React / tasks.json
yxma's picture
Docs: mark tactile latency CORRECTED in-data (no loader compensation needed)
e63e43b verified
Raw
History Blame Contribute Delete
3.61 kB
{
"dataset": "React",
"format": "video (LeRobot-style: per-camera MP4 + per-episode parquet; depth as FFV1 16-bit MKV)",
"resolution": "640x480",
"fps": 30,
"video_streams": [
"view_left",
"view_middle",
"view_right",
"tactile_left",
"tactile_right"
],
"depth_streams": [
"depth_left",
"depth_middle",
"depth_right"
],
"parquet_columns": [
"frame_idx",
"timestamp",
"sensor_left_pose",
"sensor_right_pose",
"object_pose",
"tactile_{L,R}_{intensity,area,mixed}",
"source_h5_frame"
],
"decoded_color": "RGB",
"tasks": {
"motherboard": {
"n_episodes": 32,
"dates": [
"2026-05-10",
"2026-05-11",
"2026-05-19"
],
"n_frames": 194445,
"duration_min": 108.0,
"n_segments": 76,
"clean_min": 107.04,
"bad_fraction": 0.009,
"active_sensors": [
"left",
"right"
],
"calibration_id": "may-12",
"calibration_created": "2026-05-12",
"calibration_rmse_unit": "mm",
"object_tracked_episodes": 32,
"depth_available_episodes": 32,
"depth_units": "mm",
"depth_invalid_value": 0,
"gelsight_left_serial": "2BGLKZNT/2DUPB53G",
"note": "Bimanual handheld tactile-visual interaction. Object pose (the board) tracked. 05-19 has a redefined OptiTrack world origin; offset (0.23,0,0.175)m baked into poses."
},
"pushT": {
"n_episodes": 4,
"dates": [
"2026-06-18"
],
"n_frames": 45595,
"duration_min": 25.3,
"n_segments": 17,
"clean_min": 25.16,
"bad_fraction": 0.0067,
"active_sensors": [
"left",
"right"
],
"calibration_id": "june-26",
"calibration_created": "2026-06-26",
"calibration_rmse_unit": "px",
"object_tracked_episodes": 0,
"depth_available_episodes": 4,
"depth_units": "mm",
"depth_invalid_value": 0,
"gelsight_left_serial": "2DUPB53G",
"note": "Push-T manipulation. Recalibrated cameras (June-26). Object rigid body was not tracked (object_pose = NaN). episode_004 source H5 corrupt, excluded."
}
},
"download": {
"core_size_gb": 4.4,
"depth_size_gb": 33,
"total_size_gb": 37.4,
"depth_optional": true,
"depth_note": "Depth lives in data/<task>/depth/ as a separate subtree. Skip it with snapshot_download(ignore_patterns=[\"*/depth/*\"]) or examples/download.py --no-depth. The ReactVideoDataset loader ignores depth unless load_depth=True.",
"helpers": {
"core_only": "python examples/download.py --no-depth",
"with_depth": "python examples/download.py",
"middle_depth_only": "python examples/download.py --depth-cams middle"
}
},
"tactile_latency": {
"frames_estimate": 15,
"fps": 30,
"seconds_estimate": 0.5,
"applies_to": "all recordings up to and including 2026-06-18",
"fixed_in_rig": "2026-06-27",
"cause": "recording-side cv2.VideoCapture V4L2 buffer never flushed (throttled reads + no BUFFERSIZE=1 + default pixfmt)",
"compensation": "ReactVideoDataset(tactile_latency=15) pairs view[i] with tactile[i+15]; shifts tactile videos + contact scalars only",
"note": "Recordings up to 2026-06-18 had a ~15-frame GelSight buffer lag; it has been baked out of the published tactile videos + contact scalars. No loader compensation needed. Rig fixed 2026-06-27.",
"status": "CORRECTED_IN_DATA",
"frames_shifted": 15,
"method": "tactile streams shifted +15 frames (rebuilt from raw H5) so tactile[i] aligns with view[i]/pose[i]"
}
}