Aegis:减轻针对深度神经网络的定向比特翻转攻击

Jialai Wang, Ziyuan Zhang, Meiqi Wang, Han Qiu, Tianwei Zhang, Qi Li, Zongpeng Li, Tao Wei, Chao Zhang
{"title":"Aegis:减轻针对深度神经网络的定向比特翻转攻击","authors":"Jialai Wang, Ziyuan Zhang, Meiqi Wang, Han Qiu, Tianwei Zhang, Qi Li, Zongpeng Li, Tao Wei, Chao Zhang","doi":"10.48550/arXiv.2302.13520","DOIUrl":null,"url":null,"abstract":"Bit-flip attacks (BFAs) have attracted substantial attention recently, in which an adversary could tamper with a small number of model parameter bits to break the integrity of DNNs. To mitigate such threats, a batch of defense methods are proposed, focusing on the untargeted scenarios. Unfortunately, they either require extra trustworthy applications or make models more vulnerable to targeted BFAs. Countermeasures against targeted BFAs, stealthier and more purposeful by nature, are far from well established. In this work, we propose Aegis, a novel defense method to mitigate targeted BFAs. The core observation is that existing targeted attacks focus on flipping critical bits in certain important layers. Thus, we design a dynamic-exit mechanism to attach extra internal classifiers (ICs) to hidden layers. This mechanism enables input samples to early-exit from different layers, which effectively upsets the adversary's attack plans. Moreover, the dynamic-exit mechanism randomly selects ICs for predictions during each inference to significantly increase the attack cost for the adaptive attacks where all defense mechanisms are transparent to the adversary. We further propose a robustness training strategy to adapt ICs to the attack scenarios by simulating BFAs during the IC training phase, to increase model robustness. Extensive evaluations over four well-known datasets and two popular DNN structures reveal that Aegis could effectively mitigate different state-of-the-art targeted attacks, reducing attack success rate by 5-10$\\times$, significantly outperforming existing defense methods.","PeriodicalId":91597,"journal":{"name":"Proceedings of the ... USENIX Security Symposium. UNIX Security Symposium","volume":"42 1","pages":"2329-2346"},"PeriodicalIF":0.0000,"publicationDate":"2023-02-27","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":"3","resultStr":"{\"title\":\"Aegis: Mitigating Targeted Bit-flip Attacks against Deep Neural Networks\",\"authors\":\"Jialai Wang, Ziyuan Zhang, Meiqi Wang, Han Qiu, Tianwei Zhang, Qi Li, Zongpeng Li, Tao Wei, Chao Zhang\",\"doi\":\"10.48550/arXiv.2302.13520\",\"DOIUrl\":null,\"url\":null,\"abstract\":\"Bit-flip attacks (BFAs) have attracted substantial attention recently, in which an adversary could tamper with a small number of model parameter bits to break the integrity of DNNs. To mitigate such threats, a batch of defense methods are proposed, focusing on the untargeted scenarios. Unfortunately, they either require extra trustworthy applications or make models more vulnerable to targeted BFAs. Countermeasures against targeted BFAs, stealthier and more purposeful by nature, are far from well established. In this work, we propose Aegis, a novel defense method to mitigate targeted BFAs. The core observation is that existing targeted attacks focus on flipping critical bits in certain important layers. Thus, we design a dynamic-exit mechanism to attach extra internal classifiers (ICs) to hidden layers. This mechanism enables input samples to early-exit from different layers, which effectively upsets the adversary's attack plans. Moreover, the dynamic-exit mechanism randomly selects ICs for predictions during each inference to significantly increase the attack cost for the adaptive attacks where all defense mechanisms are transparent to the adversary. We further propose a robustness training strategy to adapt ICs to the attack scenarios by simulating BFAs during the IC training phase, to increase model robustness. Extensive evaluations over four well-known datasets and two popular DNN structures reveal that Aegis could effectively mitigate different state-of-the-art targeted attacks, reducing attack success rate by 5-10$\\\\times$, significantly outperforming existing defense methods.\",\"PeriodicalId\":91597,\"journal\":{\"name\":\"Proceedings of the ... USENIX Security Symposium. UNIX Security Symposium\",\"volume\":\"42 1\",\"pages\":\"2329-2346\"},\"PeriodicalIF\":0.0000,\"publicationDate\":\"2023-02-27\",\"publicationTypes\":\"Journal Article\",\"fieldsOfStudy\":null,\"isOpenAccess\":false,\"openAccessPdf\":\"\",\"citationCount\":\"3\",\"resultStr\":null,\"platform\":\"Semanticscholar\",\"paperid\":null,\"PeriodicalName\":\"Proceedings of the ... USENIX Security Symposium. UNIX Security Symposium\",\"FirstCategoryId\":\"1085\",\"ListUrlMain\":\"https://doi.org/10.48550/arXiv.2302.13520\",\"RegionNum\":0,\"RegionCategory\":null,\"ArticlePicture\":[],\"TitleCN\":null,\"AbstractTextCN\":null,\"PMCID\":null,\"EPubDate\":\"\",\"PubModel\":\"\",\"JCR\":\"\",\"JCRName\":\"\",\"Score\":null,\"Total\":0}","platform":"Semanticscholar","paperid":null,"PeriodicalName":"Proceedings of the ... USENIX Security Symposium. UNIX Security Symposium","FirstCategoryId":"1085","ListUrlMain":"https://doi.org/10.48550/arXiv.2302.13520","RegionNum":0,"RegionCategory":null,"ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":null,"EPubDate":"","PubModel":"","JCR":"","JCRName":"","Score":null,"Total":0}
引用次数: 3

摘要

比特翻转攻击(Bit-flip attacks, bfa)近年来引起了广泛关注,攻击者可以通过篡改少量模型参数比特来破坏深度神经网络的完整性。为了减轻这种威胁,提出了一批针对非目标场景的防御方法。不幸的是,它们要么需要额外的可信应用程序,要么使模型更容易受到目标bfa的攻击。针对目标bfa的对抗措施,从本质上来说,更隐蔽、更有目的性,远没有得到很好的建立。在这项工作中,我们提出了一种新的防御方法Aegis来减轻目标BFAs。核心观察是,现有的目标攻击专注于翻转某些重要层的关键位。因此,我们设计了一个动态退出机制,将额外的内部分类器(ic)附加到隐藏层。这种机制使输入样本能够从不同的层提前退出,从而有效地打乱了对手的攻击计划。此外,动态退出机制在每次推理过程中随机选择预测ic,大大增加了自适应攻击的攻击成本,其中所有防御机制对对手都是透明的。我们进一步提出了一种鲁棒性训练策略,通过在集成电路训练阶段模拟BFAs,使集成电路适应攻击场景,以提高模型的鲁棒性。对四个知名数据集和两种流行的DNN结构的广泛评估表明,宙斯盾可以有效地缓解不同的最先进的目标攻击,将攻击成功率降低5-10倍,显著优于现有的防御方法。
本文章由计算机程序翻译,如有差异,请以英文原文为准。
查看原文
分享 分享
微信好友 朋友圈 QQ好友 复制链接
本刊更多论文
Aegis: Mitigating Targeted Bit-flip Attacks against Deep Neural Networks
Bit-flip attacks (BFAs) have attracted substantial attention recently, in which an adversary could tamper with a small number of model parameter bits to break the integrity of DNNs. To mitigate such threats, a batch of defense methods are proposed, focusing on the untargeted scenarios. Unfortunately, they either require extra trustworthy applications or make models more vulnerable to targeted BFAs. Countermeasures against targeted BFAs, stealthier and more purposeful by nature, are far from well established. In this work, we propose Aegis, a novel defense method to mitigate targeted BFAs. The core observation is that existing targeted attacks focus on flipping critical bits in certain important layers. Thus, we design a dynamic-exit mechanism to attach extra internal classifiers (ICs) to hidden layers. This mechanism enables input samples to early-exit from different layers, which effectively upsets the adversary's attack plans. Moreover, the dynamic-exit mechanism randomly selects ICs for predictions during each inference to significantly increase the attack cost for the adaptive attacks where all defense mechanisms are transparent to the adversary. We further propose a robustness training strategy to adapt ICs to the attack scenarios by simulating BFAs during the IC training phase, to increase model robustness. Extensive evaluations over four well-known datasets and two popular DNN structures reveal that Aegis could effectively mitigate different state-of-the-art targeted attacks, reducing attack success rate by 5-10$\times$, significantly outperforming existing defense methods.
求助全文
通过发布文献求助,成功后即可免费获取论文全文。 去求助
来源期刊
自引率
0.00%
发文量
0
期刊最新文献
Research on the Security of Visual Reasoning CAPTCHA A Highly Accurate Query-Recovery Attack against Searchable Encryption using Non-Indexed Documents Hot Pixels: Frequency, Power, and Temperature Attacks on GPUs and ARM SoCs PTW: Pivotal Tuning Watermarking for Pre-Trained Image Generators Inductive Graph Unlearning
×
引用
GB/T 7714-2015
复制
MLA
复制
APA
复制
导出至
BibTeX EndNote RefMan NoteFirst NoteExpress
×
×
提示
您的信息不完整,为了账户安全,请先补充。
现在去补充
×
提示
您因"违规操作"
具体请查看互助需知
我知道了
×
提示
现在去查看 取消
×
提示
确定
0
微信
客服QQ
Book学术公众号 扫码关注我们
反馈
×
意见反馈
请填写您的意见或建议
请填写您的手机或邮箱
已复制链接
已复制链接
快去分享给好友吧!
我知道了
×
扫码分享
扫码分享
Book学术官方微信
Book学术文献互助
Book学术文献互助群
群 号:481959085
Book学术
文献互助 智能选刊 最新文献 互助须知 联系我们:info@booksci.cn
Book学术提供免费学术资源搜索服务,方便国内外学者检索中英文文献。致力于提供最便捷和优质的服务体验。
Copyright © 2023 Book学术 All rights reserved.
ghs 京公网安备 11010802042870号 京ICP备2023020795号-1