跳到主要导航 跳到搜索 跳到主要内容

Transferability of White-box Perturbations: Query-Efficient Adversarial Attacks against Commercial DNN Services

  • Beijing Institute of Technology
  • Tsinghua University

科研成果: 书/报告/会议事项章节 › 会议稿件 › 同行评审

摘要

Deep Neural Networks (DNNs) have been proven to be vulnerable to adversarial attacks. Existing decision-based adversarial attacks require large numbers of queries to find an effective adversarial example, resulting in a heavy query cost and also performance degradation under defenses. In this paper, we propose the Dispersed Sampling Attack (DSA), which is a query-efficient decision-based adversarial attack by exploiting the transferability of white-box perturbations. DSA can generate diverse examples with different locations in the embedding space, which provides more information about the adversarial region of substitute models and allows us to search for transferable perturbations. Specifically, DSA samples in a hypersphere centered on an original image, and progressively constrains the perturbation. Extensive experiments are conducted on public datasets to evaluate the performance of DSA in closed-set and open-set scenarios. DSA outperforms the state-of-the-art attacks in terms of both attack success rate (ASR) and average number of queries (AvgQ). Specifically, DSA achieves an ASR of about 90% with an AvgQ of 200 on 4 well-known commercial DNN services.

源语言英语
主期刊名Proceedings of the 33rd USENIX Security Symposium
出版商USENIX Association
页2991-3008
页数18
ISBN(电子版)9781939133441
出版状态已出版 - 2024
活动33rd USENIX Security Symposium, USENIX Security 2024 - Philadelphia, 美国
期限: 14 8月 2024 → 16 8月 2024

丛书

姓名Proceedings of the 33rd USENIX Security Symposium

会议

会议33rd USENIX Security Symposium, USENIX Security 2024
国家/地区美国
市Philadelphia
时期14/08/24 → 16/08/24

学术指纹

探究 'Transferability of White-box Perturbations: Query-Efficient Adversarial Attacks against Commercial DNN Services' 的科研主题。它们共同构成独一无二的学术指纹。

引用此