mulligan2026mit
real-square-d2-c01-dagger-mulligan
A robotics dataset from the Mulligan project containing real robot demonstrations for the square-d2 task collected from multiple camera viewpoints. Part of a larger collection used for training and evaluating robotic manipulation policies.
Downloads109
Episodes50
Why This Matters for Physical AI
This dataset contributes to physical AI research by providing real robot demonstration data with multi-view camera observations for training and evaluating robotic manipulation policies through imitation learning approaches.
Technical Profile
- Modalities
- rgb
- Environment
- lab
- Task Types
- manipulation
- Episodes
- 50
- Data Format
- parquet
- Annotation Types
- language_instructionsreward_labels
- License
- mit
Access
Need custom rgb data?
Claru builds purpose-built datasets for lab applications with dense human annotations and quality assurance.
Request a Sample Pack