Defending CycleGAN from Self-Adversarial Attack
In this repository you can find the extension of the original implementation of CycleGAN in pytorch. This extension contains two defense techniques against the self-adversarial attacks performed by the unsupervised image-to-image translation methods that hold the cycle-consistency property. More information on self-adversarial attacks can be found in our paper.
In order to reproduce our results, you can find all the configuration files in configs directory.