Are Generative Classifiers More Robust to Adversarial Attacks?

Yingzhen Li, John Bradshaw, Yash Sharma

Abstract: There is a rising interest in studying the robustness of deep neural network classifiers against adversaries, with both advanced attack and defence techniques being actively developed. However, most recent work focuses on discriminative classifiers, which only model the conditional distribution of the labels given the inputs. In this paper, we propose and investigate the deep Bayes classifier, which improves classical naive Bayes with conditional deep generative models. We further develop detection methods for adversarial examples, which reject inputs with low likelihood under the generative model. Experimental results suggest that deep Bayes classifiers are more robust than deep discriminative classifiers, and that the proposed detection methods are effective against many recently proposed attacks.

Keywords: generative models, adversarial attack, defence, detection, Bayes' rule

TL;DR: We proposed a generative classifier based on deep generative models, and show improved robustness and detection results against adversarial attacks.

Community Implementations: [ 2 code implementations](https://www.catalyzex.com/paper/arxiv:1802.06552/code)