AI RESEARCH

R3D: Revisiting 3D Policy Learning

arXiv CS.CV

ArXi:2604.15281v1 Announce Type: new 3D policy learning promises superior generalization and cross-embodiment transfer, but progress has been hindered by