Research on Gait Recognition Method Based on LFE-MGPA Network
Keywords:
Gait recognition; Convolutional neural network; Feature extraction; Gait contour diagram.Abstract
As a non-contact long-distance biometric recognition technology, gait recognition is tasked to realize the identification of long-distance pedestrians according to their walking patterns. However, gait recognition is greatly affected by external factors such as what the subject is wearing and carrying. To solve this problem, this paper focuses on gait feature extraction and proposes LFE-MGPA network based on convolutional neural network (CNN). A local feature extractor (LFE) was constructed to extract fine-grained features of gait through SConv, and MGPA module was integrated to extract collection-level information of different layers. In order to improve the recognition accuracy of the network model, a joint loss function is constructed. The experiment was carried out on the open gait dataset CASIA-B, and the recognition accuracy reached 97.53%, 94.5% and 81.0% respectively under the three walking states of normal walking, backpack and overcoat, which is higher than the current mainstream algorithms. In particular, the recognition accuracy is significantly improved in the walking state of backpack and overcoat, which proves that the method has strong robustness.
Downloads
References
Zeng F, Hu S, Xiao, K. Research on partial fingerprint recognition algorithm based on deep learning[J]. Neural computing & applications,2019,31(9):4789-4798.
Sepasmoghaddam A , Pereira F M , Correia P L .Face Recognition: A Novel Multi-Level Taxonomy based Survey[J].IET Biometrics, 2019, 9(2).
Sun Y, Wang X, Tang X. Deep Learning Face Representation from Predicting 10,000 Classes[J]. 2014 IEEE Conference on Computer Vision and Pattern Recognition, Columbus, OH, USA, 2014: 1891–1898.
Wang K , Kumar A .Periocular-Assisted Multi-Feature Collaboration for Dynamic Iris Recognition[J].IEEE Transactions on Information Forensics and Security, 2021, 16:866-879.
Han J, Bhanu B. Individual recognition using gait energy image[J]. IEEE Transactions on Pattern Analysis and Machine Intelligence, 2006, 28(2): 316-322.
Wu Z, Huang Y, Wang L, et al. A Comprehensive Study on Cross-View Gait Based Human Identification with Deep CNNs[J]. IEEE Transactions on Pattern Analysis and Machine Intelligence, 2017, 39(2): 209-226.
Wang C, Zhang J, Wang L, et al. Human identification using temporal information preserving gait template[J]. IEEE transactions on pattern analysis and machine intelligence, 2012, 34(11): 2164-2176.
Alireza S, Ali E. Deep Gait Recognition: A Survey[J]. IEEE transactions on pattern analysis and machine intelligence, 2022, 45(1): 264-284.
Shiraga K, Makihara Y, Muramatsu D, et al. GEINet: View-invariant gait recognition using a convolutional neural network[C]. Proceedings of the 2016 International Conference on Biometrics (ICB), 2016: 1-8.
Fan C, Peng Y, Cao C, et al. GaitPart: Temporal Part-Based Model for Gait Recognition[J]. Proceedings of the IEEE Computer Society Conference on Computer Vision and Pattern Recognition, 2020: 14213-14221.
Sepas-Moghaddam A, Ghorbani S, Troje NF, et al. Gait recognition using multi-scale partial representation transformation with capsules[C]. Proceedings of the 2020 25th international conference on pattern recognition (ICPR), 2021, 8045-8052.
Chao H, Wang K, He Y, et al. GaitSet: Cross-view Gait Recognition through Utilizing Gait as a Deep Set[J]. IEEE Transactions on Pattern Analysis and Machine Intelligence, 2021: 1-12.
Liu N, Lu J, Tan Y. Joint subspace learning for view-invariant gait recognition[J]. IEEE Signal Process. Lett, 2011, 18(7): 431-434.
Jean F, Albu AB, Bergevin R. Towards view-invariant gait modeling: Computing view-normalized body part trajectories[J]. Pattern Recognition, 2009, 42(11): 2936-2949.
Goffredo M, Bouchrika I, Carter JN, et al. Self-calibrating view-invariant gait biometrics[J]. IEEE Transactions on Systems, Man, and Cybernetics, Part B (Cybernetics), 2010, 40 (4): 997-1008.
Yu S, Tan D, Tan T. A framework for evaluating the effect of view angle, clothing and carrying condition on gait recognition[C]. Proceedings of the 18th international conference on pattern recognition (ICPR'06), 2006: 441-444.
Hou S , Liu X , Cao C ,et al.Set Residual Network for Silhouette-Based Gait Recognition[J].IEEE Transactions on Biometrics, Behavior, and Identity Science, 2021(3-3).DOI:10.1109/TBIOM.2021.3074963.
Downloads
Published
Issue
Section
License
Copyright (c) 2024 Academic Journal of Science and Technology
This work is licensed under a Creative Commons Attribution 4.0 International License.
