| { | |
| "schema_version": 1, | |
| "source": "vendor annotation.json (schema 2, revision 1)", | |
| "time_reference": "start_s / end_s: seconds from the start of the segment video, [start, end); t_start / t_end: the same bounds on the session time base (meta.json time_base, t = 0 at the first solved Body0 frame) = start_s + video_t_start", | |
| "primitive": "LM43", | |
| "instruction": "Bring the boxes over one at a time and stack them on top of each other, walking back for each one. Keep going until the operator says stop; one object at a time, and do not rush between trips.", | |
| "segments": [ | |
| { | |
| "id": "segment1", | |
| "duration_s": 128.966633, | |
| "video": "left_eye", | |
| "video_t_start": 0.03937667034224268 | |
| } | |
| ], | |
| "episodes": [ | |
| { | |
| "id": "5b0113f7-5948-4764-a01c-4389f38ccacf", | |
| "segment": "segment1", | |
| "action": "LM43", | |
| "quality": "valid", | |
| "start_s": 0, | |
| "end_s": 123.594, | |
| "t_start": 0.0394, | |
| "t_end": 123.6334, | |
| "instruction": "Bring the boxes over one at a time and stack them on top of each other, walking back for each one. Keep going until the operator says stop; one object at a time, and do not rush between trips." | |
| } | |
| ] | |
| } | |