计算机科学 ›› 2024, Vol. 51 ›› Issue (6A): 230800043-9.doi: 10.11896/jsjkx.230800043

• 信息安全 • 上一篇    下一篇

面向公平性联邦学习的指纹识别算法

王晨卓1, 鲁艳蓉1,2, 沈剑3   

  1. 1 中国民航大学计算机科学与技术学院 天津 300300
    2 中国民航大学安全科学与工程学院 天津 300300
    3 浙江理工大学信息科学与工程学院 杭州 310018
  • 发布日期:2024-06-06
  • 通讯作者: 鲁艳蓉(yr_lu@cauc.edu.cn)
  • 作者简介:(2021051015@cauc.edu.cn)
  • 基金资助:
    国家重点研发计划(2023YFB4302901,2023YFB2703700);国家自然科学基金(61802276,62172418,U2133205,U21A20465);天津市教委科研计划项目(2021KJ038);浙江理工大学科学基金项目(22222266Y)

Study on Fingerprint Recognition Algorithm for Fairness in Federated Learning

WANG Chenzhuo1, LU Yanrong1,2, SHEN Jian3   

  1. 1 School of Computer Science and Technology,Civil Aviation University of China,Tianjin 300300,China
    2 School of Safety Science and Engineering,Civil Aviation University of China,Tianjin 300300,China
    3 School of Informatics Science and Engineering,Zhejiang Sci-Tech University,Hangzhou 310018,China
  • Published:2024-06-06
  • About author:WANG Chenzhuo,born in 1999,postgraduate.Her main research interest is federated learning.
    LU Yanrong,born in 1985,Ph.D,asso-ciate professor.Her main research in-terests include future cybersecurity,AI security and blockchain technology.
  • Supported by:
    National Key Research and Development Program of China(2023YFB4302901,2023YFB2703700),National Na-tural Science Foundation of China(61802276,62172418,U2133205,U21A20465),Scientific Research Project of Tianjin Educational Committee(2021KJ038) and Science Foundation of Zhejiang Sci-Tech University(ZSTU)(22222266Y).

摘要: 现有的指纹识别方法大多是基于机器学习,在对海量数据集中训练时忽视了数据本身的隐私性和异质性,从而导致用户信息泄漏和识别率降低。为在隐私保护下协同优化模型精度,提出了一个全新的基于联邦学习的指纹识别算法(Federated Learning-Fingerprint Recognition,Fed-FR)。首先,通过联邦学习迭代聚合来自各终端的参数,从而提高全局模型的性能;其次,将稀疏表示理论用于低质量指纹图像去噪处理,来增强指纹的纹理结构;再次,针对客户端异构而导致的分配不公问题,提出基于水库抽样的客户端调度策略;最后,在3个真实数据集上进行仿真实验,对Fed-FR的有效性进行对比分析。实验结果表明,Fed-FR精度比局部学习提高5.32%,比联邦平均算法提高8.56%,接近于集中学习的精度;在隐私保护水平、评估准确率及可扩展性等方面具有良好的表现。研究成果首次展现了联邦学习与指纹识别结合的可行性,增强了指纹识别算法的安全性和可扩展性,给联邦学习应用于生物识别技术提供了参考。

关键词: 指纹识别, 联邦学习, 稀疏表示, 水库抽样, 隐私保护

Abstract: Most existing fingerprint recognition methods rely on machine learning,which neglects the privacy and heterogeneity of the data when training on massive databases,resulting in user information leakage and reduced recognition accuracy.To cooperatively optimize model accuracy under privacy protection,this paper proposes a novel fingerprint recognition algorithm based on federated learning,termed federated learning-fingerprint recognition(Fed-FR).Firstly,the algorithm iteratively aggregates parameters from each terminal through federated learning,thereby improving the performance of the global model.Secondly,sparse representation theory is applied to low-quality fingerprint image denoising to enhance the texture structure of the fingerprint.Thirdly,in response to the allocation inequity issue caused by client heterogeneity,this paper proposes a client scheduling strategy based on reservoir sampling.Finally,experimental results on three real-world databases show that Fed-FR significantly outperforms local learning by 5.32% and federated average by 8.56%,approaching the accuracy of centralized learning.The results demonstrate the effectiveness of Fed-FR in privacy protection,accuracy evaluation,and scalability.This study demonstrates for the first time the feasibility of combining federated learning with fingerprint recognition,enhancing the security and scalability of fingerprint recognition algorithms,and providing a reference for the application of federated learning in biometric technologies.

Key words: Fingerprint recognition, Federated learning, Sparse representation, Reservoir sampling, Privacy protection

中图分类号: 

  • TP391
[1]HU Y,HU A,LI C,et al.Towards a privacy protection-capable noise fingerprinting for numerically aggregated data[J].Computers & Security,2022,119(8):102755.
[2]ZHANG F,FENG J.High-resolution mobile fingerprint matching via deep joint KNN-triplet embedding[C]//AAAI Confe-rence on Artificial Intelligence.2017:5019-2020.
[3]NOGUEIRA R F,DE ALENCAR LOTUFO R,MACHADO R C.Fingerprint liveness detection using convolutional neural networks[J].IEEE Transactions on Information Forensics and Security,2016,11(6):1206-1213.
[4]SUN S,CAO Z,ZHU H,et al.A survey of optimization methods from a machine learning perspective[J].IEEE Transactions on Cybernetics,2019,50(8):3668-3681.
[5]ZHAO Y,LI M,LAI L,et al.Federated learning with non-iiddata[J].arXiv:1806.00582,2018.
[6]KAIROUZ P,MCMAHAN H B,AVENT B,et al.Advances and open problems in federated learning[J].Foundations and Trends © in Machine Learning,2021,14(1/2):1-210.
[7]MCMAHAN B,MOORE E,RAMAGE D,et al.Communica-tion-efficient learning of deep networks from decentralized data[C]//Artificial Intelligence and Statistics.PMLR,2017:1273-1282.
[8]COETZEE L,BOTHA E C.Fingerprint recognition in low qua-lity images[J].Pattern recognition,1993,26(10):1441-1460.
[9]LI K H.Reservoir-sampling algorithms of time complexity O(n(1+ log(N/n)))[J].ACM Transactions on Mathematical Software,1994,20(4):481-493.
[10]CAPPELLI R,FERRARA M,MALTONI D.Minutia cylinder-code:A new representation and matching technique for fingerprint recognition[J].IEEE Transactions on Pattern Analysis and Machine Intelligence,2010,32(12):2128-2141.
[11]LEE W,CHO S,CHOI H,et al.Partial fingerprint matchingusing minutiae and ridge shape features for small fingerprint scanners[J].Expert Systems with Applications,2017,87(6):183-198.
[12]ALAM B,JIN Z,YAP W S,et al.An alignment-free cancelable fingerprint template for bio-cryptosystems[J].Journal of Network and Computer Applications,2018,115(4):20-32.
[13]WANG T,ZHENG Z,BASHIR A K,et al.FinPrivacy:A privacy-preserving mechanism for fingerprint identification[J].ACM Transactions on Internet Technology,2021,21(3):1-15.
[14]ZHANG Z,LIU S,LIU M.A multi-task fully deep convolutional neural network for contactless fingerprint minutiae extraction[J].Pattern Recognition,2021,120(12):108189.
[15]CAPPELLI R,FERRARA M,MALTONI D.Minutia cylinder-code:A new representation and matching technique for fingerprint recognition[J].IEEE Transactions on Pattern Analysis and Machine Intelligence,2010,32(12):2128-2141.
[16]ZHAO J,CHANG X,FENG Y,et al.Participant selection forfederated learning with heterogeneous data in intelligent transport system[J].IEEE Transactions on Intelligent Transportation Systems,2022,24(1):1-10.
[17]ZHAO J,CHANG X,FENG Y,et al.Participant selection forfederated learning with heterogeneous data in intelligent transport system[J].IEEE Transactions on Intelligent Transportation Systems,2022,24(1):1-10.
[18]WANG H,KAPLAN Z,NIU D,et al.Optimizing federatedlearning on non-iid data with reinforcement learning[C]//IEEE INFOCOM 2020-IEEE Conference on Computer Communications.2020:1698-1707.
[19]LI X,ZHANG X,HUANG W,et al.Truncation cross entropy loss for remote sensing image captioning[J].IEEE Transactions on Geoscience and Remote Sensing,2020,59(6):5246-5257.
[20]SHEHU Y I,RUIZ-GARCIA A,PALADE V,et al.Sokoto co-ventry fingerprint dataset[J].arXiv:1807.10609,2018.
[21]CAPPELLI R,MAIO D,MALTONI D.Synthetic fingerprint-database generation[C]//International Conference on Pattern Recognition.2002:744-747.
[22]MAIO D,MALTONI D,CAPPELLI R,et al.FVC2004:Thirdfingerprint verification competition[C]//Biometric Authentication:First International Conference.2004:1-7.
[23]DA CUNHA A L,ZHOU J,DO M N.The nonsubsampled contourlet transform:theory,design,and applications[J].IEEE Transactions on Image Processing,2006,15(10):3089-3101.
[24]LI T,SAHU A K,ZAHEER M,et al.Federated optimization in heterogeneous networks[C]//Machine Learning and Systems.MLSYS,2020:429-450.
[25]NOGUEIRA R F,DE ALENCAR LOTUFO R,MACHADO R C.Fingerprint liveness detection using convolutional neural networks[J].IEEE Transactions on Information Forensics and Security,2016,11(6):1206-1213.
[26]ULIYAN D M,SADEGHI S,JALAB H A.Anti-spoofing methodfor fingerprint recognition using patch based deep learning machine[J].Engineering Science and Technology an International Journal,2020,23(2):264-273.
[27]PATIL S R,SURALKAR S R.Neural network based fingerprint classification[J].International Journal of Science and Research,2013,2(1):58-62.
[28]KOUAMO S,TANGHA C.Fingerprint recognition with artificial neural networks:application to e-learning[J].Journal of Intelligent Learning Systems and Applications,2016,8(2):39-49.
Viewed
Full text


Abstract

Cited

  Shared   
  Discussed   
No Suggested Reading articles found!