# Episode 0 of basematrix/ego-binocular-v1

- format: LeRobot v3.0
- robot: orbbec_ego_binocular
- robot family, as the index files it: Other robots
- kind, as the index files it: recording; other (inferred from the name)
- frames per second: 30
- frames in this episode: 1,144
- task: organizing snacks
- episodes in the dataset: 14
- licence: cc-by-nc-4.0
- dataset page: https://huggingface.co/datasets/basematrix/ego-binocular-v1

LeRobot v3 keeps many episodes in one file. Download the data file and keep the rows whose episode_index is 0. Each video file holds several episodes back to back: cut it to episode_seconds. Or load it with LeRobot: LeRobotDataset('basematrix/ego-binocular-v1', episodes=[0]). The data file is 107 MB and holds many episodes; data_via_index gives only this one's rows.

Columns of the data file:

- `timestamp`: float32, shape 1
- `observation.state.camera_pose`: float32, shape 7 (tx, ty, tz, qx, qy, qz, qw)
- `observation.state.camera_pose_valid`: int64, shape 1 (valid)
- `observation.state.left_wrist`: float32, shape 3 (x, y, z)
- `observation.state.right_wrist`: float32, shape 3 (x, y, z)
- `observation.state.left_keypoints_3d`: float32, shape 63
- `observation.state.right_keypoints_3d`: float32, shape 63
- `observation.state.left_contact`: int64, shape 1
- `observation.state.right_contact`: int64, shape 1
- `action.left_wrist_delta`: float32, shape 3 (dx, dy, dz)
- `action.right_wrist_delta`: float32, shape 3 (dx, dy, dz)
- `action.left_grasp`: int64, shape 1
- `action.right_grasp`: int64, shape 1
- `observation.state.left_force`: float32, shape 3 (fx, fy, fz)
- `observation.state.right_force`: float32, shape 3 (fx, fy, fz)
- `observation.state.left_tactile`: float32, shape 48
- `observation.state.right_tactile`: float32, shape 48
- `observation.state.gripper_force`: float32, shape 2 (left_n, right_n)
- `action.left_gripper`: float32, shape 7 (cx, cy, cz, dx, dy, dz, openness)
- `action.right_gripper`: float32, shape 7 (cx, cy, cz, dx, dy, dz, openness)
- `observation.state.left_mano_rot`: float32, shape 3 (rx, ry, rz)
- `observation.state.right_mano_rot`: float32, shape 3 (rx, ry, rz)
- `observation.state.left_mano_pose`: float32, shape 45
- `observation.state.right_mano_pose`: float32, shape 45
- `observation.state.left_mano_trans`: float32, shape 3 (tx, ty, tz)
- `observation.state.right_mano_trans`: float32, shape 3 (tx, ty, tz)
- `observation.state.left_mano_betas`: float32, shape 10
- `observation.state.right_mano_betas`: float32, shape 10

| file | camera | size | this episode | link |
|---|---|---|---|---|
| data |  | 107.6 MB | rows where episode_index == 0 (global frames 0 to 1143, 1144 frames) | https://huggingface.co/datasets/basematrix/ego-binocular-v1/resolve/main/data/chunk-000/file-000.parquet |
| video | observation.images.ego_left | 59.7 MB | seconds 0.0 to 38.133 | https://huggingface.co/datasets/basematrix/ego-binocular-v1/resolve/main/videos/observation.images.ego_left/chunk-000/file-000.mp4 |
| video | observation.images.ego_right | 65.5 MB | seconds 0.0 to 38.133 | https://huggingface.co/datasets/basematrix/ego-binocular-v1/resolve/main/videos/observation.images.ego_right/chunk-000/file-000.mp4 |

If your network cannot reach Hugging Face, this site fetches the episode's rows for you, and only that episode's: https://datasets.gurasees.com/episode-data/basematrix/ego-binocular-v1?episode=0 (parquet) or https://datasets.gurasees.com/episode-data/basematrix/ego-binocular-v1?episode=0&format=json (JSON). Videos are not relayed.

- next episode: https://datasets.gurasees.com/episodes/basematrix/ego-binocular-v1.md?episode=1
- last episode: https://datasets.gurasees.com/episodes/basematrix/ego-binocular-v1.md?episode=13