UnifoLM_WBT_Dataset
Collection
55 items • Updated • 96
Codebase v3.0. All cameras are AV1-encoded, 480×640×3, 30 fps.
observation.images.*)
| Key | Description |
|---|---|
head_stereo_left / head_stereo_right |
Head stereo cameras |
wrist_left / wrist_right |
Wrist cameras |
observation.state.*)
| Key | Dim | Description |
|---|---|---|
lower_body |
15 | Both legs (hip×3, knee, ankle×2) + waist (yaw, roll, pitch) |
left_arm / right_arm |
7 | shoulder×3, elbow, wrist×3 |
left_leg / right_leg |
6 | hip×3, knee, ankle×2 |
waist_state_joint |
3 | yaw, roll, pitch |
hand_state |
2 | Hand state, 1 per hand (see Hand Convention) |
left_gripper / right_gripper |
1 | Gripper position (see Hand Convention) |
left_ee_pose_gripper_base / right_ee_pose_gripper_base |
6 | EE pose (x,y,z,roll,pitch,yaw) in base frame |
left_ee_pose_gripper_torso / right_ee_pose_gripper_torso |
6 | EE pose in torso frame |
recomputed_ee_valid |
1 | 1 if recomputed EE poses are valid |
state_torso |
6 | Torso pose (x,y,z,roll,pitch,yaw) |
state_d435 |
6 | D435 camera pose |
state_base_pose |
7 | Base pose (x,y,z,qx,qy,qz,qw) |
state_base_rot |
6 | Gravity vector + angular velocity (gx,gy,gz,wx,wy,wz) |
state_base_rot_quat |
4 | Base orientation (qx,qy,qz,qw) |
action.*)
| Key | Dim | Description |
|---|---|---|
lower_body |
15 | Both legs + waist |
left_arm / right_arm |
7 | shoulder×3, elbow, wrist×3 |
left_leg / right_leg |
6 | hip×3, knee, ankle×2 |
waist_action_joint |
3 | yaw, roll, pitch |
hand_cmd |
2 | Hand commands, 1 per hand (see Hand Convention) |
left_gripper / right_gripper |
1 | Gripper command (see Hand Convention) |
left_ee_pose_gripper_base / right_ee_pose_gripper_base |
6 | Target EE pose in base frame |
left_ee_pose_gripper_torso / right_ee_pose_gripper_torso |
6 | Target EE pose in torso frame |
action_torso |
6 | Target torso pose |
action_d435 |
6 | Target D435 pose |
action_base_pose |
7 | Target base pose (x,y,z,qx,qy,qz,qw) |
action_base_rot |
4 | Target base orientation (qx,qy,qz,qw) |
base_command |
4 | vx, vy, angle_z, height |
timestamp, frame_index, episode_index, index, task_index.
This dataset uses Dex1 grippers (left_gripper / right_gripper, 1 channel
each), mirrored in hand_state / hand_cmd (idx 0 = left, idx 1 = right).
This dataset was created using the LeRobot framework.