APE-GAN: Adversarial Perturbation Elimination with GAN

Guoqing Jin; Shiwei Shen; Dongming Zhang; Feng Dai; Yongdong Zhang

ICASSP 2019accepted0 citations

APE-GAN: Adversarial Perturbation Elimination with GAN

Guoqing Jin, Shiwei Shen, Dongming Zhang, Feng Dai, Yongdong Zhang

Abstract

Although Deep Neural Networks could achieve state-of-the-art performance while recongnizing images, they often suffer a tremendous defeat from adversarial examples-inputs generated by utilizing imperceptible but intentional perturbations to samples from the datasets. So far, very few methods have provided a significant defense to adversarial examples. In this paper, an effective framework based Generative Adversarial Nets(GAN) is proposed to defense against the adversarial examples. The essense of the model is to eliminate the adversarial perturbations being highly aligned with the weight vectors of nueral models. Extensive experiments on benchmark datasets MNIST, CIFAR10 and ImageNet indicate that our framework is able to defense against adversarial examples effectively.

BibTeX

@inproceedings{icassp2019_apeganadversaria,
  title = {APE-GAN: Adversarial Perturbation Elimination with GAN},
  author = {Guoqing Jin and Shiwei Shen and Dongming Zhang and Feng Dai and Yongdong Zhang},
  booktitle = {ICASSP 2019},
  year = {2019}
}