Skip to content
DR-GAN Official Tensorflow Implementation
Python Shell
Branch: master
Clone or download
Latest commit 42d531c Dec 30, 2019
Permalink
Type Name Latest commit message Commit time
Failed to load latest commit information.
images update Dec 11, 2019
LICENSE update Dec 11, 2019
Loss_AMSoftmax.py working code Dec 11, 2019
README.md update Dec 11, 2019
config.py working code Dec 11, 2019
main_DR_GAN_multi.py working code Dec 11, 2019
main_pretrain.py working code Dec 11, 2019
main_test.py working code Dec 11, 2019
main_test_Luan.py working code Dec 11, 2019
model_DR_GAN_multi.py working code Dec 11, 2019
model_pretrain.py Update model_pretrain.py Dec 30, 2019
model_pretrain_.py working code Dec 11, 2019
ops.py working code Dec 11, 2019
pretrain.sh working code Dec 11, 2019
train_DRGAN_multi.sh working code Dec 11, 2019
utils.py working code Dec 11, 2019

README.md

Disentangled Representation Learning GAN for Pose-Invariant Face Recognition (DR-GAN)

[Project page] [CVPR'17 Paper] [PAMI'18 Paper]

Teaser

Library requirements

  • Tensorflow

Data

VGG2

Please create txt files with each line in the form <image_path>;<person_id>

Run the code

Pretraining

pretrain.sh

Finetunning

train_DRGAN_multi.sh

Citation

If you find this work useful, please cite our papers with the following bibtex:

@article{ tran2018representation,
  author = { Luan Tran and Xi Yin and Xiaoming Liu },
  title = { Representation Learning by Rotating Your Faces },
  booktitle = { IEEE Transactions on Pattern Analysis and Machine Intelligence },
  month = { September },
  year = { 2018 },
}
@inproceedings{ tran2017disentangled,
  author = { Luan Tran and Xi Yin and Xiaoming Liu },
  title = { Disentangled Representation Learning GAN for Pose-Invariant Face Recognition },
  booktitle = { In Proceeding of IEEE Computer Vision and Pattern Recognition },
  address = { Honolulu, HI },
  month = { July },
  year = { 2017 },
}

Contacts

If you have any questions, feel free to drop an email to tranluan@msu.edu.

You can’t perform that action at this time.