Skip to content

PeizhiYan/ArchitectHead

Repository files navigation

ArchitectHead: Continuous Level of Detail Control for 3D Gaussian Head Avatars

🌐 Homepage

Setup

# create conda virtual environment
conda create --name architect -y python=3.10
conda activate architect

# install cuda toolkit
conda install -c "nvidia/label/cuda-11.7.1" cuda-toolkit ninja
conda install -c conda-forge cudatoolkit=11.7 cudatoolkit-dev=11.7

# install PyTorch
pip install torch==2.0.1 torchvision --index-url https://download.pytorch.org/whl/cu117

# install other dependencies
pip install -r requirements.txt

Model Files

FLAME 2020

  • Download FLAME 2020 (fixed mouth, improved expressions, more data) from https://flame.is.tue.mpg.de/ and extract to ./model_data/FLAME2020

    • As an alternative to manually downloading, you can run ./download_FLAME.sh to automatically download and extract the model files.
  • Download https://github.com/yfeng95/DECA/blob/master/data/head_template.obj and save to ./model_data/head_template.obj

Pre-trained Avatar Models

Please download from https://github.com/PeizhiYan/ArchitectHead/releases/tag/trained-avatars, and unzip to ./saved_ckpts/

Demo

python avatar_viewer.py

About

[WACV 2026] Official code repo of paper ArchitectHead: Continuous Level of Detail Control for 3D Gaussian Head Avatars.

Topics

Resources

License

Stars

Watchers

Forks

Contributors

Languages