HunyuanVideo-Avatar

A multi-modal diffusion transformer(MM-DiT)-based model capable of generating dynamic, emotion-controllable, and multi-character dialogue videos.