huzheyuanapache-2.0
3lego whole-episode sessions: round4
Dual-arm YAM robot block-sorting dataset with whole-episode sessions reconverted for in-context learning, featuring 18 episodes of lego block manipulation with multi-camera RGB and full kinematic state/action data.
Downloads0
Episodes18
Hours1.00
Why This Matters for Physical AI
This dataset enables in-context learning research by preserving whole task sessions with latent permutation structure, providing multi-view RGB and full dual-arm kinematics for learning manipulation from teleoperated demonstrations.
Technical Profile
- Modalities
- rgbproprioception
- Robot Embodiments
- Dual-YAM
- Action Space
- end_effector_delta
- Environment
- lab
- Task Types
- manipulationpick_and_placesorting
- Episodes
- 18
- Total Hours
- 1.00
- Data Format
- LeRobot
- Annotation Types
- language_instructionsaction_labels
- License
- apache-2.0
Access
Need custom rgb data?
Claru builds purpose-built datasets for lab applications with dense human annotations and quality assurance.
Request a Sample Pack