Protection against adversarial attacks with randomization of recognition algorithm

IF 1.5 Q3 COMPUTER SCIENCE, INFORMATION SYSTEMS
Grigory Marshalko, Svetlana Koreshkova
{"title":"Protection against adversarial attacks with randomization of recognition algorithm","authors":"Grigory Marshalko, Svetlana Koreshkova","doi":"10.1007/s11416-023-00503-z","DOIUrl":null,"url":null,"abstract":"We study a randomized variant of one type of biometric recognition algorithms, which is intended to mitigate adversarial attacks. We show that the problem of an estimation of the security of the proposed algorithm can be formulated in the form of an estimation of statistical distance between the probability distributions, induced by the initial and the randomized algorithm. A variant of practical password-based implementation is discussed. The results of experimental evaluation are given. The preliminary verison of this research was presented at CTCrypt 2020 workshop.","PeriodicalId":15545,"journal":{"name":"Journal of Computer Virology and Hacking Techniques","volume":null,"pages":null},"PeriodicalIF":1.5000,"publicationDate":"2023-10-05","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":"0","resultStr":null,"platform":"Semanticscholar","paperid":null,"PeriodicalName":"Journal of Computer Virology and Hacking Techniques","FirstCategoryId":"1085","ListUrlMain":"https://doi.org/10.1007/s11416-023-00503-z","RegionNum":0,"RegionCategory":null,"ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":null,"EPubDate":"","PubModel":"","JCR":"Q3","JCRName":"COMPUTER SCIENCE, INFORMATION SYSTEMS","Score":null,"Total":0}
引用次数: 0

Abstract

We study a randomized variant of one type of biometric recognition algorithms, which is intended to mitigate adversarial attacks. We show that the problem of an estimation of the security of the proposed algorithm can be formulated in the form of an estimation of statistical distance between the probability distributions, induced by the initial and the randomized algorithm. A variant of practical password-based implementation is discussed. The results of experimental evaluation are given. The preliminary verison of this research was presented at CTCrypt 2020 workshop.
基于随机识别算法的对抗性攻击防护
我们研究了一种生物识别算法的随机变体,旨在减轻对抗性攻击。我们证明了该算法的安全性估计问题可以用初始算法和随机算法引起的概率分布之间的统计距离估计的形式来表示。讨论了一种实用的基于密码的实现方法。给出了实验评价结果。这项研究的初步版本在CTCrypt 2020研讨会上发表。
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 Computer Virology and Hacking Techniques
Journal of Computer Virology and Hacking Techniques COMPUTER SCIENCE, INFORMATION SYSTEMS-
CiteScore
4.00
自引率
13.30%
发文量
41
期刊介绍: The field of computer virus prevention has rapidly taken an important position in our technological and information society. Viral attacks increase year after year, and antiviral efforts continually face new challenges. Beneficial applications of technologies based on scientific computer virology are still very limited. The theoretical aspects of the virus problem are only rarely considered, although many interesting and important open problems still exist. Little proactive research is focused on predicting the future of viral attacks.The Journal of Computer Virology and Hacking Techniques is an independent scientific and technical journal dedicated to viral and antiviral computer technologies. Both theoretical and experimental aspects will be considered; papers emphasizing the theoretical aspects are especially welcome. The topics covered by this journal include, but are certainly not limited to:- Mathematical aspects and theoretical fundamentals of computer virology - Algorithmics and computer virology - Computer immunology and biological models for computers - Reverse engineering (hardware and software) - Viral  and antiviral technologies - Cryptology and steganography tools and techniques - Applications in computer virology - Virology and IDS - Hardware hacking, and free and open hardware - Operating system, network, and embedded systems security - Social engineeringIn addition, since computational problems are of practical interest, papers on the computational aspects of computer virology are welcome. It is expected that the areas covered by this journal will change as new technologies, methodologies, challenges and applications develop. Hacking involves understanding technology intimately and in depth in order to use it in an operational way. Hackers are complementary to academics in that they favour the result over the methods and over the theory, while academics favour the formalization and the methods -- explaining is not operating and operating is not explaining. The aim of the journal in this respect is to build a bridge between the two communities for the benefit of technology and science.The aim of the Journal of Computer Virology and Hacking Techniques is to promote constructive research in computer virology by publishing technical and scientific results related to this research area. Submitted papers will be judged primarily by their content, their originality and their technical and scientific quality. Contributions should comprise novel and previously unpublished material.However, prior publication in conference proceedings of an abstract, summary, or other abbreviated, preliminary form of the material should not preclude publication in this journal when notice of such prior or concurrent publication is given with the submission. In addition to full-length theoretical and technical articles, short communications or notes are acceptable. Survey papers will be accepted with a prior invitation only. Special issues devoted to a single topic are also planned.The policy of the journal is to maintain strict refereeing procedures, to perform a high quality peer-review of each submitted paper, and to send notification to the author(s) with as short a delay as possible. Accepted papers will normally be published within one year of submission at the latest. The journal will be published four times a year. Note: As far as new viral techniques are concerned, the journal strongly encourages authors to consider algorithmic aspects rather than the actual source code of a particular virus. Nonetheless, papers containing viral source codes may be accepted provided that a scientific approach is maintained and that inclusion of the source code is necessary for the presentation of the research. No paper containing a viral source code will be considered or accepted unless the complete source code is communicated to the Editor-in-Chief. No publication will occur before antiviral companies receive this source code to update/upgrade their products.The final objective is, once again, proactive defence.This journal was previously known as Journal in Computer Virology. It is published by Springer France.
×
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