Improving the adversarial transferability with relational graphs ensemble adversarial attack

Author:

Pi Jiatian,Luo Chaoyang,Xia Fen,Jiang Ning,Wu Haiying,Wu Zhiyou

Abstract

In transferable black-box attacks, adversarial samples remain adversarial across multiple models and are more likely to attack unknown models. From this view, acquiring and exploiting multiple models is the key to improving transferability. For exploiting multiple models, existing approaches concentrate on differences among models but ignore the underlying complex dependencies. This exacerbates the issue of unbalanced and inadequate attacks on multiple models. To this problem, this paper proposes a novel approach, called Relational Graph Ensemble Attack (RGEA), to exploit the dependencies among multiple models. Specifically, we redefine the multi-model ensemble attack as a multi-objective optimization and create a sub-optimization problem to compute the optimal attack direction, but there are serious time-consuming problems. For this time-consuming problem, we define the vector representation of the model, extract the dependency matrix, and then equivalently simplify the sub-optimization problem by utilizing the dependency matrix. Finaly, we theoretically extend to investigate the connection between RGEA and the traditional multiple gradient descent algorithm (MGDA). Notably, combining RGEA further enhances the transferability of existing gradient-based attacks. The experiments using ten normal training models and ten defensive models on the labeled face in the wild (LFW) dataset demonstrate that RGEA improves the success rate of white-box attacks and further boosts the transferability of black-box attacks.

Publisher

Frontiers Media SA

Subject

General Neuroscience

Reference51 articles.

1. “Adversarial attacks on node embeddings via graph poisoning,”;Bojchevski;Proceedings of the 36th International Conference on Machine Learning, volume 97 of Proceedings of Machine Learning Research,2019

2. “Relational multi-task learning: Modeling relations between data and tasks,” CaoK. YouJ. LeskovecJ. International Conference on Learning Representations2022

3. “Towards evaluating the robustness of neural networks,”;Carlini;2017 IEEE Symposium on Security and Privacy (SP),2017

4. A new ensemble adversarial attack powered by long-term gradient memories;Che;Proc. AAAI Conf. Artif. Intell,2020

5. Going far boosts attack transferability, but do not do it;Chen;arXiv preprint,2021

Cited by 1 articles. 订阅此论文施引文献 订阅此论文施引文献,注册后可以免费订阅5篇论文的施引文献,订阅后可以查看论文全部施引文献

1. Layered Distillation Training: A Study of Adversarial Attacks and Defenses;2024 3rd International Conference for Innovation in Technology (INOCON);2024-03-01

同舟云学术

1.学者识别学者识别

2.学术分析学术分析

3.人才评估人才评估

"同舟云学术"是以全球学者为主线,采集、加工和组织学术论文而形成的新型学术文献查询和分析系统,可以对全球学者进行文献检索和人才价值评估。用户可以通过关注某些学科领域的顶尖人物而持续追踪该领域的学科进展和研究前沿。经过近期的数据扩容,当前同舟云学术共收录了国内外主流学术期刊6万余种,收集的期刊论文及会议论文总量共计约1.5亿篇,并以每天添加12000余篇中外论文的速度递增。我们也可以为用户提供个性化、定制化的学者数据。欢迎来电咨询!咨询电话:010-8811{复制后删除}0370

www.globalauthorid.com

TOP

Copyright © 2019-2024 北京同舟云网络信息技术有限公司
京公网安备11010802033243号  京ICP备18003416号-3