1 paper
Haoran Lu, Shang Wu, Songling Liu +9
Recent video diffusion models have achieved impressive capabilities as large-scale generative world models. However, these models often struggle with fine-grained physical consiste…