Controllable 3D Generative Adversarial Face Model via Disentangling Shape and Appearance

Fariborz Teherkhani, Aashish Rai*, Shaunak Srivastava*, Quankai Gao*, Xuanbai Chen, Fernando de la Torre, Steven Song, Aayush Prakash, Daeil Kim (* equal contribution)

Carnegie Mellon University, Facebook/Meta

This is the official Pytorch implementation of the paper.

[Project Page] [Video] [Colab Demo] [Arxiv]

Testing

Download pre-trained weights and put the "checkpoints" folder in the main directory. [Link]

Generate 3D Faces (mesh and texture)
```
python generate_faces.py
```
Generate meshes only
```
python test_gan3d.py
```
Generate textures only
```
python test_texture.py
```

Train your own model

Preprocess data

- Download Facescape dataset and specify path to the "facescape_trainset" folder.

python preprocess_traindata.py

Start training

Shape

Train AE
python train_ae.py

Generate Reduced Data
python gen_reduced_data.py

Train GAN
python train_gan3d.py

Texture

Train P-GAN
python train_texture.py --init_step 1 --batch_size 128

License

The code is available under MIT License. Please read the license terms available at [Link]

Citation

If you use find this paper/code useful, please consider citing:

@misc{taherkhani2022controllable,
      title={Controllable 3D Generative Adversarial Face Model via Disentangling Shape and Appearance}, 
      author={Fariborz Taherkhani and Aashish Rai and Quankai Gao and Shaunak Srivastava and Xuanbai Chen and Fernando de la Torre and Steven Song and Aayush Prakash and Daeil Kim},
      year={2022},
      eprint={2208.14263},
      archivePrefix={arXiv},
      primaryClass={cs.CV}
}

MarlNox / 3DFaceCAM