Generate talking-head avatars or digital humans from text or audio.
No models in this category are strong on that dimension yet.