OmniHuman

Create a lip-synced “talking head” video by animating a human figure from audio input (optionally using a character/image). The model focuses on realistic mouth movement and facial expressions to match speech, making it useful for avatar-style presentations and influencer-like clips.