FACEGOOD/FACEGOOD-Audio2Face
Audio2Face transforms speech audio into facial animation blendshape weights to drive digital characters in real-time.
Not currently ranked — collecting fresh signals.
star history
FACEGOOD Audio2Face is a neural network system that converts audio waveforms into facial blendshape weights. It uses a formant network for audio analysis followed by an articulation network that incorporates emotional state vectors, generating blendshape weights through fully-connected layers. The system drives digital characters in Unreal Engine, enabling audio-driven facial animation for virtual humans.
Frequently asked
- What is FACEGOOD/FACEGOOD-Audio2Face?
- Audio2Face transforms speech audio into facial animation blendshape weights to drive digital characters in real-time.
- Is FACEGOOD-Audio2Face open source?
- Yes — FACEGOOD/FACEGOOD-Audio2Face is open source, released under the MIT license.
- What language is FACEGOOD-Audio2Face written in?
- FACEGOOD/FACEGOOD-Audio2Face is primarily written in Python.
- How popular is FACEGOOD-Audio2Face?
- FACEGOOD/FACEGOOD-Audio2Face has 1.9k stars on GitHub.
- Where can I find FACEGOOD-Audio2Face?
- FACEGOOD/FACEGOOD-Audio2Face is on GitHub at https://github.com/FACEGOOD/FACEGOOD-Audio2Face.