文献类型：期刊文献

英文题名：Cancelable HD-SEMG Biometric Identification via Deep Feature Learning

作者：Fan, Jiahao[1,2];Jiang, Xinyu[1];Liu, Xiangyu[3];Zhao, Xian[1];Ye, Xinming[3];Dai, Chenyun[1];Akay, Metin[4];Chen, Wei[1]

机构：[1]Fudan Univ, Sch Informat Sci & Technol, Ctr Intelligent Med Elect, Shanghai 200433, Peoples R China;[2]Fudan Univ, Human Phenome Inst, Shanghai 201203, Peoples R China;[3]East China Univ Sci & Technol, Sch Art Design & Media, Shanghai 200237, Peoples R China;[4]Univ Houston, Dept Biomed Engn, Houston, TX 77204 USA

年份：2022

卷号：26

期号：4

起止页码：1782

外文期刊名：IEEE JOURNAL OF BIOMEDICAL AND HEALTH INFORMATICS

收录：;EI(收录号：20214511139831);WOS:【SCI-EXPANDED(收录号:WOS:000803121800040)】；

基金：This work was supported in part by the National Key R&D Program of China under Grant 2017YFE0112000, in part by Shanghai Municipal Science and Technology International R&D Collaboration Project under Grant 20510710500, in part by Shanghai Pujiang Program under Grant 19PJ1401100, and in part by Shanghai Natural Science Foundation under Grant 20ZR1403400.

语种：英文

外文关键词：Biometrics (access control); Task analysis; Biological system modeling; Brain modeling; Bioinformatics; Face recognition; Password; Person identification; biometrics; cancelability; high-density sEMG; feature learning

摘要：Conventional biometric modalities, such as the face, fingerprint, and iris, are vulnerable against imitation and circumvention. Accordingly, secure biometric modalities with cancelable properties are needed for personal identification, especially in smart healthcare applications. Here we developed a person identification model using high-density surface electromyography (HD-sEMG) as biometric traits. In this model, the HD-sEMG biometric templates are cancelable and could be customized by the users through finger isometric contractions. A deep feature learning approach, implemented by convolutional neural networks (CNNs) is used to capture user-specific patterns from HD-sEMG signals and make identification decisions. This model has been validated on twenty-two subjects, with training and testing data acquired from two different days. The rank-1 identification accuracy and equal error rate for 44 identities (22 subjects x 2 accounts) can reach 87.23% and 4.66%, respectively. The cross-day identification accuracy of the proposed model is higher than the results of previous methods reported in the literature. The usability and efficiency of the proposed model are also investigated, indicating its potentials for practical applications.