基于柯西问题理解视觉变形器的对抗鲁棒性

Zheng Wang, Wenjie Ruan
{"title":"基于柯西问题理解视觉变形器的对抗鲁棒性","authors":"Zheng Wang, Wenjie Ruan","doi":"10.48550/arXiv.2208.00906","DOIUrl":null,"url":null,"abstract":"Recent research on the robustness of deep learning has shown that Vision Transformers (ViTs) surpass the Convolutional Neural Networks (CNNs) under some perturbations, e.g., natural corruption, adversarial attacks, etc. Some papers argue that the superior robustness of ViT comes from the segmentation of its input images; others say that the Multi-head Self-Attention (MSA) is the key to preserving the robustness. In this paper, we aim to introduce a principled and unified theoretical framework to investigate such an argument on ViT's robustness. We first theoretically prove that, unlike Transformers in Natural Language Processing, ViTs are Lipschitz continuous. Then we theoretically analyze the adversarial robustness of ViTs from the perspective of the Cauchy Problem, via which we can quantify how the robustness propagates through layers. We demonstrate that the first and last layers are the critical factors to affect the robustness of ViTs. Furthermore, based on our theory, we empirically show that unlike the claims from existing research, MSA only contributes to the adversarial robustness of ViTs under weak adversarial attacks, e.g., FGSM, and surprisingly, MSA actually comprises the model's adversarial robustness under stronger attacks, e.g., PGD attacks.","PeriodicalId":74091,"journal":{"name":"Machine learning and knowledge discovery in databases : European Conference, ECML PKDD ... : proceedings. ECML PKDD (Conference)","volume":null,"pages":null},"PeriodicalIF":0.0000,"publicationDate":"2022-08-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":"5","resultStr":"{\"title\":\"Understanding Adversarial Robustness of Vision Transformers via Cauchy Problem\",\"authors\":\"Zheng Wang, Wenjie Ruan\",\"doi\":\"10.48550/arXiv.2208.00906\",\"DOIUrl\":null,\"url\":null,\"abstract\":\"Recent research on the robustness of deep learning has shown that Vision Transformers (ViTs) surpass the Convolutional Neural Networks (CNNs) under some perturbations, e.g., natural corruption, adversarial attacks, etc. Some papers argue that the superior robustness of ViT comes from the segmentation of its input images; others say that the Multi-head Self-Attention (MSA) is the key to preserving the robustness. In this paper, we aim to introduce a principled and unified theoretical framework to investigate such an argument on ViT's robustness. We first theoretically prove that, unlike Transformers in Natural Language Processing, ViTs are Lipschitz continuous. Then we theoretically analyze the adversarial robustness of ViTs from the perspective of the Cauchy Problem, via which we can quantify how the robustness propagates through layers. We demonstrate that the first and last layers are the critical factors to affect the robustness of ViTs. Furthermore, based on our theory, we empirically show that unlike the claims from existing research, MSA only contributes to the adversarial robustness of ViTs under weak adversarial attacks, e.g., FGSM, and surprisingly, MSA actually comprises the model's adversarial robustness under stronger attacks, e.g., PGD attacks.\",\"PeriodicalId\":74091,\"journal\":{\"name\":\"Machine learning and knowledge discovery in databases : European Conference, ECML PKDD ... : proceedings. ECML PKDD (Conference)\",\"volume\":null,\"pages\":null},\"PeriodicalIF\":0.0000,\"publicationDate\":\"2022-08-01\",\"publicationTypes\":\"Journal Article\",\"fieldsOfStudy\":null,\"isOpenAccess\":false,\"openAccessPdf\":\"\",\"citationCount\":\"5\",\"resultStr\":null,\"platform\":\"Semanticscholar\",\"paperid\":null,\"PeriodicalName\":\"Machine learning and knowledge discovery in databases : European Conference, ECML PKDD ... : proceedings. ECML PKDD (Conference)\",\"FirstCategoryId\":\"1085\",\"ListUrlMain\":\"https://doi.org/10.48550/arXiv.2208.00906\",\"RegionNum\":0,\"RegionCategory\":null,\"ArticlePicture\":[],\"TitleCN\":null,\"AbstractTextCN\":null,\"PMCID\":null,\"EPubDate\":\"\",\"PubModel\":\"\",\"JCR\":\"\",\"JCRName\":\"\",\"Score\":null,\"Total\":0}","platform":"Semanticscholar","paperid":null,"PeriodicalName":"Machine learning and knowledge discovery in databases : European Conference, ECML PKDD ... : proceedings. ECML PKDD (Conference)","FirstCategoryId":"1085","ListUrlMain":"https://doi.org/10.48550/arXiv.2208.00906","RegionNum":0,"RegionCategory":null,"ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":null,"EPubDate":"","PubModel":"","JCR":"","JCRName":"","Score":null,"Total":0}
引用次数: 5

摘要

最近对深度学习鲁棒性的研究表明,视觉变形(ViTs)在一些扰动下,如自然腐败、对抗性攻击等,超过了卷积神经网络(cnn)。一些论文认为ViT的鲁棒性来自于对输入图像的分割;也有人认为多头自注意(MSA)是保持鲁棒性的关键。在本文中,我们旨在引入一个原则性和统一的理论框架来研究这种关于ViT鲁棒性的论点。我们首先从理论上证明,与自然语言处理中的变形金刚不同,vit是利普希茨连续的。然后从柯西问题的角度对vit的对抗鲁棒性进行了理论分析,量化了鲁棒性如何在各层间传播。我们证明了第一层和最后一层是影响vit鲁棒性的关键因素。此外,根据我们的理论,我们通过经验表明,与现有研究的说法不同,MSA仅有助于vit在弱对抗性攻击(例如FGSM)下的对抗性鲁棒性,而令人惊讶的是,MSA实际上包含了模型在更强攻击(例如PGD攻击)下的对抗性鲁棒性。
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
Understanding Adversarial Robustness of Vision Transformers via Cauchy Problem
Recent research on the robustness of deep learning has shown that Vision Transformers (ViTs) surpass the Convolutional Neural Networks (CNNs) under some perturbations, e.g., natural corruption, adversarial attacks, etc. Some papers argue that the superior robustness of ViT comes from the segmentation of its input images; others say that the Multi-head Self-Attention (MSA) is the key to preserving the robustness. In this paper, we aim to introduce a principled and unified theoretical framework to investigate such an argument on ViT's robustness. We first theoretically prove that, unlike Transformers in Natural Language Processing, ViTs are Lipschitz continuous. Then we theoretically analyze the adversarial robustness of ViTs from the perspective of the Cauchy Problem, via which we can quantify how the robustness propagates through layers. We demonstrate that the first and last layers are the critical factors to affect the robustness of ViTs. Furthermore, based on our theory, we empirically show that unlike the claims from existing research, MSA only contributes to the adversarial robustness of ViTs under weak adversarial attacks, e.g., FGSM, and surprisingly, MSA actually comprises the model's adversarial robustness under stronger attacks, e.g., PGD attacks.
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
Explaining Full-disk Deep Learning Model for Solar Flare Prediction using Attribution Methods Offline Reinforcement Learning with On-Policy Q-Function Regularization Visualizing Overlapping Biclusterings and Boolean Matrix Factorizations An Examination of Wearable Sensors and Video Data Capture for Human Exercise Classification Online Network Source Optimization with Graph-Kernel MAB
×
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