Experimental work-in-progress project

An open-source, cross-platform project that converts audio input into realistic facial expressions in real-time following the MPEG-4 (FBA) standard.

Core (uv)

uv sync

MFA (micromamba)

micromamba create -n mfa -c conda-forge python=3.12 montreal-forced-aligner
micromamba activate mfa

mfa model download acoustic english_us_arpa
mfa model download dictionary english_us_arpa
mfa model download g2p english_us_arpa

Dataset Download is now integrated in the training script.

python training/train.py --config training/recipes/tcn_config.toml

This project uses the LibriSpeech ASR corpus (CC BY 4.0 license).