SM4RT: Learning Structured Motion Geometry for 4D Reconstruction

By Shing Ho J. Lin · Paper · cs.CV

Geometry Foundation Models (GFMs) have substantially advanced monocular 3D reconstruction, yet extending this capability to 4D dynamic understanding remains a fundamental challenge. Most existing motion perception methods (e.g., sparse tracking, dense point-wise flow) treat motio

Cs.cv

View original

HomeResourceLoading…