CodeTalker

A fork of codetalker repo. This repo containes an audio to vertices/flame parameters model that generate realistic facial animation in a speed close to realtime.

Environment

• Linux
• Python 3.6+
• Pytorch 2.0

Other necessary packages:

pip install -r requirements.txt

• ffmpeg (rendering output videos)
• Vertices2flame (Optional for flame blenshapes)

Dataset Preparation

Download "FLAME_sample.ply" from voca and put it in dataset/model.

VOCASET

Request the VOCASET data from https://voca.is.tue.mpg.de/. Place the downloaded files data_verts.npy, raw_audio_fixed.pkl, templates.pkl and subj_seq_to_idx.pkl in the folder vocaset/. Read the vertices/audio data and convert them to .npy/.wav files stored in vocaset/vertices_npy and vocaset/wav:

cd vocaset
python process_voca_data.py

VoxCeleb

Request the Flame model from https://flame.is.tue.mpg.de/. Place the download files flame_static_embedding.pkl, flame_dynamic_embedding.npy and generic_model.pkl in dataset/model. Request the dataset from (TODO)

Demo

Download the pretrained models and put the pretrained models under vox folder. Given the audio signal

sh scripts/download_weights.sh

• to animate a mesh in FLAME topology, run:

   sh scripts/demo.sh vox
  

  This script will automatically generate the rendered videos in the demo/output folder. You can also put your own test audio file (.wav format) under the demo/wav folder and specify the arguments in DEMO section of config/<dataset>/demo.yaml.

Training

The training operation shares a similar command:

sh scripts/train.sh <exp_name> config/<vox|vocaset>/<stage1|stage2>.yaml <vox|vocaset> <1|2|3>

Please replace <exp_name> with your own experiment name, <vox|vocaset"> by the name of your target dataset, i.e., vox or vocaset. Change the exp_dir in scripts/train.sh if needed. We just take an example for default commands below.

Training for Discrete Motion Prior

sh scripts/train.sh CodeTalker_s1 config/vox/stage1.yaml vox 1

Training for Speech-Driven Motion Synthesis

Make sure the paths of pre-trained models are correct, i.e., vqvae_pretrained_path and wav2vec2model_path in config/<vox|vocaset>/stage2.yaml.

sh scripts/train.sh CodeTalker_s2 config/vox/stage2.yaml vox 2

Deployment

Launch server endpoint:

python server.py --config config/vox/demo.yaml

Acknowledgement

We heavily borrow the code from FaceFormer, Learn2Listen, and VOCA. Thanks for sharing their code and huggingface-transformers for their wav2vec2 implementation. We also gratefully acknowledge the ETHZ-CVL for providing the B3D(AC)2 dataset and MPI-IS for releasing the VOCASET dataset. Any third-party packages are owned by their respective authors and must be used under their respective licenses.