Unrestricted Facial Geometry Reconstruction Using Image-to-Image Translation

[Arxiv] [Video]

Evaluation code for Unrestricted Facial Geometry Reconstruction Using Image-to-Image Translation. Given a single image, the code outputs the reconstructed mesh.

Recent Updates

2020.05.07: We have released a PyTorch version of this repository and encourage you to give it a try!

2018.07.20: Spiltted postprocess into the different steps composing the pipeline, changed models to float to save space.

Setup & Usage

The project was tested on Ubuntu 14.04 LTS with Matlab R2015b, to run it follow these instructions:

• Make sure you have Torch installed on your machine.

• Install the mattorch and nngraph packages.

  luarocks install mattorch    
  luarocks install nngraph

• Download the model files and extract them into the models directroy.

• Run the runme.m script in Matlab.

Citation

If you use this code for your research, please cite our paper Unrestricted Facial Geometry Reconstruction Using Image-to-Image Translation:

@article{sela2017unrestricted,
  title={Unrestricted Facial Geometry Reconstruction Using Image-to-Image Translation},
  author={Sela, Matan and Richardson, Elad and Kimmel, Ron},
  journal={arxiv},
  year={2017}
}