Preprint
Jul 2026
GeoProp: Grounding Robot State in Vision for Generalist Manipulation
Proprioception is fundamental to robotic manipulation, yet standard fusion methods often treat it as an isolated vector lacking explicit alignment with visual tokens. Without a direct correspondence between 3D kinematics and 2D feature maps, manipulation policies struggle to ground the robot's state within the scene, f...
Guoyang Zhao, Quanhao Qian, Gongjie Zhang et al.
· 0 citations