xuejf2026

EgoSuite-Open1K

A large-scale egocentric dataset containing 10,000 hours of real-world human activity across seven scene families, organized into six composable research SKUs with progressive supervision layers including pose and semantic annotations.

Downloads0
Hours10000

Why This Matters for Physical AI

EgoSuite-Open1K provides large-scale egocentric video with structured pose and semantic supervision across real-world domestic and professional environments, enabling training of embodied AI systems for natural human-object interaction understanding and imitation learning.

Technical Profile

Modalities
rgbhand_posebody_posesemantic_annotation
Environment
homehospitalityretailsportslogisticsofficeindustry
Task Types
manipulationdaily_routinesservice_flowspickingcheckouttrainingsortingpackingdesktop_workassembly
Total Hours
10000
Data Format
WebDataset
Annotation Types
hand_posebody_posesemantic_labelslanguage_instructions
Part of the EgoSuite-Open1K family

Access

Need custom rgb data?

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

Request a Sample Pack

Related Datasets