MIT-Media-Lab2021cc-by-nc-4.0

DexYCB VITRA Hybrid Streaming Release

A completed VITRA Stage-1 conversion of 1,000 physical DexYCB sequences containing fixed-camera, allocentric human hand-object interaction data with multi-view RGB video, hand pose annotations, and object trajectory labels.

Downloads328
Episodes1000

Why This Matters for Physical AI

This dataset provides large-scale, precisely annotated human hand-object interaction data with calibrated multi-view RGB and hand pose supervision, serving as valuable supplemental training signal for vision-language-action models in dexterous manipulation research.

Technical Profile

Modalities
rgbhand_posedepthsegmentationcamera_intrinsicscamera_extrinsics
Environment
lab
Task Types
hand_pose_estimationobject_manipulationpick_and_place
Episodes
1000
Data Format
NumPy
Annotation Types
hand_pose_keypointsMANO_parametersobject_trajectorylanguage_instructionscamera_calibrationsegmentation_masks
License
cc-by-nc-4.0
Part of the DexYCB family

Access

Need custom rgb data?

Claru builds purpose-built datasets for lab applications with dense human annotations and quality assurance.

Request a Sample Pack

Related Datasets