DMRP: Privacy-Preserving Deep Learning Model with Dynamic Masking and Random Permutation

IF 3.8 2区 计算机科学 Q2 COMPUTER SCIENCE, INFORMATION SYSTEMS
Chongzhen Zhang , Zhiwang Hu , Xiangrui Xu , Yong Liu , Bin Wang , Jian Shen , Tao Li , Yu Huang , Baigen Cai , Wei Wang
{"title":"DMRP: Privacy-Preserving Deep Learning Model with Dynamic Masking and Random Permutation","authors":"Chongzhen Zhang ,&nbsp;Zhiwang Hu ,&nbsp;Xiangrui Xu ,&nbsp;Yong Liu ,&nbsp;Bin Wang ,&nbsp;Jian Shen ,&nbsp;Tao Li ,&nbsp;Yu Huang ,&nbsp;Baigen Cai ,&nbsp;Wei Wang","doi":"10.1016/j.jisa.2025.103987","DOIUrl":null,"url":null,"abstract":"<div><div>Large AI models exhibit significant efficiency and precision in addressing complex problems. Despite their considerable advantages in various domains, these models encounter numerous challenges, notably high training costs. Currently, the training of distributed large AI models offers a solution to mitigate these elevated costs. However, distributed large AI models remain susceptible to data reconstruction attacks. A malicious server could leverage the intermediate results uploaded by clients to reconstruct the original data within the framework of distributed large AI models. This study first examines the underlying principles of data reconstruction attacks and proposes a privacy protection scheme. Our approach begins by obfuscating the mapping relationship between embeddings and the original data to ensure privacy protection. Specifically, during the upload of embedding data by clients to the server, genuine embeddings are concealed to prevent unauthorized access by malicious servers. Building on this concept, we introduce <em>DMRP</em>, a defensive mechanism featuring Dynamic Masking and Random Permutation, designed to mitigate data reconstruction attacks while maintaining the accuracy of the primary task. Our experiments, conducted across three models and four datasets, demonstrate the effectiveness of DMRP in countering data reconstruction attacks within distributed large-scale AI models.</div></div>","PeriodicalId":48638,"journal":{"name":"Journal of Information Security and Applications","volume":"89 ","pages":"Article 103987"},"PeriodicalIF":3.8000,"publicationDate":"2025-02-10","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":"0","resultStr":null,"platform":"Semanticscholar","paperid":null,"PeriodicalName":"Journal of Information Security and Applications","FirstCategoryId":"94","ListUrlMain":"https://www.sciencedirect.com/science/article/pii/S2214212625000250","RegionNum":2,"RegionCategory":"计算机科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":null,"EPubDate":"","PubModel":"","JCR":"Q2","JCRName":"COMPUTER SCIENCE, INFORMATION SYSTEMS","Score":null,"Total":0}
引用次数: 0

Abstract

Large AI models exhibit significant efficiency and precision in addressing complex problems. Despite their considerable advantages in various domains, these models encounter numerous challenges, notably high training costs. Currently, the training of distributed large AI models offers a solution to mitigate these elevated costs. However, distributed large AI models remain susceptible to data reconstruction attacks. A malicious server could leverage the intermediate results uploaded by clients to reconstruct the original data within the framework of distributed large AI models. This study first examines the underlying principles of data reconstruction attacks and proposes a privacy protection scheme. Our approach begins by obfuscating the mapping relationship between embeddings and the original data to ensure privacy protection. Specifically, during the upload of embedding data by clients to the server, genuine embeddings are concealed to prevent unauthorized access by malicious servers. Building on this concept, we introduce DMRP, a defensive mechanism featuring Dynamic Masking and Random Permutation, designed to mitigate data reconstruction attacks while maintaining the accuracy of the primary task. Our experiments, conducted across three models and four datasets, demonstrate the effectiveness of DMRP in countering data reconstruction attacks within distributed large-scale AI models.
DMRP:具有动态掩蔽和随机置换的隐私保护深度学习模型
大型人工智能模型在解决复杂问题时表现出显著的效率和精度。尽管这些模型在各个领域具有相当大的优势,但它们遇到了许多挑战,尤其是高昂的培训成本。目前,分布式大型人工智能模型的训练提供了一个解决方案,以减轻这些增加的成本。然而,分布式大型人工智能模型仍然容易受到数据重建攻击。恶意服务器可以利用客户端上传的中间结果,在分布式大型人工智能模型框架内重构原始数据。本研究首先考察了数据重构攻击的基本原理,并提出了一种隐私保护方案。我们的方法首先模糊嵌入和原始数据之间的映射关系,以确保隐私保护。具体来说,在客户端向服务器上传嵌入数据的过程中,隐藏了真实的嵌入,以防止恶意服务器的未经授权访问。在此概念的基础上,我们引入了DMRP,一种具有动态掩蔽和随机排列的防御机制,旨在减轻数据重构攻击,同时保持主要任务的准确性。我们在三个模型和四个数据集上进行的实验证明了DMRP在对抗分布式大规模人工智能模型中的数据重建攻击方面的有效性。
本文章由计算机程序翻译,如有差异,请以英文原文为准。
求助全文
约1分钟内获得全文 求助全文
来源期刊
Journal of Information Security and Applications
Journal of Information Security and Applications Computer Science-Computer Networks and Communications
CiteScore
10.90
自引率
5.40%
发文量
206
审稿时长
56 days
期刊介绍: Journal of Information Security and Applications (JISA) focuses on the original research and practice-driven applications with relevance to information security and applications. JISA provides a common linkage between a vibrant scientific and research community and industry professionals by offering a clear view on modern problems and challenges in information security, as well as identifying promising scientific and "best-practice" solutions. JISA issues offer a balance between original research work and innovative industrial approaches by internationally renowned information security experts and researchers.
×
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
确定
请完成安全验证×
copy
已复制链接
快去分享给好友吧!
我知道了
右上角分享
点击右上角分享
0
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 布克学术 All rights reserved.
京ICP备2023020795号-1
ghs 京公网安备 11010802042870号
Book学术文献互助
Book学术文献互助群
群 号:481959085
Book学术官方微信