1 paper
Lechao Cheng, Jiahui Li, Jingxuan He +5
Reconstructing dynamic 3D scenes from monocular video remains fundamentally challenging due to the need to jointly infer motion, structure, and appearance from limited observations…