macrodatacc-by-nc-sa-4.0

WGO-Bench

A manually annotated benchmark for evaluating vision-language models on temporal segmentation of robot and egocentric manipulation videos into timestamped subtask annotations. Contains 100 episodes with 743 gold subtask segments across three source families.

Downloads887
Likes11

Technical Profile

Modalities
rgblanguage
Robot Embodiments
mobile_manipulatorhumanoidquadruped
Environment
lab
Task Types
manipulationgraspingpouringpick_and_place
Data Format
Parquet
License
cc-by-nc-sa-4.0
Part of the WGO-Bench family

Community Signals

Top 50% by downloads

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