This repository contains demo code of our CVPR2019 paper. It contains code for the training and evaluation of a DFGAN with learned noise on the CIFAR-10 dataset.
The code is based on Python 2.7 and tensorflow 1.12.
How to use it
- Set the paths to the data and log directories in constants.py.
- Run init_datasets.py to download and convert the CIFAR-10 dataset.
2. DFGAN training and evaluation
- To train and evaluate a DFGAN with learned noise on CIFAR-10 run run_DFGAN_ln.py.
- To train and evaluate a standard GAN on CIFAR-10 run run_standard_GAN.py.
A PyTorch implementation of DFGAN is provided by: https://github.com/Johnson-yue/pytorch-DFGAN