Haotian Yang

Latest

• Vmoba: Mixture-of-block attention for video diffusion models