DS-CNN: A pre-trained Xception model based on depth-wise separable convolutional neural network for finger vein recognition

被引:103
|
作者
Shaheed, Kashif [1 ]
Mao, Aihua [1 ]
Qureshi, Imran [2 ,3 ]
Kumar, Munish [4 ]
Hussain, Sumaira [5 ]
Ullah, Inam [5 ]
Zhang, Xingming [1 ]
机构
[1] South China Univ Technol, Sch Comp Sci & Engn, Guangzhou, Peoples R China
[2] Natl Univ Sci & Technol MCS NUST, Mil Coll Signals, Dept Comp Software Engn, Islamabad 44000, Pakistan
[3] Nanjing Univ Aeronaut & Astronaut, Coll Astronaut, Key Lab Space Photoelect Detect & Percept, Minist Ind & Informat Technol, Nanjing 211106, Jiangsu, Peoples R China
[4] Maharaja Ranjit Singh Punjab Tech Univ, Dept Computat Sci, Bathinda, India
[5] Shandong Univ, Sch Software Engn, TIME Lab, Jinan, Peoples R China
关键词
Biometric; Convolutional neural network; Classification; Deep learning; Finger vein recognition; Transfer learning; Xception; EXTRACTION; LOCALIZATION;
D O I
10.1016/j.eswa.2021.116288
中图分类号
TP18 [人工智能理论];
学科分类号
081104 ; 0812 ; 0835 ; 1405 ;
摘要
Finger vein recognition received special attention among all other biometric traits due to its high security. Adequate recognition and classification accuracy ensure the security of personal authentication. Many convolutional neural networks (CNNs) have been proposed with a promising performance in biometric finger vein recognition. However, their architectures have several problems, such as high complexity, extraction of robust features, degraded performance, etc. Considering the issues of CNNs, the authors present a pre-trained CNN network named Xception model based on depth-wise separable CNNs with residual connection, which is considered to be a more effective, less complex neural network to extract robust features. Our work can be seen as a three-stage process. Initially, the concept of data pre-processing is applied to convert the raw input samples into the standard format. Afterward, data augmentation using different geometrical techniques is incorporated to overcome the lack of training samples required for training the deep learning model. Finally, the feature extraction and classification task is performed through the pre-trained Xception architecture to verify the person's identity. SDUMLA and THU-FVFDT2 datasets are utilized to test and evaluate the proposed multi-layered CNN model performance with existing arts. Our proposed method for the SDUMLA database achieved an accuracy of 99% with an F1-score of 98%. While on THU-FVFDT2, the proposed method obtained an accuracy of 90% with an F1-score of 88%. Experimental results conclude that the proposed work obtained excellent performance compared to existing methods.
引用
收藏
页数:18
相关论文
共 50 条
  • [31] Finger Vein Recognition Based on Multi-Receptive Field Bilinear Convolutional Neural Network
    Wang, Kaixuan
    Chen, Guanghua
    Chu, Hongjia
    IEEE SIGNAL PROCESSING LETTERS, 2021, 28 : 1590 - 1594
  • [32] CURVATURE GRAY FEATURE DECOMPOSITION BASED FINGER VEIN RECOGNITION WITH AN IMPROVED CONVOLUTIONAL NEURAL NETWORK
    Zhao, Jia-Yi
    Gong, Jin
    Ma, Si-Teng
    Lu, Zhe-Ming
    INTERNATIONAL JOURNAL OF INNOVATIVE COMPUTING INFORMATION AND CONTROL, 2020, 16 (01): : 77 - 90
  • [33] Finger Vein Recognition Based on Multi-Receptive Field Bilinear Convolutional Neural Network
    Wang, Kaixuan
    Chen, Guanghua
    Chu, Hongjia
    IEEE Signal Processing Letters, 2021, 28 : 1590 - 1594
  • [34] Convolutional neural network based on multi-directional local coding for finger vein recognition
    Zhang, Zhongxia
    Zhou, Zhengchun
    Yang, Xue
    Meng, Hua
    Wu, Guohua
    INFORMATION SCIENCES, 2023, 623 : 633 - 647
  • [35] Convolutional Neural Network-based Finger Vein Recognition using Near Infrared Images
    Fairuz, Subha
    Habaebi, Mohamed Hadi
    An, Elsheikh Mohamed Ahmed Elsheikh
    Chebil, Jalel
    PROCEEDINGS OF THE 2018 7TH INTERNATIONAL CONFERENCE ON COMPUTER AND COMMUNICATION ENGINEERING (ICCCE), 2018, : 453 - 458
  • [36] Finger vein recognition scheme based on convolutional neural network using curvature gray image
    Zhao, Jia-Yi
    Gong, Jin
    Ma, Si-Teng
    Lu, Zhe-Ming
    Chu, Shu-Chuan
    Roddick, John F.
    Journal of Network Intelligence, 2019, 4 (03): : 114 - 123
  • [37] Lithography Hotspot Detection Method Based on Transfer Learning Using Pre-Trained Deep Convolutional Neural Network
    Liao, Lufeng
    Li, Sikun
    Che, Yongqiang
    Shi, Weijie
    Wang, Xiangzhao
    APPLIED SCIENCES-BASEL, 2022, 12 (04):
  • [38] Pre-Trained Deep Convolutional Neural Network for Clostridioides Difficile Bacteria Cytotoxicity Classification Based on Fluorescence Images
    Brodzicki, Andrzej
    Jaworek-Korjakowska, Joanna
    Kleczek, Pawel
    Garland, Megan
    Bogyo, Matthew
    SENSORS, 2020, 20 (23) : 1 - 17
  • [39] The Effect of Pre-processing on a Convolutional Neural Network Model for Dorsal Hand Vein Recognition
    Tarawneh, Omar
    Saber, Qotadeh
    Almaghthawi, Ahmed
    Abu Owida, Hamza
    Issa, Abedalhakeem
    Alshdaifat, Nawaf
    Jaradat, Ghaith
    Abuowaida, Suhaila
    Arabiat, Mohammad
    INTERNATIONAL JOURNAL OF ADVANCED COMPUTER SCIENCE AND APPLICATIONS, 2024, 15 (03) : 1284 - 1289
  • [40] Convolutional Neural Network-Based Finger-Vein Recognition Using NIR Image Sensors
    Hong, Hyung Gil
    Lee, Min Beom
    Park, Kang Ryoung
    SENSORS, 2017, 17 (06)