Generative Adversarial Trainer: Defense to Adversarial Perturbations with GAN

TL;DR

This paper introduces a GAN-based adversarial training method that enhances neural network robustness against adversarial attacks, reduces overfitting, and improves generalization in supervised learning tasks.

Contribution

It presents the first use of GANs for adversarial training to improve classifier robustness and generalization in supervised learning.

Findings

Significantly lowers generalization error on CIFAR datasets.

Outperforms other regularization methods like Dropout.

Effectively generates adversarial examples to train more robust classifiers.

Abstract

We propose a novel technique to make neural network robust to adversarial examples using a generative adversarial network. We alternately train both classifier and generator networks. The generator network generates an adversarial perturbation that can easily fool the classifier network by using a gradient of each image. Simultaneously, the classifier network is trained to classify correctly both original and adversarial images generated by the generator. These procedures help the classifier network to become more robust to adversarial perturbations. Furthermore, our adversarial training framework efficiently reduces overfitting and outperforms other regularization methods such as Dropout. We applied our method to supervised learning for CIFAR datasets, and experimantal results show that our method significantly lowers the generalization error of the network. To the best of our knowledge, this is the first method which uses GAN to improve supervised learning.