LearningFromVideoapache-2.0
LearningFromVideo rollout_eval-push-milk-right
An evaluation rollout dataset for a push-milk-right task collected using LeRobot, containing 25 episodes of robot manipulation with multi-view RGB observations and end-effector state/action data.
Downloads9
Episodes25
Why This Matters for Physical AI
This evaluation dataset demonstrates learning from video for robotic manipulation, providing multi-view observations and precise end-effector control data critical for training and benchmarking vision-based manipulation policies.
Technical Profile
- Modalities
- rgbproprioception
- Robot Embodiments
- dk1_follower
- Action Space
- end_effector_delta
- Environment
- lab
- Task Types
- manipulationpush
- Episodes
- 25
- Data Format
- LeRobot
- 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