cjwbw / dreamtalk

RESEARCH/NON-COMMERCIAL USE ONLY: diffusion-based audio-driven expressive talking head generation

  • Public
  • 985 runs
  • GitHub
  • Paper
  • License