Skip to content
Code for “Adversarial Learning with Local Coordinate Coding”
Branch: master
Clone or download
Latest commit ccece16 Mar 27, 2019
Type Name Latest commit message Commit time
Failed to load latest commit information.
images update images Aug 9, 2018 update README Aug 14, 2018 update model_definer Aug 14, 2018 update parameter name Mar 25, 2019 fix import error Mar 27, 2019


Pytorch implementation for “Adversarial Learning with Local Coordinate Coding”.

Architecture of LCCGAN

  • AutoEncoder (AE) learns embeddings on the latent manifold.
  • Local Coordinate Coding (LCC) learns local coordinate systems.
  • The LCC sampling method is conducted on the latent manifold.

Gometric Views of LCC Sampling

  • With the help of LCC, we obtain local coordinate systems for sampling on the latent manifold.
  • Using the local coordinate systems, LCC-GANs always sample some meaningful points to generate new images with different attributes.

Training Algorithm


python 2.7



In our paper, to sample different images, we train our model on four datasets, respectively.


  • Train LCCGAN on Oxford-102 Flowers dataset.
python --dataset flowers --dataroot your_images_folder --batchSize 64 --imageSize 64 --cuda
  • 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 --dataset name_o_dataset --dataroot path_of_dataset


  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}
You can’t perform that action at this time.