OmniEgoCap Turns Any Egocentric Device into a Full-Body 3D Mocap System
Scobleizer · x · 2026-10-26
Seoul National University researchers present OmniEgoCap at NeurIPS 2026: from just head trajectory plus whichever hands the device sees, it reconstructs full-body 3D motion with a single body shape — no dedicated mocap suit needed. The model is device-agnostic, working across Aria, GoPro, Quest 3, and iPhone without retraining, and the team releases OmniEgoDB, the first mocap benchmark spanning multiple consumer egocentric devices, along with code, checkpoints, and an interactive 3D demo. The broader thesis: if egocentric wearables become everyday items that double as full-body mocap systems, daily life becomes a scalable source of 3D human motion data for embodied AI and robot learning.
More from Embodied
- CoRL paper: flow-matching VLAs can run 1-step Euler inference with no retraining — kaixhin · 2026-10-26
- Robotic dexterity should be judged on speed, force control and tool use, not copying humans — DJiafei · 2026-10-26
- Shenzhen company combines humanoid robots with industrial robot arms — RoboBalaji · 2026-10-26
- Realbotix, a sex-doll maker, unveils Echo humanoid with lifelike face — CyberRobooo · 2026-10-26
- Track bolt robot tightens railway fasteners in 6–7 seconds with sub-mm precision — RoboBalaji · 2026-10-26
- Tesla FSD (Supervised) Approved in Latvia, Its 10th European Country — XFreeze · 2026-10-26