Yazdi9/Talking_Face_Avatar

Combines Leonardo.ai image generation and ElevenLabs text-to-speech APIs to produce talking head videos from AI-generated portraits and synthesized audio. Uses SadTalker's audio-driven facial animation pipeline with ExpNet and PoseVAE models for expression and pose prediction, plus GFPGAN for face enhancement and Wav2Lip for lip-sync accuracy. Supports multiple generation modes (still, reference, resize) with configurable preprocessing and can process both static images and video sources via CLI or web UI.

232 stars. No commits in the last 6 months.