Skip to content
An implement of Disentangled Representation Learning GAN for Pose-Invariant Face Recognition
Branch: master
Clone or download
Permalink
Type Name Latest commit message Commit time
Failed to load latest commit information.
cfp-dataset Add cfp-dataset. May 20, 2018
data First version. May 19, 2018
model
options Fixing some pytorch compabilities issues, changing paths to be relative Jul 18, 2018
util First version. May 19, 2018
.gitattributes Initial commit May 17, 2018
README.md Add illustration of tensorboardX. Jan 30, 2019
exam.py First version. May 19, 2018
test.py First version. May 19, 2018
torch First version. May 19, 2018
train.py Change tensorflow to tensorboardX Jan 30, 2019

README.md

DR-GAN-by-pytorch

Disentangled Representation Learning GAN for Pose-Invariant Face Recognition

  • Authors: Luan Tran, Xi Yin, Xiaoming Liu
  • CVPR2017: http://cvlab.cse.msu.edu/pdfs/Tran_Yin_Liu_CVPR2017.pdf
  • Pytorch implimentation of DR-GAN (updated version in "Representation Learning by Rotating Your Faces")
  • Added a pretrained ResNet18 to offer a feature loss in order to improve Generator's performance. (Only in Multi_DRGAN)

Requirements

  • python 3.x
  • pytorch
  • torchvision
  • numpy
  • scipy
  • matplotlib
  • pillow
  • tensorboardX

How to use

Single-Image DR-GAN

  1. Modify model function at base_options.py to define single model.

    • Data needs to have ID and pose lables corresponds to each image.
    • If you don't have, default dataset is CFP_dataset. Modify dataroot function at base_options.py.
  2. Run train.py to train models

    • Trained models and Loss_log will be saved at "checkpoints" by default. Generated pictures will be saved at "result".

    python train.py

    • You can also use tensorboard to watch the loss graphs in real-time. (Install tensorboard before doing it.)

    tensorboard --logdir=/home/zhangjunhao/logs (Or the address of dir 'logs' in your folder.)

  3. Generate Image with arbitrary pose

    • Change the "save_path" in base_model.py.
    • Specify leaned model's filename by "--pretrained_G" option in base_options.py.
    • Generated images will be saved at specified result directory.

    python test.py

Multi-Image DR-GAN

  1. Modify model function at base_options.py to define multi model.

    • Data needs to have ID and pose lables corresponds to each image.
    • If you don't have, default dataset is CFP_dataset. Modify dataroot function at base_options.py.
  2. Run train.py to train models

    • Trained models and Loss_log will be saved at "checkpoints" by default. Generated pictures will be saved at "result".

    python train.py

    • You can also use tensorboard to watch the loss graphs in real-time. (Install tensorboard before doing it.)

    tensorboard --logdir=/home/zhangjunhao/logs (Or the address of dir 'logs' in your folder.)

  3. Generate Image with arbitrary pose

    • Change the "save_path" in base_model.py.
    • Specify leaned model's filename by "--pretrained_G" option in base_options.py.
    • Generated images will be saved at specified result directory.

    python test.py

You can’t perform that action at this time.