Skip to content
master
Switch branches/tags
Go to file
Code

Latest commit

 

Git stats

Files

Permalink
Failed to load latest commit information.
Type
Name
Latest commit message
Commit time
 
 
 
 
 
 
 
 
 
 
 
 

README.md

LCCGAN++

Pytorch implementation for “Improving Generative Adversarial Networks with Local Coordinate Coding”.

architecture

  • AutoEncoder (AE) learns the embeddings on the latent manifold.

  • Local Coordinate Coding (LCC) learns local coordinate systems. Specifically, we train LCCGAN-v1 with q=2 and LCCGAN++ with q=3.

  • The LCC sampling method is conducted on the latent manifold.

  • The LCCGAN is a general framework that can be applied to different GAN methods.

Dependencies

python 2.7

Pytorch 0.4

Training Method

  • Train LCCGAN++ on MNIST dataset.

    • python trainer.py --dataset mnist --dataroot ./mnist --nc 1
  • Train LCCGAN++ on Oxford-102 Flowers dataset.

    • python trainer.py --dataset Oxford-102 --dataroot your_images_folder
  • If you want to train the model on Large-scale CelebFaces Attributes (CelebA), Large-scale Scene Understanding (LSUN) or your own dataset. Just replace the hyperparameter like these:

    • python trainer.py --dataset name_o_dataset --dataroot path_of_dataset

Citation

@InProceedings{pmlr-v80-cao18a,
  title = 	 {Adversarial Learning with Local Coordinate Coding},
  author = 	 {Cao, Jiezhang and Guo, Yong and Wu, Qingyao and Shen, Chunhua and Huang, Junzhou and Tan, Mingkui},
  booktitle = 	 {Proceedings of the 35th International Conference on Machine Learning},
  pages = 	 {707--715},
  year = 	 {2018},
  editor = 	 {Dy, Jennifer and Krause, Andreas},
  volume = 	 {80},
  series = 	 {Proceedings of Machine Learning Research},
  address = 	 {Stockholmsmässan, Stockholm Sweden},
  month = 	 {10--15 Jul},
  publisher = 	 {PMLR}
}

About

Pytorch implementation of LCCGAN++.

Resources

Releases

No releases published

Packages

No packages published

Languages