Defense-GAN: Protecting Classifiers Against Adversarial Attacks Using Generative Models

This repository contains the minimal implementation of the Defense-GAN by pytorch, it was pulished on ICLR 2018.

https://arxiv.org/abs/1805.06605

Samangouei, P., Kabkab, M., & Chellappa, R. (2018). Defense-Gan: Protecting classifiers against adversarial attacks using generative models. In 6th International Conference on Learning Representations, ICLR 2018.