1 paper
Tao Feng, Xianbing Zhao, Zhenhua Chen +4
Recent advances in diffusion-based and autoregressive video generation models have achieved remarkable visual realism. However, these models typically lack accurate physical alignm…