Improving Adversarial Robustness to Sensitivity and Invariance Attacks with Deep Metric Learning

TL;DR

This paper proposes a novel deep metric learning approach to improve adversarial robustness against sensitivity and invariance attacks by framing adversarial regularization as an optimal transport problem, showing promising preliminary results.

Contribution

It introduces a new framework using metric learning and optimal transport to enhance defense against both sensitivity and invariance adversarial attacks beyond Euclidean metrics.

Findings

Regularizing over invariant perturbations improves robustness.

The approach outperforms traditional Euclidean-based defenses.

Preliminary results show promising robustness enhancements.

Abstract

Intentionally crafted adversarial samples have effectively exploited weaknesses in deep neural networks. A standard method in adversarial robustness assumes a framework to defend against samples crafted by minimally perturbing a sample such that its corresponding model output changes. These sensitivity attacks exploit the model's sensitivity toward task-irrelevant features. Another form of adversarial sample can be crafted via invariance attacks, which exploit the model underestimating the importance of relevant features. Previous literature has indicated a tradeoff in defending against both attack types within a strictly L_p bounded defense. To promote robustness toward both types of attacks beyond Euclidean distance metrics, we use metric learning to frame adversarial regularization as an optimal transport problem. Our preliminary results indicate that regularizing over invariant perturbations in our framework improves both invariant and sensitivity defense.