WUJI open-sources MINT: 3D hand trajectory and camera pose reconstruction from egocentric video

chris_j_paxton · x · 2026-09-12

Robotics company WUJI has open-sourced MINT, a system that reconstructs 3D camera trajectories and bimanual hand motion in a shared world coordinate system from first-person RGB video, with retargeting to the WUJI Hand. The release includes model weights, training/inference code, visualization and evaluation tools, EgoPipeline reference code, and structured annotations for 560,649 episodes spanning 1,021 hours of egocentric video — valuable for teaching robots manipulation skills from human videos.

Original post →

More from Embodied

Embodied channel →