Code for "Pixel-wise Conditioning of Generative Adversarial Networks", ESANN2019 and Neurocomputing 2020
Usage
Conditional gan
python src/train_cgan.py config_file dataset_path
Regularized gan
python src/train_rcgan.py config_file dataset_path
Regularized gan with PacGAN
python src/train_rcgan_pac.py config_file dataset_path
Dataset formatting
Saved in HDF5 format. Use the code from scripts/datasets to format a dataset from a large single image or from classical datasets.
Figures
The figures are generated from the scrips/plot scripts