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
  • [41] Weather and surface condition detection based on road-side webcams: Application of pre-trained Convolutional Neural Network
    Khan, Md Nasim
    Ahmed, Mohamed M.
    INTERNATIONAL JOURNAL OF TRANSPORTATION SCIENCE AND TECHNOLOGY, 2022, 11 (03) : 468 - 483
  • [42] Automatic Topic Labeling model with Paired-Attention based on Pre-trained Deep Neural Network
    He, Dongbin
    Ren, Yanzhao
    Khattak, Abdul Mateen
    Liu, Xinliang
    Tao, Sha
    Gao, Wanlin
    2021 INTERNATIONAL JOINT CONFERENCE ON NEURAL NETWORKS (IJCNN), 2021,
  • [43] Implementing heuristic-based multiscale depth-wise separable adaptive temporal convolutional network for ambient air quality prediction using real time data
    Sundaramoorthy, Raj Anand
    Ananth, Antony Dennis
    Seerangan, Koteeswaran
    Nandagopal, Malarvizhi
    Balusamy, Balamurugan
    Selvarajan, Shitharth
    SCIENTIFIC REPORTS, 2024, 14 (01):
  • [44] An effective stacked autoencoder based depth separable convolutional neural network model for face mask detection
    Balasubaramanian, Sundaravadivazhagan
    Cyriac, Robin
    Roshan, Sahana
    Paramasivam, Kulandaivel Maruthamuthu
    Jose, Boby Chellanthara
    ARRAY, 2023, 19
  • [45] Robust Finger Vein Presentation Attack Detection Using XceptionNet-based Modified Depthwise Separable Convolutional Neural Network
    Alsawalqah, Ahmad Atallah
    Rosdi, Bakhtiar Affendi
    JORDAN JOURNAL OF ELECTRICAL ENGINEERING, 2025, 11 (01): : 66 - 88
  • [46] Fine-tuning a pre-trained Convolutional Neural Network Model to translate American Sign Language in Real-time
    Cayamcela, Manuel Eugenio Morocho
    Lim, Wansu
    2019 INTERNATIONAL CONFERENCE ON COMPUTING, NETWORKING AND COMMUNICATIONS (ICNC), 2019, : 100 - 104
  • [47] Research on work-stress recognition for deep ground miners based on depth-separable convolutional neural network
    Chen, Ying
    Liu, Yuehan
    Lu, Caiwu
    Chai, Peishuo
    Li, Shengwei
    Zou, Yuan
    JOURNAL OF LOSS PREVENTION IN THE PROCESS INDUSTRIES, 2024, 91
  • [48] GDNet-EEG: An attention-aware deep neural network based on group depth-wise convolution for SSVEP stimulation frequency recognition
    Wan, Zhijiang
    Cheng, Wangxinjun
    Li, Manyu
    Zhu, Renping
    Duan, Wenfeng
    FRONTIERS IN NEUROSCIENCE, 2023, 17
  • [49] A Hybrid Neural Network BERT-Cap Based on Pre-Trained Language Model and Capsule Network for User Intent Classification
    Liu, Hai
    Liu, Yuanxia
    Wong, Leung-Pun
    Lee, Lap-Kei
    Hao, Tianyong
    COMPLEXITY, 2020, 2020
  • [50] Maize leaf disease detection using convolutional neural network: a mobile application based on pre-trained VGG16 architecture
    Paul, Hansamali
    Udayangani, Hirunika
    Umesha, Kalani
    Lankasena, Nalaka
    Liyanage, Chamara
    Thambugala, Kasun
    NEW ZEALAND JOURNAL OF CROP AND HORTICULTURAL SCIENCE, 2024,