Hannibal52Barcaapache-2.0
Demo Joint Space
A bimanual teleoperation dataset derived from the ICL demo dataset, reconstructed with joint-space targets computed via inverse kinematics from recorded Cartesian end-effector poses.
Downloads0
Episodes285
Why This Matters for Physical AI
This dataset provides high-quality joint-space action reconstruction for bimanual manipulation learning, enabling imitation learning models to learn from teleoperated demonstrations with corrected kinematic targets and repaired orientation glitches.
Technical Profile
- Modalities
- proprioceptionjoint_positions
- Robot Embodiments
- bimanual_manipulator
- Action Space
- joint_positions
- Environment
- lab
- Task Types
- manipulationteleoperation
- Episodes
- 285
- Data Format
- LeRobot
- Annotation Types
- action_labelsepisode_flags
- License
- apache-2.0
Access
Need custom proprioception data?
Claru builds purpose-built datasets for lab applications with dense human annotations and quality assurance.
Request a Sample Pack