AI RESEARCH

WristCompass: Kinematic Coupling as a Learnable Visual Concept for Ego-Camera Orientation

arXiv CS.CV

ArXi:2605.30671v1 Announce Type: new Recovering ego-camera orientation from manipulation video is a prerequisite for disentangling hand motion from camera motion, a key step in imitation learning from egocentric nstrations. The obvious approach, inferring orientation from scene geometry, fails when hands occlude the frame: VGGT, a 1B-parameter scene reconstruction model, scores worse than a constant predictor on the TACO benchmark.