XM_156_636", "action_id": "action_wearing", "mentions": [ { "text": "wearing", "ranges": [ [ 42, 49 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "interval": { "start_s": 8.0, "end_s": 16.0, "source_start_s": 13.21, "source_end_s": 21.21 }, "confidence": "high", "temporal_status": "supported" } ], "id": "action_group_002", "segment_count": 1, "caption_indices": [ 1 ] }, { "action": "being supported", "types": [ "passive" ], "segments": [ { "caption_index": 2, "loom_clip_id": "v_50MmJszF2XM_636_1014", "action_id": "action_6", "mentions": [ { "text": "being supported", "ranges": [ [ 168, 183 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "interval": { "start_s": 0.0, "end_s": 12.619999999999997, "source_start_s": 21.21, "source_end_s": 33.83 }, "confidence": "high", "temporal_status": "localized" } ], "id": "action_group_003", "segment_count": 1, "caption_indices": [ 2 ] }, { "action": "bends", "types": [ "simple" ], "segments": [ { "caption_index": 2, "loom_clip_id": "v_50MmJszF2XM_636_1014", "action_id": "action_2", "mentions": [ { "text": "bends", "ranges": [ [ 41, 46 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "interval": { "start_s": 0.0, "end_s": 3.0, "source_start_s": 21.21, "source_end_s": 24.21 }, "confidence": "high", "temporal_status": "localized" } ], "id": "action_group_004", "segment_count": 1, "caption_indices": [ 2 ] }, { "action": "swings", "types": [ "simple" ], "segments": [ { "caption_index": 2, "loom_clip_id": "v_50MmJszF2XM_636_1014", "action_id": "action_1", "mentions": [ { "text": "swings", "ranges": [ [ 9, 15 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "interval": { "start_s": 0.0, "end_s": 6.0, "source_start_s": 21.21, "source_end_s": 27.21 }, "confidence": "high", "temporal_status": "localized" }, { "caption_index": 2, "loom_clip_id": "v_50MmJszF2XM_636_1014", "action_id": "action_3", "mentions": [ { "text": "swings", "ranges": [ [ 80, 86 ] ], "occurrence": 2, "particle": { "text": null, "occurrence": 1 } } ], "interval": { "start_s": 6.0, "end_s": 12.0, "source_start_s": 27.21, "source_end_s": 33.21 }, "confidence": "high", "temporal_status": "localized" } ], "id": "action_group_005", "segment_count": 2, "caption_indices": [ 2 ] }, { "action": "arching", "types": [ "simple" ], "segments": [ { "caption_index": 2, "loom_clip_id": "v_50MmJszF2XM_636_1014", "action_id": "action_5", "mentions": [ { "text": "arching", "ranges": [ [ 145, 152 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "interval": { "start_s": 12.0, "end_s": 12.619999999999997, "source_start_s": 33.21, "source_end_s": 33.83 }, "confidence": "medium", "temporal_status": "localized" } ], "id": "action_group_006", "segment_count": 1, "caption_indices": [ 2 ] }, { "action": "repeats the upward arm swing", "types": [ "light_verb" ], "segments": [ { "caption_index": 2, "loom_clip_id": "v_50MmJszF2XM_636_1014", "action_id": "action_4", "mentions": [ { "text": "repeats the upward arm swing", "ranges": [ [ 115, 143 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "interval": { "start_s": 12.0, "end_s": 12.619999999999997, "source_start_s": 33.21, "source_end_s": 33.83 }, "confidence": "high", "temporal_status": "localized" } ], "id": "action_group_007", "segment_count": 1, "caption_indices": [ 2 ] }, { "action": "jumps", "types": [ "simple" ], "segments": [ { "caption_index": 3, "loom_clip_id": "v_50MmJszF2XM_1014_1299", "action_id": "action_jump", "mentions": [ { "text": "jumps", "ranges": [ [ 79, 84 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "interval": { "start_s": 3.0, "end_s": 5.0, "source_start_s": 36.83, "source_end_s": 38.83 }, "confidence": "high", "temporal_status": "localized" } ], "id": "action_group_008", "segment_count": 1, "caption_indices": [ 3 ] }, { "action": "performs a complete back handspring", "types": [ "light_verb" ], "segments": [ { "caption_index": 3, "loom_clip_id": "v_50MmJszF2XM_1014_1299", "action_id": "action_handspring", "mentions": [ { "text": "performs a complete back handspring", "ranges": [ [ 38, 73 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "interval": { "start_s": 3.0, "end_s": 7.0, "source_start_s": 36.83, "source_end_s": 40.83 }, "confidence": "high", "temporal_status": "localized" } ], "id": "action_group_009", "segment_count": 1, "caption_indices": [ 3 ] }, { "action": "springs", "types": [ "simple" ], "segments": [ { "caption_index": 3, "loom_clip_id": "v_50MmJszF2XM_1014_1299", "action_id": "action_spring", "mentions": [ { "text": "springs", "ranges": [ [ 113, 120 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "interval": { "start_s": 4.0, "end_s": 7.0, "source_start_s": 37.83, "source_end_s": 40.83 }, "confidence": "high", "temporal_status": "localized" } ], "id": "action_group_010", "segment_count": 1, "caption_indices": [ 3 ] }, { "action": "standing", "types": [ "posture" ], "segments": [ { "caption_index": 3, "loom_clip_id": "v_50MmJszF2XM_1014_1299", "action_id": "action_standing", "mentions": [ { "text": "standing", "ranges": [ [ 155, 163 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "interval": { "start_s": 6.0, "end_s": 9.510000000000005, "source_start_s": 39.83, "source_end_s": 43.34 }, "confidence": "high", "temporal_status": "localized" }, { "caption_index": 4, "loom_clip_id": "v_50MmJszF2XM_1299_1595", "action_id": "action_2", "mentions": [ { "text": "standing", "ranges": [ [ 51, 59 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "interval": { "start_s": 1.0, "end_s": 7.0, "source_start_s": 44.34, "source_end_s": 50.34 }, "confidence": "high", "temporal_status": "localized" }, { "caption_index": 4, "loom_clip_id": "v_50MmJszF2XM_1299_1595", "action_id": "action_7", "mentions": [ { "text": "standing", "ranges": [ [ 193, 201 ] ], "occurrence": 2, "particle": { "text": null, "occurrence": 1 } } ], "interval": { "start_s": 9.0, "end_s": 9.869999999999997, "source_start_s": 52.34, "source_end_s": 53.21 }, "confidence": "high", "temporal_status": "localized" }, { "caption_index": 6, "loom_clip_id": "v_50MmJszF2XM_2089_2545", "action_id": "action_1", "mentions": [ { "text": "standing", "ranges": [ [ 6, 14 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "interval": { "start_s": 0.0, "end_s": 10.0, "source_start_s": 69.71, "source_end_s": 79.71 }, "confidence": "high", "temporal_status": "localized_sparse" }, { "caption_index": 8, "loom_clip_id": "v_50MmJszF2XM_3018_3318", "action_id": "A1", "mentions": [ { "text": "standing", "ranges": [ [ 6, 14 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "interval": { "start_s": 0.0, "end_s": 10.0, "source_start_s": 100.71, "source_end_s": 110.71 }, "confidence": "high", "temporal_status": "localized" }, { "caption_index": 11, "loom_clip_id": "v_50MmJszF2XM_4097_4486", "action_id": "action_6", "mentions": [ { "text": "standing", "ranges": [ [ 192, 200 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "interval": { "start_s": 12.0, "end_s": 12.969999999999999, "source_start_s": 148.71, "source_end_s": 149.68 }, "confidence": "high", "temporal_status": "localized" } ], "id": "action_group_011", "segment_count": 6, "caption_indices": [ 3, 4, 6, 8, 11 ] }, { "action": "performs a back walkover", "types": [ "light_verb" ], "segments": [ { "caption_index": 4, "loom_clip_id": "v_50MmJszF2XM_1299_1595", "action_id": "action_1", "mentions": [ { "text": "performs a back walkover", "ranges": [ [ 9, 33 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "interval": { "start_s": 6.0, "end_s": 9.869999999999997, "source_start_s": 49.34, "source_end_s": 53.21 }, "confidence": "medium", "temporal_status": "localized" }, { "caption_index": 9, "loom_clip_id": "v_50MmJszF2XM_3318_3857", "action_id": "A1", "mentions": [ { "text": "performs a back walkover", "ranges": [ [ 9, 33 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "interval": { "start_s": 7.0, "end_s": 18.000000000000014, "source_start_s": 117.71, "source_end_s": 128.71 }, "confidence": "high", "temporal_status": "localized" } ], "id": "action_group_012", "segment_count": 2, "caption_indices": [ 4, 9 ] }, { "action": "raises", "types": [ "simple" ], "segments": [ { "caption_index": 4, "loom_clip_id": "v_50MmJszF2XM_1299_1595", "action_id": "action_3", "mentions": [ { "text": "raises", "ranges": [ [ 74, 80 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "interval": { "start_s": 6.0, "end_s": 8.0, "source_start_s": 49.34, "source_end_s": 51.34 }, "confidence": "high", "temporal_status": "localized" } ], "id": "action_group_013", "segment_count": 1, "caption_indices": [ 4 ] }, { "action": "arches", "types": [ "simple" ], "segments": [ { "caption_index": 4, "loom_clip_id": "v_50MmJszF2XM_1299_1595", "action_id": "action_4", "mentions": [ { "text": "arches", "ranges": [ [ 91, 97 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "interval": { "start_s": 7.0, "end_s": 9.0, "source_start_s": 50.34, "source_end_s": 52.34 }, "confidence": "high", "temporal_status": "localized" }, { "caption_index": 9, "loom_clip_id": "v_50MmJszF2XM_3318_3857", "action_id": "A3", "mentions": [ { "text": "arches", "ranges": [ [ 142, 148 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "interval": { "start_s": 10.0, "end_s": 12.0, "source_start_s": 120.71, "source_end_s": 122.71 }, "confidence": "high", "temporal_status": "localized" }, { "caption_index": 11, "loom_clip_id": "v_50MmJszF2XM_4097_4486", "action_id": "action_2", "mentions": [ { "text": "arches", "ranges": [ [ 83, 89 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "interval": { "start_s": 11.0, "end_s": 12.0, "source_start_s": 147.71, "source_end_s": 148.71 }, "confidence": "high", "temporal_status": "localized" } ], "id": "action_group_014", "segment_count": 3, "caption_indices": [ 4, 9, 11 ] }, { "action": "returns", "types": [ "simple" ], "segments": [ { "caption_index": 4, "loom_clip_id": "v_50MmJszF2XM_1299_1595", "action_id": "action_6", "mentions": [ { "text": "returns", "ranges": [ [ 180, 187 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "interval": { "start_s": 9.0, "end_s": 9.869999999999997, "source_start_s": 52.34, "source_end_s": 53.21 }, "confidence": "medium", "temporal_status": "localized" } ], "id": "action_group_015", "segment_count": 1, "caption_indices": [ 4 ] }, { "action": "lowers", "types": [ "simple" ], "segments": [ { "caption_index": 5, "loom_clip_id": "v_50MmJszF2XM_1595_2089", "action_id": "action_1", "mentions": [ { "text": "lowers", "ranges": [ [ 4, 10 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "interval": { "start_s": 0.0, "end_s": 1.0, "source_start_s": 53.21, "source_end_s": 54.21 }, "confidence": "high", "temporal_status": "localized" }, { "caption_index": 10, "loom_clip_id": "v_50MmJszF2XM_3857_4097", "action_id": "action_1", "mentions": [ { "text": "lowers", "ranges": [ [ 39, 45 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "interval": { "start_s": 1.0, "end_s": 4.0, "source_start_s": 129.71, "source_end_s": 132.71 }, "confidence": "high", "temporal_status": "supported" } ], "id": "action_group_016", "segment_count": 2, "caption_indices": [ 5, 10 ] }, { "action": "turns", "types": [ "simple" ], "segments": [ { "caption_index": 5, "loom_clip_id": "v_50MmJszF2XM_1595_2089", "action_id": "action_2", "mentions": [ { "text": "turns", "ranges": [ [ 21, 26 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "interval": { "start_s": 1.0, "end_s": 2.0, "source_start_s": 54.21, "source_end_s": 55.21 }, "confidence": "medium", "temporal_status": "localized" }, { "caption_index": 10, "loom_clip_id": "v_50MmJszF2XM_3857_4097", "action_id": "action_3", "mentions": [ { "text": "turns", "ranges": [ [ 87, 92 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "interval": { "start_s": 2.0, "end_s": 4.0, "source_start_s": 130.71, "source_end_s": 132.71 }, "confidence": "medium", "temporal_status": "supported" } ], "id": "action_group_017", "segment_count": 2, "caption_indices": [ 5, 10 ] }, { "action": "walks", "types": [ "simple" ], "segments": [ { "caption_index": 5, "loom_clip_id": "v_50MmJszF2XM_1595_2089", "action_id": "action_3", "mentions": [ { "text": "walks", "ranges": [ [ 32, 37 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "interval": { "start_s": 2.0, "end_s": 4.0, "source_start_s": 55.21, "source_end_s": 57.21 }, "confidence": "high", "temporal_status": "localized" }, { "caption_index": 7, "loom_clip_id": "v_50MmJszF2XM_2545_3018", "action_id": "action_2", "mentions": [ { "text": "walks", "ranges": [ [ 21, 26 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "interval": { "start_s": 1.0, "end_s": 5.0, "source_start_s": 85.92, "source_end_s": 89.92 }, "confidence": "high", "temporal_status": "localized" }, { "caption_index": 10, "loom_clip_id": "v_50MmJszF2XM_3857_4097", "action_id": "action_2", "mentions": [ { "text": "walks", "ranges": [ [ 59, 64 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "interval": { "start_s": 1.0, "end_s": 3.0, "source_start_s": 129.71, "source_end_s": 131.71 }, "confidence": "medium", "temporal_status": "supported" } ], "id": "action_group_018", "segment_count": 3, "caption_indices": [ 5, 7, 10 ] }, { "action": "turns back", "types": [ "phrasal" ], "segments": [ { "caption_index": 5, "loom_clip_id": "v_50MmJszF2XM_1595_2089", "action_id": "action_4", "mentions": [ { "text": "turns back", "ranges": [ [ 72, 82 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "interval": { "start_s": 3.0, "end_s": 4.0, "source_start_s": 56.21, "source_end_s": 57.21 }, "confidence": "medium", "temporal_status": "localized" } ], "id": "action_group_019", "segment_count": 1, "caption_indices": [ 5 ] }, { "action": "lunges", "types": [ "simple" ], "segments": [ { "caption_index": 6, "loom_clip_id": "v_50MmJszF2XM_2089_2545", "action_id": "action_3", "mentions": [ { "text": "lunges", "ranges": [ [ 69, 75 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "interval": { "start_s": 10.0, "end_s": 13.0, "source_start_s": 79.71, "source_end_s": 82.71 }, "confidence": "high", "temporal_status": "localized_sparse" } ], "id": "action_group_020", "segment_count": 1, "caption_indices": [ 6 ] }, { "action": "performs a roundoff", "types": [ "light_verb" ], "segments": [ { "caption_index": 6, "loom_clip_id": "v_50MmJszF2XM_2089_2545", "action_id": "action_2", "mentions": [ { "text": "performs a roundoff", "ranges": [ [ 44, 63 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "interval": { "start_s": 12.0, "end_s": 14.0, "source_start_s": 81.71, "source_end_s": 83.71 }, "confidence": "high", "temporal_status": "localized_sparse" } ], "id": "action_group_021", "segment_count": 1, "caption_indices": [ 6 ] }, { "action": "facing", "types": [ "posture" ], "segments": [ { "caption_index": 5, "loom_clip_id": "v_50MmJszF2XM_1595_2089", "action_id": "action_6", "mentions": [ { "text": "facing", "ranges": [ [ 126, 132 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "interval": null, "confidence": "low", "temporal_status": "uncertain" }, { "caption_index": 6, "loom_clip_id": "v_50MmJszF2XM_2089_2545", "action_id": "action_7", "mentions": [ { "text": "facing", "ranges": [ [ 162, 168 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "interval": { "start_s": 13.0, "end_s": 15.210000000000008, "source_start_s": 82.71, "source_end_s": 84.92 }, "confidence": "high", "temporal_status": "localized_sparse" } ], "id": "action_group_022", "segment_count": 2, "caption_indices": [ 5, 6 ] }, { "action": "landing", "types": [ "simple" ], "segments": [ { "caption_index": 6, "loom_clip_id": "v_50MmJszF2XM_2089_2545", "action_id": "action_6", "mentions": [ { "text": "landing", "ranges": [ [ 141, 148 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "interval": { "start_s": 13.0, "end_s": 14.0, "source_start_s": 82.71, "source_end_s": 83.71 }, "confidence": "high", "temporal_status": "localized_sparse" } ], "id": "action_group_023", "segment_count": 1, "caption_indices": [ 6 ] }, { "action": "finishes", "types": [ "simple" ], "segments": [ { "caption_index": 6, "loom_clip_id": "v_50MmJszF2XM_2089_2545", "action_id": "action_8", "mentions": [ { "text": "finishes", "ranges": [ [ 197, 205 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "interval": { "start_s": 14.0, "end_s": 15.0, "source_start_s": 83.71, "source_end_s": 84.71 }, "confidence": "medium", "temporal_status": "localized_sparse" } ], "id": "action_group_024", "segment_count": 1, "caption_indices": [ 6 ] }, { "action": "turns around", "types": [ "phrasal" ], "segments": [ { "caption_index": 7, "loom_clip_id": "v_50MmJszF2XM_2545_3018", "action_id": "action_1", "mentions": [ { "text": "turns around", "ranges": [ [ 4, 16 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "interval": { "start_s": 0.0, "end_s": 2.0, "source_start_s": 84.92, "source_end_s": 86.92 }, "confidence": "high", "temporal_status": "localized" } ], "id": "action_group_025", "segment_count": 1, "caption_indices": [ 7 ] }, { "action": "holds", "types": [ "simple" ], "segments": [ { "caption_index": 7, "loom_clip_id": "v_50MmJszF2XM_2545_3018", "action_id": "action_3", "mentions": [ { "text": "holds", "ranges": [ [ 186, 191 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "interval": { "start_s": 7.0, "end_s": 15.789999999999992, "source_start_s": 91.92, "source_end_s": 100.71 }, "confidence": "high", "temporal_status": "localized" } ], "id": "action_group_026", "segment_count": 1, "caption_indices": [ 7 ] }, { "action": "practices arm movements", "types": [ "light_verb" ], "segments": [ { "caption_index": 8, "loom_clip_id": "v_50MmJszF2XM_3018_3318", "action_id": "A2", "mentions": [ { "text": "practices arm movements", "ranges": [ [ 20, 43 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "interval": { "start_s": 0.0, "end_s": 4.0, "source_start_s": 100.71, "source_end_s": 104.71 }, "confidence": "high", "temporal_status": "localized" } ], "id": "action_group_027", "segment_count": 1, "caption_indices": [ 8 ] }, { "action": "transitioning", "types": [ "prepositional" ], "segments": [ { "caption_index": 8, "loom_clip_id": "v_50MmJszF2XM_3018_3318", "action_id": "A3", "mentions": [ { "text": "transitioning", "ranges": [ [ 45, 58 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "interval": { "start_s": 0.0, "end_s": 4.0, "source_start_s": 100.71, "source_end_s": 104.71 }, "confidence": "high", "temporal_status": "localized" } ], "id": "action_group_028", "segment_count": 1, "caption_indices": [ 8 ] }, { "action": "extends", "types": [ "simple" ], "segments": [ { "caption_index": 8, "loom_clip_id": "v_50MmJszF2XM_3018_3318", "action_id": "A4", "mentions": [ { "text": "extends", "ranges": [ [ 138, 145 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "interval": { "start_s": 3.0, "end_s": 6.0, "source_start_s": 103.71, "source_end_s": 106.71 }, "confidence": "high", "temporal_status": "localized" } ], "id": "action_group_029", "segment_count": 1, "caption_indices": [ 8 ] }, { "action": "holds the pose", "types": [ "light_verb" ], "segments": [ { "caption_index": 8, "loom_clip_id": "v_50MmJszF2XM_3018_3318", "action_id": "A5", "mentions": [ { "text": "holds the pose", "ranges": [ [ 183, 197 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "interval": { "start_s": 5.0, "end_s": 10.0, "source_start_s": 105.71, "source_end_s": 110.71 }, "confidence": "high", "temporal_status": "localized" } ], "id": "action_group_030", "segment_count": 1, "caption_indices": [ 8 ] }, { "action": "being assisted", "types": [ "passive" ], "segments": [ { "caption_index": 9, "loom_clip_id": "v_50MmJszF2XM_3318_3857", "action_id": "A2", "mentions": [ { "text": "being assisted", "ranges": [ [ 40, 54 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "interval": { "start_s": 0.0, "end_s": 18.000000000000014, "source_start_s": 110.71, "source_end_s": 128.71 }, "confidence": "high", "temporal_status": "localized" } ], "id": "action_group_031", "segment_count": 1, "caption_indices": [ 9 ] }, { "action": "kicks over", "types": [ "phrasal" ], "segments": [ { "caption_index": 9, "loom_clip_id": "v_50MmJszF2XM_3318_3857", "action_id": "A5", "mentions": [ { "text": "kicks over", "ranges": [ [ 194, 204 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "interval": { "start_s": 10.0, "end_s": 18.000000000000014, "source_start_s": 120.71, "source_end_s": 128.71 }, "confidence": "high", "temporal_status": "localized" } ], "id": "action_group_032", "segment_count": 1, "caption_indices": [ 9 ] }, { "action": "places", "types": [ "simple" ], "segments": [ { "caption_index": 6, "loom_clip_id": "v_50MmJszF2XM_2089_2545", "action_id": "action_4", "mentions": [ { "text": "places", "ranges": [ [ 85, 91 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "interval": null, "confidence": "high", "temporal_status": "not_observed_in_sampled_frames" }, { "caption_index": 9, "loom_clip_id": "v_50MmJszF2XM_3318_3857", "action_id": "A4", "mentions": [ { "text": "places", "ranges": [ [ 159, 165 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "interval": { "start_s": 11.0, "end_s": 13.0, "source_start_s": 121.71, "source_end_s": 123.71 }, "confidence": "high", "temporal_status": "localized" }, { "caption_index": 11, "loom_clip_id": "v_50MmJszF2XM_4097_4486", "action_id": "action_3", "mentions": [ { "text": "places", "ranges": [ [ 100, 106 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "interval": null, "confidence": "high", "temporal_status": "uncertain" } ], "id": "action_group_033", "segment_count": 3, "caption_indices": [ 6, 9, 11 ] }, { "action": "face", "types": [ "posture" ], "segments": [ { "caption_index": 10, "loom_clip_id": "v_50MmJszF2XM_3857_4097", "action_id": "action_4", "mentions": [ { "text": "face", "ranges": [ [ 96, 100 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "interval": { "start_s": 3.0, "end_s": 4.0, "source_start_s": 131.71, "source_end_s": 132.71 }, "confidence": "high", "temporal_status": "supported" }, { "caption_index": 10, "loom_clip_id": "v_50MmJszF2XM_3857_4097", "action_id": "action_4", "mentions": [ { "text": "face", "ranges": [ [ 96, 100 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "interval": { "start_s": 6.0, "end_s": 8.0, "source_start_s": 134.71, "source_end_s": 136.71 }, "confidence": "high", "temporal_status": "supported" } ], "id": "action_group_034", "segment_count": 2, "caption_indices": [ 10 ] }, { "action": "performs an unassisted back walkover", "types": [ "light_verb" ], "segments": [ { "caption_index": 11, "loom_clip_id": "v_50MmJszF2XM_4097_4486", "action_id": "action_1", "mentions": [ { "text": "performs an unassisted back walkover", "ranges": [ [ 9, 45 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "interval": { "start_s": 9.0, "end_s": 12.0, "source_start_s": 145.71, "source_end_s": 148.71 }, "confidence": "high", "temporal_status": "localized" } ], "id": "action_group_035", "segment_count": 1, "caption_indices": [ 11 ] }, { "action": "lowering", "types": [ "simple" ], "segments": [ { "caption_index": 11, "loom_clip_id": "v_50MmJszF2XM_4097_4486", "action_id": "action_7", "mentions": [ { "text": "lowering", "ranges": [ [ 229, 237 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "interval": { "start_s": 11.0, "end_s": 12.0, "source_start_s": 147.71, "source_end_s": 148.71 }, "confidence": "high", "temporal_status": "localized" } ], "id": "action_group_036", "segment_count": 1, "caption_indices": [ 11 ] }, { "action": "kicks", "types": [ "simple" ], "segments": [ { "caption_index": 4, "loom_clip_id": "v_50MmJszF2XM_1299_1595", "action_id": "action_5", "mentions": [ { "text": "kicks", "ranges": [ [ 144, 149 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "interval": null, "confidence": "medium", "temporal_status": "uncertain" }, { "caption_index": 11, "loom_clip_id": "v_50MmJszF2XM_4097_4486", "action_id": "action_4", "mentions": [ { "text": "kicks", "ranges": [ [ 131, 136 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "interval": null, "confidence": "high", "temporal_status": "uncertain" } ], "id": "action_group_037", "segment_count": 2, "caption_indices": [ 4, 11 ] }, { "action": "lands", "types": [ "simple" ], "segments": [ { "caption_index": 11, "loom_clip_id": "v_50MmJszF2XM_4097_4486", "action_id": "action_5", "mentions": [ { "text": "lands", "ranges": [ [ 156, 161 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "interval": null, "confidence": "high", "temporal_status": "uncertain" } ], "id": "action_group_038", "segment_count": 1, "caption_indices": [ 11 ] }, { "action": "propels", "types": [ "simple" ], "segments": [ { "caption_index": 6, "loom_clip_id": "v_50MmJszF2XM_2089_2545", "action_id": "action_5", "mentions": [ { "text": "propels", "ranges": [ [ 118, 125 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "interval": null, "confidence": "high", "temporal_status": "not_observed_in_sampled_frames" } ], "id": "action_group_039", "segment_count": 1, "caption_indices": [ 6 ] }, { "action": "stands up", "types": [ "phrasal" ], "segments": [ { "caption_index": 9, "loom_clip_id": "v_50MmJszF2XM_3318_3857", "action_id": "A6", "mentions": [ { "text": "stands up", "ranges": [ [ 210, 219 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "interval": null, "confidence": "high", "temporal_status": "uncertain" } ], "id": "action_group_040", "segment_count": 1, "caption_indices": [ 9 ] } ] }, { "source_video_id": "v_6gZuc4umTPk", "grouping_rule": "Exact mention phrase set after case/whitespace normalization; action type is retained as metadata but does not split groups.", "actions": [ { "action": "bring", "types": [ "simple" ], "segments": [ { "caption_index": 1, "loom_clip_id": "v_6gZuc4umTPk_1_426", "action_id": "action_4", "mentions": [ { "text": "bring", "ranges": [ [ 120, 125 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "interval": { "start_s": 0.0, "end_s": 1.0, "source_start_s": 0.0, "source_end_s": 1.0 }, "confidence": "high", "temporal_status": "localized" } ], "id": "action_group_001", "segment_count": 1, "caption_indices": [ 1 ] }, { "action": "sits", "types": [ "posture" ], "segments": [ { "caption_index": 1, "loom_clip_id": "v_6gZuc4umTPk_1_426", "action_id": "action_2", "mentions": [ { "text": "sits", "ranges": [ [ 88, 92 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "interval": { "start_s": 0.0, "end_s": 14.2, "source_start_s": 0.0, "source_end_s": 14.2 }, "confidence": "high", "temporal_status": "localized" }, { "caption_index": 2, "loom_clip_id": "v_6gZuc4umTPk_475_685", "action_id": "action_2", "mentions": [ { "text": "sits", "ranges": [ [ 101, 105 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "interval": { "start_s": 1.0, "end_s": 6.999999999999998, "source_start_s": 16.83, "source_end_s": 22.83 }, "confidence": "high", "temporal_status": "localized" } ], "id": "action_group_002", "segment_count": 2, "caption_indices": [ 1, 2 ] }, { "action": "speaks", "types": [ "simple" ], "segments": [ { "caption_index": 1, "loom_clip_id": "v_6gZuc4umTPk_1_426", "action_id": "action_3", "mentions": [ { "text": "speaks", "ranges": [ [ 97, 103 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "interval": { "start_s": 0.0, "end_s": 1.0, "source_start_s": 0.0, "source_end_s": 1.0 }, "confidence": "medium", "temporal_status": "localized" }, { "caption_index": 1, "loom_clip_id": "v_6gZuc4umTPk_1_426", "action_id": "action_3", "mentions": [ { "text": "speaks", "ranges": [ [ 97, 103 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "interval": { "start_s": 2.0, "end_s": 4.0, "source_start_s": 2.0, "source_end_s": 4.0 }, "confidence": "medium", "temporal_status": "localized" }, { "caption_index": 1, "loom_clip_id": "v_6gZuc4umTPk_1_426", "action_id": "action_3", "mentions": [ { "text": "speaks", "ranges": [ [ 97, 103 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "interval": { "start_s": 9.0, "end_s": 14.0, "source_start_s": 9.0, "source_end_s": 14.0 }, "confidence": "medium", "temporal_status": "localized" } ], "id": "action_group_003", "segment_count": 3, "caption_indices": [ 1 ] }, { "action": "wearing", "types": [ "simple" ], "segments": [ { "caption_index": 1, "loom_clip_id": "v_6gZuc4umTPk_1_426", "action_id": "action_1", "mentions": [ { "text": "wearing", "ranges": [ [ 31, 38 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "interval": { "start_s": 0.0, "end_s": 14.2, "source_start_s": 0.0, "source_end_s": 14.2 }, "confidence": "high", "temporal_status": "localized" }, { "caption_index": 2, "loom_clip_id": "v_6gZuc4umTPk_475_685", "action_id": "action_1", "mentions": [ { "text": "wearing", "ranges": [ [ 32, 39 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "interval": { "start_s": 1.0, "end_s": 6.999999999999998, "source_start_s": 16.83, "source_end_s": 22.83 }, "confidence": "high", "temporal_status": "localized" } ], "id": "action_group_004", "segment_count": 2, "caption_indices": [ 1, 2 ] }, { "action": "place", "types": [ "simple" ], "segments": [ { "caption_index": 1, "loom_clip_id": "v_6gZuc4umTPk_1_426", "action_id": "action_5", "mentions": [ { "text": "place", "ranges": [ [ 162, 167 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "interval": { "start_s": 3.0, "end_s": 4.0, "source_start_s": 3.0, "source_end_s": 4.0 }, "confidence": "high", "temporal_status": "localized" }, { "caption_index": 3, "loom_clip_id": "v_6gZuc4umTPk_685_1045", "action_id": "A5", "mentions": [ { "text": "place", "ranges": [ [ 182, 187 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "interval": { "start_s": 9.0, "end_s": 12.0, "source_start_s": 31.83, "source_end_s": 34.83 }, "confidence": "high", "temporal_status": "supported" }, { "caption_index": 11, "loom_clip_id": "v_6gZuc4umTPk_4525_4990", "action_id": "action_4", "mentions": [ { "text": "place", "ranges": [ [ 89, 94 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "interval": null, "confidence": "high", "temporal_status": "uncertain" } ], "id": "action_group_005", "segment_count": 3, "caption_indices": [ 1, 3, 11 ] }, { "action": "engage", "types": [ "prepositional" ], "segments": [ { "caption_index": 1, "loom_clip_id": "v_6gZuc4umTPk_1_426", "action_id": "action_6", "mentions": [ { "text": "engage", "ranges": [ [ 230, 236 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "interval": { "start_s": 5.0, "end_s": 11.0, "source_start_s": 5.0, "source_end_s": 11.0 }, "confidence": "high", "temporal_status": "localized" } ], "id": "action_group_006", "segment_count": 1, "caption_indices": [ 1 ] }, { "action": "leaning", "types": [ "simple" ], "segments": [ { "caption_index": 1, "loom_clip_id": "v_6gZuc4umTPk_1_426", "action_id": "action_7", "mentions": [ { "text": "leaning", "ranges": [ [ 268, 275 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "interval": { "start_s": 13.0, "end_s": 14.2, "source_start_s": 13.0, "source_end_s": 14.2 }, "confidence": "high", "temporal_status": "localized" } ], "id": "action_group_007", "segment_count": 1, "caption_indices": [ 1 ] }, { "action": "open", "types": [ "simple" ], "segments": [ { "caption_index": 2, "loom_clip_id": "v_6gZuc4umTPk_475_685", "action_id": "action_4", "mentions": [ { "text": "open", "ranges": [ [ 179, 183 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "interval": { "start_s": 2.0, "end_s": 4.0, "source_start_s": 17.83, "source_end_s": 19.83 }, "confidence": "medium", "temporal_status": "localized" }, { "caption_index": 11, "loom_clip_id": "v_6gZuc4umTPk_4525_4990", "action_id": "action_3", "mentions": [ { "text": "open", "ranges": [ [ 69, 73 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "interval": { "start_s": 2.0, "end_s": 7.0, "source_start_s": 152.83, "source_end_s": 157.83 }, "confidence": "high", "temporal_status": "localized" } ], "id": "action_group_008", "segment_count": 2, "caption_indices": [ 2, 11 ] }, { "action": "apply", "types": [ "simple" ], "segments": [ { "caption_index": 3, "loom_clip_id": "v_6gZuc4umTPk_685_1045", "action_id": "A3", "mentions": [ { "text": "apply", "ranges": [ [ 100, 105 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "interval": { "start_s": 0.0, "end_s": 8.0, "source_start_s": 22.83, "source_end_s": 30.83 }, "confidence": "high", "temporal_status": "supported" }, { "caption_index": 5, "loom_clip_id": "v_6gZuc4umTPk_1330_1930", "action_id": "action_4", "mentions": [ { "text": "apply", "ranges": [ [ 129, 134 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "interval": { "start_s": 13.0, "end_s": 20.0, "source_start_s": 57.33, "source_end_s": 64.33 }, "confidence": "high", "temporal_status": "localized" }, { "caption_index": 12, "loom_clip_id": "v_6gZuc4umTPk_4990_5755", "action_id": "action_4", "mentions": [ { "text": "apply", "ranges": [ [ 94, 99 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "interval": { "start_s": 18.0, "end_s": 25.5, "source_start_s": 184.33, "source_end_s": 191.83 }, "confidence": "high", "temporal_status": "localized_from_sparse_samples" } ], "id": "action_group_009", "segment_count": 3, "caption_indices": [ 3, 5, 12 ] }, { "action": "holding", "types": [ "simple" ], "segments": [ { "caption_index": 3, "loom_clip_id": "v_6gZuc4umTPk_685_1045", "action_id": "A2", "mentions": [ { "text": "holding", "ranges": [ [ 26, 33 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "interval": { "start_s": 0.0, "end_s": 9.0, "source_start_s": 22.83, "source_end_s": 31.83 }, "confidence": "high", "temporal_status": "supported" }, { "caption_index": 8, "loom_clip_id": "v_6gZuc4umTPk_2575_3265", "action_id": "action_4", "mentions": [ { "text": "holding", "ranges": [ [ 80, 87 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "interval": { "start_s": 12.0, "end_s": 17.0, "source_start_s": 97.83, "source_end_s": 102.83 }, "confidence": "high", "temporal_status": "supported" } ], "id": "action_group_010", "segment_count": 2, "caption_indices": [ 3, 8 ] }, { "action": "leans", "types": [ "posture" ], "segments": [ { "caption_index": 3, "loom_clip_id": "v_6gZuc4umTPk_685_1045", "action_id": "A1", "mentions": [ { "text": "leans", "ranges": [ [ 11, 16 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "interval": { "start_s": 0.0, "end_s": 8.0, "source_start_s": 22.83, "source_end_s": 30.83 }, "confidence": "high", "temporal_status": "supported" } ], "id": "action_group_011", "segment_count": 1, "caption_indices": [ 3 ] }, { "action": "sit up", "types": [ "phrasal" ], "segments": [ { "caption_index": 3, "loom_clip_id": "v_6gZuc4umTPk_685_1045", "action_id": "A4", "mentions": [ { "text": "sit up", "ranges": [ [ 171, 177 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "interval": { "start_s": 8.0, "end_s": 10.0, "source_start_s": 30.83, "source_end_s": 32.83 }, "confidence": "high", "temporal_status": "supported" } ], "id": "action_group_012", "segment_count": 1, "caption_indices": [ 3 ] }, { "action": "screws", "types": [ "prepositional", "simple" ], "segments": [ { "caption_index": 4, "loom_clip_id": "v_6gZuc4umTPk_1045_1330", "action_id": "action_1", "mentions": [ { "text": "screws", "ranges": [ [ 11, 17 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "interval": { "start_s": 1.0, "end_s": 4.0, "source_start_s": 35.83, "source_end_s": 38.83 }, "confidence": "high", "temporal_status": "supported" }, { "caption_index": 6, "loom_clip_id": "v_6gZuc4umTPk_1930_2245", "action_id": "action_2", "mentions": [ { "text": "screws", "ranges": [ [ 48, 54 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "interval": { "start_s": 3.0, "end_s": 7.0, "source_start_s": 67.33, "source_end_s": 71.33 }, "confidence": "high", "temporal_status": "localized" } ], "id": "action_group_013", "segment_count": 2, "caption_indices": [ 4, 6 ] }, { "action": "sets", "types": [ "phrasal" ], "segments": [ { "caption_index": 4, "loom_clip_id": "v_6gZuc4umTPk_1045_1330", "action_id": "action_2", "mentions": [ { "text": "sets", "ranges": [ [ 67, 71 ] ], "occurrence": 1, "particle": { "text": "down", "occurrence": 1, "range": [ 75, 79 ] } } ], "interval": { "start_s": 4.0, "end_s": 5.0, "source_start_s": 38.83, "source_end_s": 39.83 }, "confidence": "medium", "temporal_status": "supported" }, { "caption_index": 6, "loom_clip_id": "v_6gZuc4umTPk_1930_2245", "action_id": "action_3", "mentions": [ { "text": "sets", "ranges": [ [ 87, 91 ] ], "occurrence": 1, "particle": { "text": "aside", "occurrence": 1, "range": [ 95, 100 ] } } ], "interval": { "start_s": 7.0, "end_s": 8.0, "source_start_s": 71.33, "source_end_s": 72.33 }, "confidence": "medium", "temporal_status": "localized" } ], "id": "action_group_014", "segment_count": 2, "caption_indices": [ 4, 6 ] }, { "action": "pick up", "types": [ "phrasal" ], "segments": [ { "caption_index": 2, "loom_clip_id": "v_6gZuc4umTPk_475_685", "action_id": "action_3", "mentions": [ { "text": "pick up", "ranges": [ [ 128, 135 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "interval": null, "confidence": "low", "temporal_status": "uncertain" }, { "caption_index": 4, "loom_clip_id": "v_6gZuc4umTPk_1045_1330", "action_id": "action_3", "mentions": [ { "text": "pick up", "ranges": [ [ 91, 98 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "interval": { "start_s": 5.0, "end_s": 7.0, "source_start_s": 39.83, "source_end_s": 41.83 }, "confidence": "medium", "temporal_status": "supported" }, { "caption_index": 4, "loom_clip_id": "v_6gZuc4umTPk_1045_1330", "action_id": "action_6", "mentions": [ { "text": "pick up", "ranges": [ [ 163, 170 ] ], "occurrence": 2, "particle": { "text": null, "occurrence": 1 } } ], "interval": { "start_s": 8.0, "end_s": 9.5, "source_start_s": 42.83, "source_end_s": 44.33 }, "confidence": "medium", "temporal_status": "supported" }, { "caption_index": 14, "loom_clip_id": "v_6gZuc4umTPk_6370_6730", "action_id": "action_pick_up_bottle_1", "mentions": [ { "text": "pick up", "ranges": [ [ 90, 97 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "interval": { "start_s": 9.0, "end_s": 11.0, "source_start_s": 221.33, "source_end_s": 223.33 }, "confidence": "medium", "temporal_status": "localized" } ], "id": "action_group_015", "segment_count": 4, "caption_indices": [ 2, 4, 14 ] }, { "action": "inspect", "types": [ "simple" ], "segments": [ { "caption_index": 4, "loom_clip_id": "v_6gZuc4umTPk_1045_1330", "action_id": "action_4", "mentions": [ { "text": "inspect", "ranges": [ [ 115, 122 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "interval": { "start_s": 6.0, "end_s": 7.0, "source_start_s": 40.83, "source_end_s": 41.83 }, "confidence": "medium", "temporal_status": "supported" } ], "id": "action_group_016", "segment_count": 1, "caption_indices": [ 4 ] }, { "action": "set", "types": [ "phrasal" ], "segments": [ { "caption_index": 4, "loom_clip_id": "v_6gZuc4umTPk_1045_1330", "action_id": "action_5", "mentions": [ { "text": "set", "ranges": [ [ 131, 134 ] ], "occurrence": 1, "particle": { "text": "down", "occurrence": 2, "range": [ 143, 147 ] } } ], "interval": { "start_s": 7.0, "end_s": 8.0, "source_start_s": 41.83, "source_end_s": 42.83 }, "confidence": "medium", "temporal_status": "supported" }, { "caption_index": 8, "loom_clip_id": "v_6gZuc4umTPk_2575_3265", "action_id": "action_6", "mentions": [ { "text": "set", "ranges": [ [ 131, 134 ] ], "occurrence": 1, "particle": { "text": "down", "occurrence": 1, "range": [ 138, 142 ] } } ], "interval": null, "confidence": "low", "temporal_status": "not_supported_in_sampled_frames" }, { "caption_index": 10, "loom_clip_id": "v_6gZuc4umTPk_3820_4525", "action_id": "action_set_bottle_down", "mentions": [ { "text": "set", "ranges": [ [ 85, 88 ] ], "occurrence": 1, "particle": { "text": "down", "occurrence": 1, "range": { "start": 100, "end": 104 } } } ], "interval": { "start_s": 12.0, "end_s": 14.0, "source_start_s": 139.32999999999998, "source_end_s": 141.32999999999998 }, "confidence": "medium", "temporal_status": "localized_with_sparse_support" } ], "id": "action_group_017", "segment_count": 3, "caption_indices": [ 4, 8, 10 ] }, { "action": "opens", "types": [ "simple" ], "segments": [ { "caption_index": 5, "loom_clip_id": "v_6gZuc4umTPk_1330_1930", "action_id": "action_1", "mentions": [ { "text": "opens", "ranges": [ [ 11, 16 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "interval": { "start_s": 0.0, "end_s": 2.0, "source_start_s": 44.33, "source_end_s": 46.33 }, "confidence": "medium", "temporal_status": "localized" }, { "caption_index": 9, "loom_clip_id": "v_6gZuc4umTPk_3265_3820", "action_id": "action_open", "mentions": [ { "text": "opens", "ranges": [ [ 51, 56 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "interval": { "start_s": 13.0, "end_s": 18.0, "source_start_s": 121.83, "source_end_s": 126.83 }, "confidence": "high", "temporal_status": "localized_sparse_support" } ], "id": "action_group_018", "segment_count": 2, "caption_indices": [ 5, 9 ] }, { "action": "take hold", "types": [ "light_verb" ], "segments": [ { "caption_index": 5, "loom_clip_id": "v_6gZuc4umTPk_1330_1930", "action_id": "action_3", "mentions": [ { "text": "take hold", "ranges": [ [ 78, 87 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "interval": { "start_s": 2.0, "end_s": 4.0, "source_start_s": 46.33, "source_end_s": 48.33 }, "confidence": "medium", "temporal_status": "localized" }, { "caption_index": 5, "loom_clip_id": "v_6gZuc4umTPk_1330_1930", "action_id": "action_3", "mentions": [ { "text": "take hold", "ranges": [ [ 78, 87 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "interval": { "start_s": 11.0, "end_s": 20.0, "source_start_s": 55.33, "source_end_s": 64.33 }, "confidence": "medium", "temporal_status": "localized" } ], "id": "action_group_019", "segment_count": 2, "caption_indices": [ 5 ] }, { "action": "holds", "types": [ "simple", "phrasal" ], "segments": [ { "caption_index": 5, "loom_clip_id": "v_6gZuc4umTPk_1330_1930", "action_id": "action_2", "mentions": [ { "text": "holds", "ranges": [ [ 51, 56 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "interval": { "start_s": 5.0, "end_s": 20.0, "source_start_s": 49.33, "source_end_s": 64.33 }, "confidence": "high", "temporal_status": "localized" }, { "caption_index": 13, "loom_clip_id": "v_6gZuc4umTPk_5755_6370", "action_id": "A1", "mentions": [ { "text": "holds", "ranges": [ [ 11, 16 ] ], "occurrence": 1, "particle": { "text": "up", "occurrence": 1, "ranges": [ [ 41, 43 ] ] } } ], "interval": { "start_s": 2.0, "end_s": 17.0, "source_start_s": 193.83, "source_end_s": 208.83 }, "confidence": "high", "temporal_status": "supported" } ], "id": "action_group_020", "segment_count": 2, "caption_indices": [ 5, 13 ] }, { "action": "pull", "types": [ "simple" ], "segments": [ { "caption_index": 6, "loom_clip_id": "v_6gZuc4umTPk_1930_2245", "action_id": "action_4", "mentions": [ { "text": "pull", "ranges": [ [ 112, 116 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "interval": { "start_s": 8.0, "end_s": 10.5, "source_start_s": 72.33, "source_end_s": 74.83 }, "confidence": "high", "temporal_status": "localized" } ], "id": "action_group_021", "segment_count": 1, "caption_indices": [ 6 ] }, { "action": "taps · tapping", "types": [ "simple" ], "segments": [ { "caption_index": 7, "loom_clip_id": "v_6gZuc4umTPk_2245_2575", "action_id": "action_3", "mentions": [ { "text": "taps", "ranges": [ [ 84, 88 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } }, { "text": "tapping", "ranges": [ [ 156, 163 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "interval": { "start_s": 4.0, "end_s": 7.0, "source_start_s": 78.83, "source_end_s": 81.83 }, "confidence": "high", "temporal_status": "supported" } ], "id": "action_group_022", "segment_count": 1, "caption_indices": [ 7 ] }, { "action": "claps · clapping", "types": [ "simple" ], "segments": [ { "caption_index": 7, "loom_clip_id": "v_6gZuc4umTPk_2245_2575", "action_id": "action_1", "mentions": [ { "text": "claps", "ranges": [ [ 11, 16 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } }, { "text": "clapping", "ranges": [ [ 143, 151 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "interval": { "start_s": 7.0, "end_s": 11.0, "source_start_s": 81.83, "source_end_s": 85.83 }, "confidence": "high", "temporal_status": "supported" } ], "id": "action_group_023", "segment_count": 1, "caption_indices": [ 7 ] }, { "action": "puts down", "types": [ "phrasal" ], "segments": [ { "caption_index": 8, "loom_clip_id": "v_6gZuc4umTPk_2575_3265", "action_id": "action_1", "mentions": [ { "text": "puts down", "ranges": [ [ 11, 20 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "interval": { "start_s": 0.0, "end_s": 1.0, "source_start_s": 85.83, "source_end_s": 86.83 }, "confidence": "medium", "temporal_status": "supported" } ], "id": "action_group_024", "segment_count": 1, "caption_indices": [ 8 ] }, { "action": "picks up", "types": [ "phrasal" ], "segments": [ { "caption_index": 7, "loom_clip_id": "v_6gZuc4umTPk_2245_2575", "action_id": "action_2", "mentions": [ { "text": "picks up", "ranges": [ [ 36, 44 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "interval": null, "confidence": "high", "temporal_status": "uncertain" }, { "caption_index": 8, "loom_clip_id": "v_6gZuc4umTPk_2575_3265", "action_id": "action_2", "mentions": [ { "text": "picks up", "ranges": [ [ 33, 41 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "interval": { "start_s": 1.0, "end_s": 3.0, "source_start_s": 86.83, "source_end_s": 88.83 }, "confidence": "medium", "temporal_status": "supported" }, { "caption_index": 9, "loom_clip_id": "v_6gZuc4umTPk_3265_3820", "action_id": "action_pick_up", "mentions": [ { "text": "picks up", "ranges": [ [ 11, 19 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "interval": { "start_s": 10.0, "end_s": 13.0, "source_start_s": 118.83, "source_end_s": 121.83 }, "confidence": "medium", "temporal_status": "localized_sparse_support" }, { "caption_index": 11, "loom_clip_id": "v_6gZuc4umTPk_4525_4990", "action_id": "action_2", "mentions": [ { "text": "picks up", "ranges": [ [ 23, 31 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "interval": { "start_s": 1.0, "end_s": 3.0, "source_start_s": 151.83, "source_end_s": 153.83 }, "confidence": "high", "temporal_status": "localized" }, { "caption_index": 14, "loom_clip_id": "v_6gZuc4umTPk_6370_6730", "action_id": "action_pick_up_object_1", "mentions": [ { "text": "picks up", "ranges": [ [ 11, 19 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "interval": { "start_s": 2.0, "end_s": 4.0, "source_start_s": 214.33, "source_end_s": 216.33 }, "confidence": "medium", "temporal_status": "localized" }, { "caption_index": 15, "loom_clip_id": "v_6gZuc4umTPk_6730_6990", "action_id": "action_pick_up", "mentions": [ { "text": "picks up", "ranges": [ [ 11, 19 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "interval": { "start_s": 1.0, "end_s": 4.0, "source_start_s": 225.33, "source_end_s": 228.33 }, "confidence": "high", "temporal_status": "supported" } ], "id": "action_group_025", "segment_count": 6, "caption_indices": [ 7, 8, 9, 11, 14, 15 ] }, { "action": "uncap", "types": [ "simple" ], "segments": [ { "caption_index": 8, "loom_clip_id": "v_6gZuc4umTPk_2575_3265", "action_id": "action_3", "mentions": [ { "text": "uncap", "ranges": [ [ 67, 72 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "interval": { "start_s": 5.0, "end_s": 12.0, "source_start_s": 90.83, "source_end_s": 97.83 }, "confidence": "medium", "temporal_status": "supported" } ], "id": "action_group_026", "segment_count": 1, "caption_indices": [ 8 ] }, { "action": "recap", "types": [ "simple" ], "segments": [ { "caption_index": 8, "loom_clip_id": "v_6gZuc4umTPk_2575_3265", "action_id": "action_5", "mentions": [ { "text": "recap", "ranges": [ [ 110, 115 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "interval": { "start_s": 17.0, "end_s": 22.0, "source_start_s": 102.83, "source_end_s": 107.83 }, "confidence": "medium", "temporal_status": "supported" } ], "id": "action_group_027", "segment_count": 1, "caption_indices": [ 8 ] }, { "action": "hold", "types": [ "simple" ], "segments": [ { "caption_index": 9, "loom_clip_id": "v_6gZuc4umTPk_3265_3820", "action_id": "action_hold", "mentions": [ { "text": "hold", "ranges": [ [ 66, 70 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "interval": { "start_s": 13.0, "end_s": 18.5, "source_start_s": 121.83, "source_end_s": 127.33 }, "confidence": "medium", "temporal_status": "localized_sparse_support" }, { "caption_index": 12, "loom_clip_id": "v_6gZuc4umTPk_4990_5755", "action_id": "action_3", "mentions": [ { "text": "hold", "ranges": [ [ 60, 64 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "interval": { "start_s": 3.0, "end_s": 4.0, "source_start_s": 169.33, "source_end_s": 170.33 }, "confidence": "high", "temporal_status": "localized_from_sparse_samples" }, { "caption_index": 12, "loom_clip_id": "v_6gZuc4umTPk_4990_5755", "action_id": "action_3", "mentions": [ { "text": "hold", "ranges": [ [ 60, 64 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "interval": { "start_s": 17.0, "end_s": 25.5, "source_start_s": 183.33, "source_end_s": 191.83 }, "confidence": "high", "temporal_status": "localized_from_sparse_samples" } ], "id": "action_group_028", "segment_count": 3, "caption_indices": [ 9, 12 ] }, { "action": "look", "types": [ "prepositional", "simple" ], "segments": [ { "caption_index": 9, "loom_clip_id": "v_6gZuc4umTPk_3265_3820", "action_id": "action_look", "mentions": [ { "text": "look", "ranges": [ [ 85, 89 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "interval": { "start_s": 13.0, "end_s": 18.0, "source_start_s": 121.83, "source_end_s": 126.83 }, "confidence": "medium", "temporal_status": "localized_sparse_support" }, { "caption_index": 13, "loom_clip_id": "v_6gZuc4umTPk_5755_6370", "action_id": "A5", "mentions": [ { "text": "look", "ranges": [ [ 137, 141 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1, "ranges": [] } } ], "interval": { "start_s": 20.0, "end_s": 20.5, "source_start_s": 211.83, "source_end_s": 212.33 }, "confidence": "medium", "temporal_status": "supported" } ], "id": "action_group_029", "segment_count": 2, "caption_indices": [ 9, 13 ] }, { "action": "talk", "types": [ "simple" ], "segments": [ { "caption_index": 9, "loom_clip_id": "v_6gZuc4umTPk_3265_3820", "action_id": "action_talk", "mentions": [ { "text": "talk", "ranges": [ [ 101, 105 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "interval": { "start_s": 13.0, "end_s": 18.5, "source_start_s": 121.83, "source_end_s": 127.33 }, "confidence": "medium", "temporal_status": "localized_sparse_support" } ], "id": "action_group_030", "segment_count": 1, "caption_indices": [ 9 ] }, { "action": "gesturing", "types": [ "simple" ], "segments": [ { "caption_index": 9, "loom_clip_id": "v_6gZuc4umTPk_3265_3820", "action_id": "action_gesture", "mentions": [ { "text": "gesturing", "ranges": [ [ 112, 121 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "interval": { "start_s": 15.0, "end_s": 18.5, "source_start_s": 123.83, "source_end_s": 127.33 }, "confidence": "medium", "temporal_status": "localized_sparse_support" } ], "id": "action_group_031", "segment_count": 1, "caption_indices": [ 9 ] }, { "action": "puts", "types": [ "phrasal", "prepositional" ], "segments": [ { "caption_index": 10, "loom_clip_id": "v_6gZuc4umTPk_3820_4525", "action_id": "action_put_brush_back", "mentions": [ { "text": "puts", "ranges": [ [ 11, 15 ] ], "occurrence": 1, "particle": { "text": "back", "occurrence": 1, "range": { "start": 26, "end": 30 } } } ], "interval": { "start_s": 9.0, "end_s": 11.0, "source_start_s": 136.32999999999998, "source_end_s": 138.32999999999998 }, "confidence": "high", "temporal_status": "localized" }, { "caption_index": 15, "loom_clip_id": "v_6gZuc4umTPk_6730_6990", "action_id": "action_put", "mentions": [ { "text": "puts", "ranges": [ [ 40, 44 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "interval": { "start_s": 3.0, "end_s": 5.0, "source_start_s": 227.33, "source_end_s": 229.33 }, "confidence": "high", "temporal_status": "supported" } ], "id": "action_group_032", "segment_count": 2, "caption_indices": [ 10, 15 ] }, { "action": "closes", "types": [ "simple" ], "segments": [ { "caption_index": 10, "loom_clip_id": "v_6gZuc4umTPk_3820_4525", "action_id": "action_close_bottle", "mentions": [ { "text": "closes", "ranges": [ [ 69, 75 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1, "ranges": [] } } ], "interval": { "start_s": 10.0, "end_s": 11.0, "source_start_s": 137.32999999999998, "source_end_s": 138.32999999999998 }, "confidence": "medium", "temporal_status": "localized" } ], "id": "action_group_033", "segment_count": 1, "caption_indices": [ 10 ] }, { "action": "bend · leaning", "types": [ "posture" ], "segments": [ { "caption_index": 10, "loom_clip_id": "v_6gZuc4umTPk_3820_4525", "action_id": "action_bend_and_lean", "mentions": [ { "text": "bend", "ranges": [ [ 114, 118 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1, "ranges": [] } }, { "text": "leaning", "ranges": [ [ 128, 135 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1, "ranges": [] } } ], "interval": { "start_s": 15.0, "end_s": 23.0, "source_start_s": 142.32999999999998, "source_end_s": 150.32999999999998 }, "confidence": "high", "temporal_status": "localized" } ], "id": "action_group_034", "segment_count": 1, "caption_indices": [ 10 ] }, { "action": "sits up", "types": [ "posture" ], "segments": [ { "caption_index": 11, "loom_clip_id": "v_6gZuc4umTPk_4525_4990", "action_id": "action_1", "mentions": [ { "text": "sits up", "ranges": [ [ 11, 18 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "interval": { "start_s": 0.0, "end_s": 2.0, "source_start_s": 150.83, "source_end_s": 152.83 }, "confidence": "high", "temporal_status": "localized" } ], "id": "action_group_035", "segment_count": 1, "caption_indices": [ 11 ] }, { "action": "reaches", "types": [ "simple" ], "segments": [ { "caption_index": 12, "loom_clip_id": "v_6gZuc4umTPk_4990_5755", "action_id": "action_1", "mentions": [ { "text": "reaches", "ranges": [ [ 11, 18 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "interval": { "start_s": 0.0, "end_s": 3.0, "source_start_s": 166.33, "source_end_s": 169.33 }, "confidence": "medium", "temporal_status": "localized_from_sparse_samples" } ], "id": "action_group_036", "segment_count": 1, "caption_indices": [ 12 ] }, { "action": "take", "types": [ "simple" ], "segments": [ { "caption_index": 12, "loom_clip_id": "v_6gZuc4umTPk_4990_5755", "action_id": "action_2", "mentions": [ { "text": "take", "ranges": [ [ 27, 31 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "interval": { "start_s": 2.0, "end_s": 4.0, "source_start_s": 168.33, "source_end_s": 170.33 }, "confidence": "medium", "temporal_status": "localized_from_sparse_samples" } ], "id": "action_group_037", "segment_count": 1, "caption_indices": [ 12 ] }, { "action": "blows", "types": [ "prepositional" ], "segments": [ { "caption_index": 13, "loom_clip_id": "v_6gZuc4umTPk_5755_6370", "action_id": "A2", "mentions": [ { "text": "blows", "ranges": [ [ 45, 50 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1, "ranges": [] } } ], "interval": { "start_s": 3.0, "end_s": 6.0, "source_start_s": 194.83, "source_end_s": 197.83 }, "confidence": "medium", "temporal_status": "supported" } ], "id": "action_group_038", "segment_count": 1, "caption_indices": [ 13 ] }, { "action": "examines", "types": [ "simple" ], "segments": [ { "caption_index": 13, "loom_clip_id": "v_6gZuc4umTPk_5755_6370", "action_id": "A3", "mentions": [ { "text": "examines", "ranges": [ [ 85, 93 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1, "ranges": [] } } ], "interval": { "start_s": 6.0, "end_s": 17.0, "source_start_s": 197.83, "source_end_s": 208.83 }, "confidence": "high", "temporal_status": "supported" } ], "id": "action_group_039", "segment_count": 1, "caption_indices": [ 13 ] }, { "action": "release", "types": [ "simple" ], "segments": [ { "caption_index": 13, "loom_clip_id": "v_6gZuc4umTPk_5755_6370", "action_id": "A4", "mentions": [ { "text": "release", "ranges": [ [ 116, 123 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1, "ranges": [] } } ], "interval": { "start_s": 17.0, "end_s": 20.0, "source_start_s": 208.83, "source_end_s": 211.83 }, "confidence": "high", "temporal_status": "supported" } ], "id": "action_group_040", "segment_count": 1, "caption_indices": [ 13 ] }, { "action": "places", "types": [ "simple" ], "segments": [ { "caption_index": 14, "loom_clip_id": "v_6gZuc4umTPk_6370_6730", "action_id": "action_place_object_1", "mentions": [ { "text": "places", "ranges": [ [ 46, 52 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "interval": { "start_s": 4.0, "end_s": 8.0, "source_start_s": 216.33, "source_end_s": 220.33 }, "confidence": "medium", "temporal_status": "localized" } ], "id": "action_group_041", "segment_count": 1, "caption_indices": [ 14 ] }, { "action": "close", "types": [ "simple" ], "segments": [ { "caption_index": 14, "loom_clip_id": "v_6gZuc4umTPk_6370_6730", "action_id": "action_close_box_1", "mentions": [ { "text": "close", "ranges": [ [ 132, 137 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "interval": { "start_s": 11.0, "end_s": 12.0, "source_start_s": 223.33, "source_end_s": 224.33 }, "confidence": "medium", "temporal_status": "localized" } ], "id": "action_group_042", "segment_count": 1, "caption_indices": [ 14 ] }, { "action": "give a thumbs-up gesture", "types": [ "light_verb" ], "segments": [ { "caption_index": 15, "loom_clip_id": "v_6gZuc4umTPk_6730_6990", "action_id": "action_thumbs_up", "mentions": [ { "text": "give a thumbs-up gesture", "ranges": [ [ 72, 96 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "interval": { "start_s": 5.0, "end_s": 6.0, "source_start_s": 229.33, "source_end_s": 230.33 }, "confidence": "high", "temporal_status": "supported" } ], "id": "action_group_043", "segment_count": 1, "caption_indices": [ 15 ] }, { "action": "waving", "types": [ "simple" ], "segments": [ { "caption_index": 15, "loom_clip_id": "v_6gZuc4umTPk_6730_6990", "action_id": "action_wave", "mentions": [ { "text": "waving", "ranges": [ [ 104, 110 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "interval": { "start_s": 7.0, "end_s": 8.669999999999987, "source_start_s": 231.33, "source_end_s": 233.0 }, "confidence": "high", "temporal_status": "supported" } ], "id": "action_group_044", "segment_count": 1, "caption_indices": [ 15 ] }, { "action": "applying", "types": [ "simple" ], "segments": [ { "caption_index": 6, "loom_clip_id": "v_6gZuc4umTPk_1930_2245", "action_id": "action_1", "mentions": [ { "text": "applying", "ranges": [ [ 20, 28 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "interval": null, "confidence": "medium", "temporal_status": "not_confidently_localized" } ], "id": "action_group_045", "segment_count": 1, "caption_indices": [ 6 ] }, { "action": "looking around", "types": [ "phrasal" ], "segments": [ { "caption_index": 11, "loom_clip_id": "v_6gZuc4umTPk_4525_4990", "action_id": "action_5", "mentions": [ { "text": "looking around", "ranges": [ [ 128, 142 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "interval": null, "confidence": "high", "temporal_status": "uncertain" } ], "id": "action_group_046", "segment_count": 1, "caption_indices": [ 11 ] } ] }, { "source_video_id": "v_9A0BwCxAKAg", "grouping_rule": "Exact mention phrase set after case/whitespace normalization; action type is retained as metadata but does not split groups.", "actions": [ { "action": "is speaking", "types": [ "simple" ], "segments": [ { "caption_index": 1, "loom_clip_id": "v_9A0BwCxAKAg_1_195", "action_id": "action_3", "mentions": [ { "text": "is speaking", "ranges": [ [ 159, 170 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "interval": { "start_s": 0.0, "end_s": 1.0, "source_start_s": 0.0, "source_end_s": 1.0 }, "confidence": "medium", "temporal_status": "localized_sparse" }, { "caption_index": 1, "loom_clip_id": "v_9A0BwCxAKAg_1_195", "action_id": "action_3", "mentions": [ { "text": "is speaking", "ranges": [ [ 159, 170 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "interval": { "start_s": 2.0, "end_s": 3.0, "source_start_s": 2.0, "source_end_s": 3.0 }, "confidence": "medium", "temporal_status": "localized_sparse" }, { "caption_index": 1, "loom_clip_id": "v_9A0BwCxAKAg_1_195", "action_id": "action_3", "mentions": [ { "text": "is speaking", "ranges": [ [ 159, 170 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "interval": { "start_s": 5.0, "end_s": 6.5, "source_start_s": 5.0, "source_end_s": 6.5 }, "confidence": "medium", "temporal_status": "localized_sparse" } ], "id": "action_group_001", "segment_count": 3, "caption_indices": [ 1 ] }, { "action": "is wearing", "types": [ "simple" ], "segments": [ { "caption_index": 1, "loom_clip_id": "v_9A0BwCxAKAg_1_195", "action_id": "action_2", "mentions": [ { "text": "is wearing", "ranges": [ [ 107, 117 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "interval": { "start_s": 0.0, "end_s": 6.5, "source_start_s": 0.0, "source_end_s": 6.5 }, "confidence": "high", "temporal_status": "localized" } ], "id": "action_group_002", "segment_count": 1, "caption_indices": [ 1 ] }, { "action": "looking", "types": [ "prepositional" ], "segments": [ { "caption_index": 1, "loom_clip_id": "v_9A0BwCxAKAg_1_195", "action_id": "action_4", "mentions": [ { "text": "looking", "ranges": [ [ 177, 184 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "interval": { "start_s": 0.0, "end_s": 6.5, "source_start_s": 0.0, "source_end_s": 6.5 }, "confidence": "medium", "temporal_status": "localized" } ], "id": "action_group_003", "segment_count": 1, "caption_indices": [ 1 ] }, { "action": "stands", "types": [ "posture" ], "segments": [ { "caption_index": 1, "loom_clip_id": "v_9A0BwCxAKAg_1_195", "action_id": "action_1", "mentions": [ { "text": "stands", "ranges": [ [ 66, 72 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "interval": { "start_s": 0.0, "end_s": 6.5, "source_start_s": 0.0, "source_end_s": 6.5 }, "confidence": "medium", "temporal_status": "localized" } ], "id": "action_group_004", "segment_count": 1, "caption_indices": [ 1 ] }, { "action": "turns", "types": [ "simple" ], "segments": [ { "caption_index": 2, "loom_clip_id": "v_9A0BwCxAKAg_195_450", "action_id": "action_1", "mentions": [ { "text": "turns", "ranges": [ [ 11, 16 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "interval": { "start_s": 0.0, "end_s": 1.0, "source_start_s": 6.5, "source_end_s": 7.5 }, "confidence": "medium", "temporal_status": "localized_sparse_support" }, { "caption_index": 3, "loom_clip_id": "v_9A0BwCxAKAg_450_1004", "action_id": "action_2", "mentions": [ { "text": "turns", "ranges": [ [ 41, 46 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "interval": null, "confidence": "medium", "temporal_status": "uncertain" }, { "caption_index": 9, "loom_clip_id": "v_9A0BwCxAKAg_4900_5783", "action_id": "action_1", "mentions": [ { "text": "turns", "ranges": [ [ 11, 16 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "interval": { "start_s": 0.0, "end_s": 3.0, "source_start_s": 163.5, "source_end_s": 166.5 }, "confidence": "high", "temporal_status": "supported" } ], "id": "action_group_005", "segment_count": 3, "caption_indices": [ 2, 3, 9 ] }, { "action": "face", "types": [ "posture" ], "segments": [ { "caption_index": 2, "loom_clip_id": "v_9A0BwCxAKAg_195_450", "action_id": "action_2", "mentions": [ { "text": "face", "ranges": [ [ 42, 46 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "interval": { "start_s": 1.0, "end_s": 7.0, "source_start_s": 7.5, "source_end_s": 13.5 }, "confidence": "high", "temporal_status": "localized_sparse_support" }, { "caption_index": 14, "loom_clip_id": "v_9A0BwCxAKAg_6653_6918", "action_id": "action_6", "mentions": [ { "text": "face", "ranges": [ [ 146, 150 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "interval": { "start_s": 6.0, "end_s": 8.840000000000003, "source_start_s": 227.99, "source_end_s": 230.83 }, "confidence": "high", "temporal_status": "localized" } ], "id": "action_group_006", "segment_count": 2, "caption_indices": [ 2, 14 ] }, { "action": "gestures · pointing", "types": [ "simple" ], "segments": [ { "caption_index": 2, "loom_clip_id": "v_9A0BwCxAKAg_195_450", "action_id": "action_3", "mentions": [ { "text": "gestures", "ranges": [ [ 71, 79 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } }, { "text": "pointing", "ranges": [ [ 100, 108 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "interval": { "start_s": 1.0, "end_s": 7.0, "source_start_s": 7.5, "source_end_s": 13.5 }, "confidence": "high", "temporal_status": "localized_sparse_support" } ], "id": "action_group_007", "segment_count": 1, "caption_indices": [ 2 ] }, { "action": "turning", "types": [ "simple" ], "segments": [ { "caption_index": 2, "loom_clip_id": "v_9A0BwCxAKAg_195_450", "action_id": "action_4", "mentions": [ { "text": "turning", "ranges": [ [ 147, 154 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "interval": { "start_s": 7.0, "end_s": 8.5, "source_start_s": 13.5, "source_end_s": 15.0 }, "confidence": "medium", "temporal_status": "localized_sparse_support" } ], "id": "action_group_008", "segment_count": 1, "caption_indices": [ 2 ] }, { "action": "picks up", "types": [ "phrasal" ], "segments": [ { "caption_index": 3, "loom_clip_id": "v_9A0BwCxAKAg_450_1004", "action_id": "action_1", "mentions": [ { "text": "picks up", "ranges": [ [ 3, 11 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "interval": { "start_s": 0.0, "end_s": 1.0, "source_start_s": 15.0, "source_end_s": 16.0 }, "confidence": "medium", "temporal_status": "sparsely_supported" }, { "caption_index": 5, "loom_clip_id": "v_9A0BwCxAKAg_1424_2068", "action_id": "a4", "mentions": [ { "text": "picks up", "ranges": [ [ 100, 108 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "interval": { "start_s": 8.0, "end_s": 12.0, "source_start_s": 55.5, "source_end_s": 59.5 }, "confidence": "medium", "temporal_status": "localized" }, { "caption_index": 8, "loom_clip_id": "v_9A0BwCxAKAg_4136_4900", "action_id": "action_2", "mentions": [ { "text": "picks up", "ranges": [ [ 19, 27 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "interval": { "start_s": 1.0, "end_s": 3.0, "source_start_s": 139.0, "source_end_s": 141.0 }, "confidence": "high", "temporal_status": "localized" }, { "caption_index": 11, "loom_clip_id": "v_9A0BwCxAKAg_6009_6233", "action_id": "action_2", "mentions": [ { "text": "picks up", "ranges": [ [ 30, 38 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "interval": { "start_s": 1.0, "end_s": 2.0, "source_start_s": 201.49, "source_end_s": 202.49 }, "confidence": "high", "temporal_status": "localized_sparse_support" } ], "id": "action_group_009", "segment_count": 4, "caption_indices": [ 3, 5, 8, 11 ] }, { "action": "moving", "types": [ "simple" ], "segments": [ { "caption_index": 3, "loom_clip_id": "v_9A0BwCxAKAg_450_1004", "action_id": "action_5", "mentions": [ { "text": "moving", "ranges": [ [ 129, 135 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "interval": { "start_s": 1.0, "end_s": 17.0, "source_start_s": 16.0, "source_end_s": 32.0 }, "confidence": "high", "temporal_status": "supported" }, { "caption_index": 3, "loom_clip_id": "v_9A0BwCxAKAg_450_1004", "action_id": "action_5", "mentions": [ { "text": "moving", "ranges": [ [ 129, 135 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "interval": { "start_s": 18.0, "end_s": 18.5, "source_start_s": 33.0, "source_end_s": 33.5 }, "confidence": "high", "temporal_status": "supported" }, { "caption_index": 12, "loom_clip_id": "v_9A0BwCxAKAg_6233_6428", "action_id": "action_3", "mentions": [ { "text": "moving", "ranges": [ [ 68, 74 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "interval": { "start_s": 3.0, "end_s": 6.0, "source_start_s": 210.99, "source_end_s": 213.99 }, "confidence": "medium", "temporal_status": "localized" } ], "id": "action_group_010", "segment_count": 3, "caption_indices": [ 3, 12 ] }, { "action": "Using", "types": [ "simple" ], "segments": [ { "caption_index": 3, "loom_clip_id": "v_9A0BwCxAKAg_450_1004", "action_id": "action_3", "mentions": [ { "text": "Using", "ranges": [ [ 67, 72 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "interval": { "start_s": 1.0, "end_s": 17.0, "source_start_s": 16.0, "source_end_s": 32.0 }, "confidence": "high", "temporal_status": "supported" }, { "caption_index": 3, "loom_clip_id": "v_9A0BwCxAKAg_450_1004", "action_id": "action_3", "mentions": [ { "text": "Using", "ranges": [ [ 67, 72 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "interval": { "start_s": 18.0, "end_s": 18.5, "source_start_s": 33.0, "source_end_s": 33.5 }, "confidence": "high", "temporal_status": "supported" }, { "caption_index": 9, "loom_clip_id": "v_9A0BwCxAKAg_4900_5783", "action_id": "action_3", "mentions": [ { "text": "using", "ranges": [ [ 37, 42 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "interval": { "start_s": 4.0, "end_s": 24.0, "source_start_s": 167.5, "source_end_s": 187.5 }, "confidence": "high", "temporal_status": "supported" } ], "id": "action_group_011", "segment_count": 3, "caption_indices": [ 3, 9 ] }, { "action": "wets", "types": [ "simple" ], "segments": [ { "caption_index": 3, "loom_clip_id": "v_9A0BwCxAKAg_450_1004", "action_id": "action_4", "mentions": [ { "text": "wets", "ranges": [ [ 92, 96 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "interval": { "start_s": 1.0, "end_s": 17.0, "source_start_s": 16.0, "source_end_s": 32.0 }, "confidence": "high", "temporal_status": "supported" }, { "caption_index": 3, "loom_clip_id": "v_9A0BwCxAKAg_450_1004", "action_id": "action_4", "mentions": [ { "text": "wets", "ranges": [ [ 92, 96 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "interval": { "start_s": 18.0, "end_s": 18.5, "source_start_s": 33.0, "source_end_s": 33.5 }, "confidence": "high", "temporal_status": "supported" } ], "id": "action_group_012", "segment_count": 2, "caption_indices": [ 3 ] }, { "action": "bends", "types": [ "simple" ], "segments": [ { "caption_index": 4, "loom_clip_id": "v_9A0BwCxAKAg_1004_1424", "action_id": "action_2", "mentions": [ { "text": "bends", "ranges": [ [ 45, 50 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "interval": { "start_s": 7.0, "end_s": 11.0, "source_start_s": 40.5, "source_end_s": 44.5 }, "confidence": "high", "temporal_status": "localized" }, { "caption_index": 7, "loom_clip_id": "v_9A0BwCxAKAg_3087_4136", "action_id": "action_3", "mentions": [ { "text": "bends", "ranges": [ [ 187, 192 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "interval": { "start_s": 29.0, "end_s": 31.0, "source_start_s": 132.0, "source_end_s": 134.0 }, "confidence": "high", "temporal_status": "localized" } ], "id": "action_group_013", "segment_count": 2, "caption_indices": [ 4, 7 ] }, { "action": "disappearing", "types": [ "simple" ], "segments": [ { "caption_index": 4, "loom_clip_id": "v_9A0BwCxAKAg_1004_1424", "action_id": "action_3", "mentions": [ { "text": "disappearing", "ranges": [ [ 57, 69 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "interval": { "start_s": 8.0, "end_s": 11.0, "source_start_s": 41.5, "source_end_s": 44.5 }, "confidence": "medium", "temporal_status": "localized" } ], "id": "action_group_014", "segment_count": 1, "caption_indices": [ 4 ] }, { "action": "straightens up", "types": [ "phrasal" ], "segments": [ { "caption_index": 4, "loom_clip_id": "v_9A0BwCxAKAg_1004_1424", "action_id": "action_4", "mentions": [ { "text": "straightens up", "ranges": [ [ 107, 118 ], [ 124, 126 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "interval": { "start_s": 10.0, "end_s": 12.0, "source_start_s": 43.5, "source_end_s": 45.5 }, "confidence": "high", "temporal_status": "localized" }, { "caption_index": 8, "loom_clip_id": "v_9A0BwCxAKAg_4136_4900", "action_id": "action_1", "mentions": [ { "text": "straightens up", "ranges": [ [ 3, 17 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "interval": null, "confidence": "high", "temporal_status": "uncertain" }, { "caption_index": 11, "loom_clip_id": "v_9A0BwCxAKAg_6009_6233", "action_id": "action_1", "mentions": [ { "text": "straightens up", "ranges": [ [ 11, 25 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "interval": { "start_s": 0.0, "end_s": 1.0, "source_start_s": 200.49, "source_end_s": 201.49 }, "confidence": "high", "temporal_status": "localized_sparse_support" } ], "id": "action_group_015", "segment_count": 3, "caption_indices": [ 4, 8, 11 ] }, { "action": "looks", "types": [ "prepositional" ], "segments": [ { "caption_index": 4, "loom_clip_id": "v_9A0BwCxAKAg_1004_1424", "action_id": "action_6", "mentions": [ { "text": "looks", "ranges": [ [ 164, 169 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "interval": { "start_s": 11.0, "end_s": 14.0, "source_start_s": 44.5, "source_end_s": 47.5 }, "confidence": "high", "temporal_status": "localized" } ], "id": "action_group_016", "segment_count": 1, "caption_indices": [ 4 ] }, { "action": "uses", "types": [ "simple" ], "segments": [ { "caption_index": 5, "loom_clip_id": "v_9A0BwCxAKAg_1424_2068", "action_id": "a1", "mentions": [ { "text": "uses", "ranges": [ [ 3, 7 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "interval": { "start_s": 1.0, "end_s": 5.0, "source_start_s": 48.5, "source_end_s": 52.5 }, "confidence": "medium", "temporal_status": "localized" }, { "caption_index": 12, "loom_clip_id": "v_9A0BwCxAKAg_6233_6428", "action_id": "action_1", "mentions": [ { "text": "uses", "ranges": [ [ 11, 15 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "interval": { "start_s": 2.0, "end_s": 6.0, "source_start_s": 209.99, "source_end_s": 213.99 }, "confidence": "high", "temporal_status": "localized" } ], "id": "action_group_017", "segment_count": 2, "caption_indices": [ 5, 12 ] }, { "action": "wipe", "types": [ "simple" ], "segments": [ { "caption_index": 5, "loom_clip_id": "v_9A0BwCxAKAg_1424_2068", "action_id": "a2", "mentions": [ { "text": "wipe", "ranges": [ [ 21, 25 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "interval": { "start_s": 1.0, "end_s": 5.0, "source_start_s": 48.5, "source_end_s": 52.5 }, "confidence": "high", "temporal_status": "localized" } ], "id": "action_group_018", "segment_count": 1, "caption_indices": [ 5 ] }, { "action": "puts", "types": [ "phrasal" ], "segments": [ { "caption_index": 5, "loom_clip_id": "v_9A0BwCxAKAg_1424_2068", "action_id": "a3", "mentions": [ { "text": "puts", "ranges": [ [ 79, 83 ] ], "occurrence": 1, "particle": { "text": "down", "occurrence": 1, "ranges": [ [ 94, 98 ] ] } } ], "interval": { "start_s": 5.0, "end_s": 9.0, "source_start_s": 52.5, "source_end_s": 56.5 }, "confidence": "medium", "temporal_status": "localized" }, { "caption_index": 9, "loom_clip_id": "v_9A0BwCxAKAg_4900_5783", "action_id": "action_5", "mentions": [ { "text": "puts", "ranges": [ [ 111, 115 ] ], "occurrence": 1, "particle": { "text": "down", "occurrence": 1, "range": [ 129, 133 ] } } ], "interval": { "start_s": 24.0, "end_s": 26.0, "source_start_s": 187.5, "source_end_s": 189.5 }, "confidence": "medium", "temporal_status": "supported" } ], "id": "action_group_019", "segment_count": 2, "caption_indices": [ 5, 9 ] }, { "action": "wiping", "types": [ "simple" ], "segments": [ { "caption_index": 5, "loom_clip_id": "v_9A0BwCxAKAg_1424_2068", "action_id": "a5", "mentions": [ { "text": "wiping", "ranges": [ [ 138, 144 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "interval": { "start_s": 12.0, "end_s": 21.5, "source_start_s": 59.5, "source_end_s": 69.0 }, "confidence": "high", "temporal_status": "localized" }, { "caption_index": 11, "loom_clip_id": "v_9A0BwCxAKAg_6009_6233", "action_id": "action_7", "mentions": [ { "text": "wiping", "ranges": [ [ 178, 184 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "interval": { "start_s": 6.0, "end_s": 7.5, "source_start_s": 206.49, "source_end_s": 207.99 }, "confidence": "high", "temporal_status": "localized_sparse_support" } ], "id": "action_group_020", "segment_count": 2, "caption_indices": [ 5, 11 ] }, { "action": "hold", "types": [ "simple" ], "segments": [ { "caption_index": 6, "loom_clip_id": "v_9A0BwCxAKAg_2068_3087", "action_id": "action_1", "mentions": [ { "text": "hold", "ranges": [ [ 14, 18 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "interval": { "start_s": 0.0, "end_s": 2.0, "source_start_s": 69.0, "source_end_s": 71.0 }, "confidence": "high", "temporal_status": "localized_sparse" }, { "caption_index": 6, "loom_clip_id": "v_9A0BwCxAKAg_2068_3087", "action_id": "action_1", "mentions": [ { "text": "hold", "ranges": [ [ 14, 18 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "interval": { "start_s": 5.0, "end_s": 11.0, "source_start_s": 74.0, "source_end_s": 80.0 }, "confidence": "high", "temporal_status": "localized_sparse" }, { "caption_index": 6, "loom_clip_id": "v_9A0BwCxAKAg_2068_3087", "action_id": "action_1", "mentions": [ { "text": "hold", "ranges": [ [ 14, 18 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "interval": { "start_s": 22.0, "end_s": 27.0, "source_start_s": 91.0, "source_end_s": 96.0 }, "confidence": "high", "temporal_status": "localized_sparse" }, { "caption_index": 6, "loom_clip_id": "v_9A0BwCxAKAg_2068_3087", "action_id": "action_1", "mentions": [ { "text": "hold", "ranges": [ [ 14, 18 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "interval": { "start_s": 31.0, "end_s": 34.0, "source_start_s": 100.0, "source_end_s": 103.0 }, "confidence": "high", "temporal_status": "localized_sparse" }, { "caption_index": 12, "loom_clip_id": "v_9A0BwCxAKAg_6233_6428", "action_id": "action_4", "mentions": [ { "text": "hold", "ranges": [ [ 100, 104 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "interval": { "start_s": 2.0, "end_s": 6.5, "source_start_s": 209.99, "source_end_s": 214.49 }, "confidence": "high", "temporal_status": "localized" } ], "id": "action_group_021", "segment_count": 5, "caption_indices": [ 6, 12 ] }, { "action": "makes several horizontal passes", "types": [ "light_verb" ], "segments": [ { "caption_index": 6, "loom_clip_id": "v_9A0BwCxAKAg_2068_3087", "action_id": "action_2", "mentions": [ { "text": "makes several horizontal passes", "ranges": [ [ 36, 67 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "interval": { "start_s": 5.0, "end_s": 10.0, "source_start_s": 74.0, "source_end_s": 79.0 }, "confidence": "high", "temporal_status": "localized_sparse" } ], "id": "action_group_022", "segment_count": 1, "caption_indices": [ 6 ] }, { "action": "cleaning", "types": [ "simple" ], "segments": [ { "caption_index": 6, "loom_clip_id": "v_9A0BwCxAKAg_2068_3087", "action_id": "action_3", "mentions": [ { "text": "cleaning", "ranges": [ [ 145, 153 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "interval": { "start_s": 13.0, "end_s": 20.0, "source_start_s": 82.0, "source_end_s": 89.0 }, "confidence": "high", "temporal_status": "localized_sparse" }, { "caption_index": 6, "loom_clip_id": "v_9A0BwCxAKAg_2068_3087", "action_id": "action_3", "mentions": [ { "text": "cleaning", "ranges": [ [ 145, 153 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "interval": { "start_s": 22.0, "end_s": 27.0, "source_start_s": 91.0, "source_end_s": 96.0 }, "confidence": "high", "temporal_status": "localized_sparse" }, { "caption_index": 6, "loom_clip_id": "v_9A0BwCxAKAg_2068_3087", "action_id": "action_3", "mentions": [ { "text": "cleaning", "ranges": [ [ 145, 153 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "interval": { "start_s": 31.0, "end_s": 34.0, "source_start_s": 100.0, "source_end_s": 103.0 }, "confidence": "high", "temporal_status": "localized_sparse" } ], "id": "action_group_023", "segment_count": 3, "caption_indices": [ 6 ] }, { "action": "clean · clear", "types": [ "simple" ], "segments": [ { "caption_index": 7, "loom_clip_id": "v_9A0BwCxAKAg_3087_4136", "action_id": "action_1", "mentions": [ { "text": "clean", "ranges": [ [ 24, 29 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } }, { "text": "clear", "ranges": [ [ 104, 109 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "interval": { "start_s": 0.0, "end_s": 29.0, "source_start_s": 103.0, "source_end_s": 132.0 }, "confidence": "high", "temporal_status": "localized" } ], "id": "action_group_024", "segment_count": 1, "caption_indices": [ 7 ] }, { "action": "using long, overlapping vertical strokes", "types": [ "light_verb" ], "segments": [ { "caption_index": 7, "loom_clip_id": "v_9A0BwCxAKAg_3087_4136", "action_id": "action_2", "mentions": [ { "text": "using long, overlapping vertical strokes", "ranges": [ [ 60, 100 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "interval": { "start_s": 0.0, "end_s": 29.0, "source_start_s": 103.0, "source_end_s": 132.0 }, "confidence": "high", "temporal_status": "localized" } ], "id": "action_group_025", "segment_count": 1, "caption_indices": [ 7 ] }, { "action": "re-wets", "types": [ "simple" ], "segments": [ { "caption_index": 8, "loom_clip_id": "v_9A0BwCxAKAg_4136_4900", "action_id": "action_3", "mentions": [ { "text": "re-wets", "ranges": [ [ 60, 67 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "interval": { "start_s": 2.0, "end_s": 6.0, "source_start_s": 140.0, "source_end_s": 144.0 }, "confidence": "high", "temporal_status": "localized" } ], "id": "action_group_026", "segment_count": 1, "caption_indices": [ 8 ] }, { "action": "switches back", "types": [ "phrasal" ], "segments": [ { "caption_index": 8, "loom_clip_id": "v_9A0BwCxAKAg_4136_4900", "action_id": "action_4", "mentions": [ { "text": "switches back", "ranges": [ [ 107, 120 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "interval": { "start_s": 12.0, "end_s": 14.0, "source_start_s": 150.0, "source_end_s": 152.0 }, "confidence": "high", "temporal_status": "localized" } ], "id": "action_group_027", "segment_count": 1, "caption_indices": [ 8 ] }, { "action": "clean", "types": [ "simple" ], "segments": [ { "caption_index": 8, "loom_clip_id": "v_9A0BwCxAKAg_4136_4900", "action_id": "action_5", "mentions": [ { "text": "clean", "ranges": [ [ 146, 151 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "interval": { "start_s": 13.0, "end_s": 24.0, "source_start_s": 151.0, "source_end_s": 162.0 }, "confidence": "high", "temporal_status": "localized" }, { "caption_index": 9, "loom_clip_id": "v_9A0BwCxAKAg_4900_5783", "action_id": "action_4", "mentions": [ { "text": "clean", "ranges": [ [ 59, 64 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "interval": { "start_s": 4.0, "end_s": 24.0, "source_start_s": 167.5, "source_end_s": 187.5 }, "confidence": "high", "temporal_status": "supported" } ], "id": "action_group_028", "segment_count": 2, "caption_indices": [ 8, 9 ] }, { "action": "gestures", "types": [ "simple" ], "segments": [ { "caption_index": 9, "loom_clip_id": "v_9A0BwCxAKAg_4900_5783", "action_id": "action_2", "mentions": [ { "text": "gestures", "ranges": [ [ 21, 29 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "interval": { "start_s": 3.0, "end_s": 4.0, "source_start_s": 166.5, "source_end_s": 167.5 }, "confidence": "medium", "temporal_status": "supported" } ], "id": "action_group_029", "segment_count": 1, "caption_indices": [ 9 ] }, { "action": "bends over", "types": [ "phrasal" ], "segments": [ { "caption_index": 9, "loom_clip_id": "v_9A0BwCxAKAg_4900_5783", "action_id": "action_6", "mentions": [ { "text": "bends over", "ranges": [ [ 135, 145 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "interval": { "start_s": 25.0, "end_s": 29.0, "source_start_s": 188.5, "source_end_s": 192.5 }, "confidence": "high", "temporal_status": "supported" } ], "id": "action_group_030", "segment_count": 1, "caption_indices": [ 9 ] }, { "action": "moves", "types": [ "simple" ], "segments": [ { "caption_index": 9, "loom_clip_id": "v_9A0BwCxAKAg_4900_5783", "action_id": "action_7", "mentions": [ { "text": "moves", "ranges": [ [ 151, 156 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "interval": { "start_s": 29.0, "end_s": 29.460000000000008, "source_start_s": 192.5, "source_end_s": 192.96 }, "confidence": "high", "temporal_status": "supported_to_clip_boundary" } ], "id": "action_group_031", "segment_count": 1, "caption_indices": [ 9 ] }, { "action": "approaches", "types": [ "simple" ], "segments": [ { "caption_index": 10, "loom_clip_id": "v_9A0BwCxAKAg_5814_6009", "action_id": "action_2", "mentions": [ { "text": "approaches", "ranges": [ [ 87, 97 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "interval": { "start_s": 1.0, "end_s": 4.0, "source_start_s": 194.99, "source_end_s": 197.99 }, "confidence": "high", "temporal_status": "supported" } ], "id": "action_group_032", "segment_count": 1, "caption_indices": [ 10 ] }, { "action": "enters", "types": [ "simple" ], "segments": [ { "caption_index": 10, "loom_clip_id": "v_9A0BwCxAKAg_5814_6009", "action_id": "action_1", "mentions": [ { "text": "enters", "ranges": [ [ 62, 68 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "interval": { "start_s": 1.0, "end_s": 2.0, "source_start_s": 194.99, "source_end_s": 195.99 }, "confidence": "high", "temporal_status": "supported" } ], "id": "action_group_033", "segment_count": 1, "caption_indices": [ 10 ] }, { "action": "use", "types": [ "simple" ], "segments": [ { "caption_index": 10, "loom_clip_id": "v_9A0BwCxAKAg_5814_6009", "action_id": "action_4", "mentions": [ { "text": "use", "ranges": [ [ 133, 136 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "interval": { "start_s": 1.0, "end_s": 4.0, "source_start_s": 194.99, "source_end_s": 197.99 }, "confidence": "medium", "temporal_status": "supported" }, { "caption_index": 13, "loom_clip_id": "v_9A0BwCxAKAg_6428_6653", "action_id": "use_cloth", "mentions": [ { "text": "use", "ranges": [ [ 80, 83 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "interval": { "start_s": 3.0, "end_s": 5.0, "source_start_s": 217.49, "source_end_s": 219.49 }, "confidence": "medium", "temporal_status": "localized" } ], "id": "action_group_034", "segment_count": 2, "caption_indices": [ 10, 13 ] }, { "action": "work", "types": [ "prepositional" ], "segments": [ { "caption_index": 10, "loom_clip_id": "v_9A0BwCxAKAg_5814_6009", "action_id": "action_5", "mentions": [ { "text": "work", "ranges": [ [ 157, 161 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "interval": { "start_s": 1.0, "end_s": 4.0, "source_start_s": 194.99, "source_end_s": 197.99 }, "confidence": "medium", "temporal_status": "supported" } ], "id": "action_group_035", "segment_count": 1, "caption_indices": [ 10 ] }, { "action": "bend over", "types": [ "phrasal" ], "segments": [ { "caption_index": 10, "loom_clip_id": "v_9A0BwCxAKAg_5814_6009", "action_id": "action_3", "mentions": [ { "text": "bend over", "ranges": [ [ 119, 128 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "interval": { "start_s": 4.0, "end_s": 6.0, "source_start_s": 197.99, "source_end_s": 199.99 }, "confidence": "high", "temporal_status": "supported" } ], "id": "action_group_036", "segment_count": 1, "caption_indices": [ 10 ] }, { "action": "remain", "types": [ "posture" ], "segments": [ { "caption_index": 10, "loom_clip_id": "v_9A0BwCxAKAg_5814_6009", "action_id": "action_6", "mentions": [ { "text": "remain", "ranges": [ [ 203, 209 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "interval": { "start_s": 5.0, "end_s": 6.5, "source_start_s": 198.99, "source_end_s": 200.49 }, "confidence": "high", "temporal_status": "supported" } ], "id": "action_group_037", "segment_count": 1, "caption_indices": [ 10 ] }, { "action": "holding", "types": [ "simple" ], "segments": [ { "caption_index": 4, "loom_clip_id": "v_9A0BwCxAKAg_1004_1424", "action_id": "action_5", "mentions": [ { "text": "holding", "ranges": [ [ 127, 134 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "interval": null, "confidence": "high", "temporal_status": "uncertain" }, { "caption_index": 11, "loom_clip_id": "v_9A0BwCxAKAg_6009_6233", "action_id": "action_6", "mentions": [ { "text": "holding", "ranges": [ [ 149, 156 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "interval": { "start_s": 2.0, "end_s": 7.5, "source_start_s": 202.49, "source_end_s": 207.99 }, "confidence": "high", "temporal_status": "localized_sparse_support" } ], "id": "action_group_038", "segment_count": 2, "caption_indices": [ 4, 11 ] }, { "action": "unfold", "types": [ "simple" ], "segments": [ { "caption_index": 11, "loom_clip_id": "v_9A0BwCxAKAg_6009_6233", "action_id": "action_3", "mentions": [ { "text": "unfold", "ranges": [ [ 74, 80 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "interval": { "start_s": 2.0, "end_s": 5.0, "source_start_s": 202.49, "source_end_s": 205.49 }, "confidence": "high", "temporal_status": "localized_sparse_support" } ], "id": "action_group_039", "segment_count": 1, "caption_indices": [ 11 ] }, { "action": "examine", "types": [ "simple" ], "segments": [ { "caption_index": 11, "loom_clip_id": "v_9A0BwCxAKAg_6009_6233", "action_id": "action_4", "mentions": [ { "text": "examine", "ranges": [ [ 95, 102 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "interval": { "start_s": 4.0, "end_s": 6.0, "source_start_s": 204.49, "source_end_s": 206.49 }, "confidence": "high", "temporal_status": "localized_sparse_support" } ], "id": "action_group_040", "segment_count": 1, "caption_indices": [ 11 ] }, { "action": "raise", "types": [ "simple" ], "segments": [ { "caption_index": 11, "loom_clip_id": "v_9A0BwCxAKAg_6009_6233", "action_id": "action_5", "mentions": [ { "text": "raise", "ranges": [ [ 125, 130 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "interval": { "start_s": 5.0, "end_s": 7.0, "source_start_s": 205.49, "source_end_s": 207.49 }, "confidence": "high", "temporal_status": "localized_sparse_support" } ], "id": "action_group_041", "segment_count": 1, "caption_indices": [ 11 ] }, { "action": "wipe · wiping", "types": [ "simple" ], "segments": [ { "caption_index": 12, "loom_clip_id": "v_9A0BwCxAKAg_6233_6428", "action_id": "action_2", "mentions": [ { "text": "wipe", "ranges": [ [ 29, 33 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } }, { "text": "wiping", "ranges": [ [ 174, 180 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "interval": { "start_s": 2.0, "end_s": 6.0, "source_start_s": 209.99, "source_end_s": 213.99 }, "confidence": "high", "temporal_status": "localized" } ], "id": "action_group_042", "segment_count": 1, "caption_indices": [ 12 ] }, { "action": "lowers", "types": [ "simple" ], "segments": [ { "caption_index": 13, "loom_clip_id": "v_9A0BwCxAKAg_6428_6653", "action_id": "lower_arms", "mentions": [ { "text": "lowers", "ranges": [ [ 41, 47 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "interval": { "start_s": 0.0, "end_s": 2.0, "source_start_s": 214.49, "source_end_s": 216.49 }, "confidence": "high", "temporal_status": "localized" } ], "id": "action_group_043", "segment_count": 1, "caption_indices": [ 13 ] }, { "action": "bends down · bent over", "types": [ "posture" ], "segments": [ { "caption_index": 13, "loom_clip_id": "v_9A0BwCxAKAg_6428_6653", "action_id": "bend_down_and_remain_bent", "mentions": [ { "text": "bends down", "ranges": [ [ 63, 73 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } }, { "text": "bent over", "ranges": [ [ 184, 193 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "interval": { "start_s": 1.0, "end_s": 7.5, "source_start_s": 215.49, "source_end_s": 221.99 }, "confidence": "high", "temporal_status": "localized" } ], "id": "action_group_044", "segment_count": 1, "caption_indices": [ 13 ] }, { "action": "wipe · clean", "types": [ "simple" ], "segments": [ { "caption_index": 13, "loom_clip_id": "v_9A0BwCxAKAg_6428_6653", "action_id": "wipe_lower_frame", "mentions": [ { "text": "wipe", "ranges": [ [ 117, 121 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } }, { "text": "clean", "ranges": [ [ 202, 207 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "interval": { "start_s": 3.0, "end_s": 5.0, "source_start_s": 217.49, "source_end_s": 219.49 }, "confidence": "high", "temporal_status": "localized" } ], "id": "action_group_045", "segment_count": 1, "caption_indices": [ 13 ] }, { "action": "observes · look", "types": [ "other" ], "segments": [ { "caption_index": 14, "loom_clip_id": "v_9A0BwCxAKAg_6653_6918", "action_id": "action_2", "mentions": [ { "text": "observes", "ranges": [ [ 34, 42 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } }, { "text": "look", "ranges": [ [ 177, 181 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "interval": { "start_s": 0.0, "end_s": 4.0, "source_start_s": 221.99, "source_end_s": 225.99 }, "confidence": "medium", "temporal_status": "localized" }, { "caption_index": 14, "loom_clip_id": "v_9A0BwCxAKAg_6653_6918", "action_id": "action_2", "mentions": [ { "text": "observes", "ranges": [ [ 34, 42 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } }, { "text": "look", "ranges": [ [ 177, 181 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "interval": { "start_s": 6.0, "end_s": 8.840000000000003, "source_start_s": 227.99, "source_end_s": 230.83 }, "confidence": "medium", "temporal_status": "localized" } ], "id": "action_group_046", "segment_count": 2, "caption_indices": [ 14 ] }, { "action": "stands up", "types": [ "phrasal" ], "segments": [ { "caption_index": 14, "loom_clip_id": "v_9A0BwCxAKAg_6653_6918", "action_id": "action_1", "mentions": [ { "text": "stands up", "ranges": [ [ 11, 20 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "interval": { "start_s": 0.0, "end_s": 2.0, "source_start_s": 221.99, "source_end_s": 223.99 }, "confidence": "medium", "temporal_status": "localized" } ], "id": "action_group_047", "segment_count": 1, "caption_indices": [ 14 ] }, { "action": "turn", "types": [ "simple" ], "segments": [ { "caption_index": 14, "loom_clip_id": "v_9A0BwCxAKAg_6653_6918", "action_id": "action_3", "mentions": [ { "text": "turn", "ranges": [ [ 74, 78 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "interval": { "start_s": 4.0, "end_s": 6.0, "source_start_s": 225.99, "source_end_s": 227.99 }, "confidence": "high", "temporal_status": "localized" } ], "id": "action_group_048", "segment_count": 1, "caption_indices": [ 14 ] }, { "action": "speak", "types": [ "simple" ], "segments": [ { "caption_index": 14, "loom_clip_id": "v_9A0BwCxAKAg_6653_6918", "action_id": "action_4", "mentions": [ { "text": "speak", "ranges": [ [ 116, 121 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "interval": { "start_s": 5.0, "end_s": 6.0, "source_start_s": 226.99, "source_end_s": 227.99 }, "confidence": "medium", "temporal_status": "localized" } ], "id": "action_group_049", "segment_count": 1, "caption_indices": [ 14 ] }, { "action": "turn back", "types": [ "phrasal" ], "segments": [ { "caption_index": 14, "loom_clip_id": "v_9A0BwCxAKAg_6653_6918", "action_id": "action_5", "mentions": [ { "text": "turn back", "ranges": [ [ 133, 142 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "interval": { "start_s": 5.0, "end_s": 7.0, "source_start_s": 226.99, "source_end_s": 228.99 }, "confidence": "high", "temporal_status": "localized" } ], "id": "action_group_050", "segment_count": 1, "caption_indices": [ 14 ] }, { "action": "sets down", "types": [ "phrasal" ], "segments": [ { "caption_index": 4, "loom_clip_id": "v_9A0BwCxAKAg_1004_1424", "action_id": "action_1", "mentions": [ { "text": "sets down", "ranges": [ [ 11, 20 ] ], "occurrence": 1, "particle": { "text": null, "occurrence": 1 } } ], "interval": null, "confidence": "high", "temporal_status": "uncertain" } ], "id": "action_group_051", "segment_count": 1, "caption_indices": [ 4 ] } ] } ], "action_grouping_rule": "Within each source video, combine action records when their exact mention phrase set matches after case/whitespace normalization. Action types are retained as metadata but do not split groups. Preserve each interval as a separate segment with caption provenance; no lemmatization or semantic merging." }