Subnets
/
SN/06
RoboCap
Video Dataset
Subnet active
Subnet overview
mission details
Summary
The RoboCap subnet collects large-scale egocentric data from everyday human activities. Contributors wear a multi-camera capture system to record natural interactions in real-world environments, generating diverse training data for Physical AI.
The network has collected 100,000+ hours of egocentric data to date. RoboWrist support adds detailed hand tracking today, with additional tactile sensing capabilities and data enrichment coming in future releases.
robotic embodiment
RoboCap
output
Video Dataset
robots from
Frodobots Lab (RoboCap)
owner
Frodobots Lab
Open to
Teleoperators
Contributors are encouraged to engage with this subnet by clicking the link below. Contribute to earn rewards.
