Enhancing All-to-X Backdoor Attacks with Optimized Target Class Mapping

TL;DR

This paper introduces a novel method to improve the success rate of All-to-X backdoor attacks in machine learning, demonstrating significant enhancements over existing approaches and highlighting the robustness of A2X attacks against defenses.

Contribution

The paper proposes an optimized attack strategy for All-to-X backdoor attacks, significantly increasing success rates and addressing an under-explored area in backdoor attack research.

Findings

Up to 28% increase in attack success rate

Average improvements of 6.7%, 16.4%, 14.1% on CIFAR10, CIFAR100, Tiny-ImageNet

A2X attacks are robust against current defenses

Abstract

Backdoor attacks pose severe threats to machine learning systems, prompting extensive research in this area. However, most existing work focuses on single-target All-to-One (A2O) attacks, overlooking the more complex All-to-X (A2X) attacks with multiple target classes, which are often assumed to have low attack success rates. In this paper, we first demonstrate that A2X attacks are robust against state-of-the-art defenses. We then propose a novel attack strategy that enhances the success rate of A2X attacks while maintaining robustness by optimizing grouping and target class assignment mechanisms. Our method improves the attack success rate by up to 28%, with average improvements of 6.7%, 16.4%, 14.1% on CIFAR10, CIFAR100, and Tiny-ImageNet, respectively. We anticipate that this study will raise awareness of A2X attacks and stimulate further research in this under-explored area. Our code is available at https://github.com/kazefjj/A2X-backdoor .