GeneGAN in PyTorch

Introduction

This is the re-implemented source code for the paper GeneGAN: Learning Object Transfiguration and Attribute Subspace from Unpaired Data. We reproduce the results using PyTorch.

Installation

• Install python libraries

  pip3 install -r requirements.txt
  

Dataset Preparation

• Download celebA dataset and unzip it into datasets directory. There are various source providers for CelebA datasets. To ensure that the size of downloaded images is correct, please run identify datasets/celebA/data/000001.jpg. The size should be 409 x 687 if you are using the same dataset. Besides, please ensure that you have the following directory tree structure.

  ├── datasets
  │   └── celebA
  │       ├── data
  │       ├── list_attr_celeba.txt
  │       ├── list_eval_partition.txt
  │       └── list_landmarks_celeba.txt
  

• Run python3 preprocess.py. It will take several miniutes to preprocess all face images. A new directory datasets/celebA/align_5p will be created.

GeneGAN train/test (Attribute migration)

• Choose an attribute for training/testing (The optional attributes are: Eyeglasses | Bangs | Male | Mouth_Slightly_Open | No_Beard ..., we select Bangs as the attribute in the following example)

  make train TASK='Bangs'      # Train
  make show_train TASK='Bangs' # Show train details
  make test TASK='Bangs'       # Test
  make show_test TASK='Bangs'  # Show test details

Citation

• If you use this code for your research, please cite our paper:

  @inproceedings{DBLP:conf/bmvc/ZhouXYFHH17,
    author    = {Shuchang Zhou and
                Taihong Xiao and
                Yi Yang and
                Dieqiao Feng and
                Qinyao He and
                Weiran He},
    title     = {GeneGAN: Learning Object Transfiguration and Attribute Subspace from Unpaired Data},
    booktitle = {Proceedings of the British Machine Vision Conference (BMVC)},
    year      = {2017},
    url       = {http://arxiv.org/abs/1705.04932},
    timestamp = {http://dblp.uni-trier.de/rec/bib/journals/corr/ZhouXYFHH17},
    bibsource = {dblp computer science bibliography, http://dblp.org}
  }
  

Acknowledgements

This code borrows heavily from pytorch-CycleGAN-and-pix2pix