Fetching the paper…

DAE-Talker: High Fidelity Speech-Driven Talking Face Generation with Diffusion Autoencoder · Around