1 paper
Tianxiao Chen, Hanmo Chen, Huajin Chen +3
Modern video generation models can synthesize visually compelling and temporally coherent clips, yet controlling their physical behavior remains difficult with standard text and im…