This is the implementation of CSAE on Pytorch

1. Download the project

 git clone https://github.com/anlongstory/CSAE.git

2. Extract the files the dataset in 'data' folder

Ubuntu:

cd path/to/CVAE/data
unzip MNIST_img.zip
unzip letter.zip

Windows:

unzip directly

3. Generate Predefine evenly-distribute class centroids

python PEDCC.py

4. Train CSAE

python main.py

5. Test Model

python inference_model.py

Tip: Here, we just take MNIST as example, you can change some parameters or paths in config.py, and chang dataset in data_transform.py

If you have any question, please feel free to contact me :)