Improvement of Face Recognition Algorithm for Residential Area Surveillance System Based on Graph Convolution Network
0
조회
0
다운로드
0
인용
0
참고문헌
키워드
초록
The construction of smart communities is a new method and important measure to ensure the security of residential areas. In order to solve the problem of low accuracy in face recognition caused by distorting facial features due to monitoring camera angles and other external factors, this paper proposes the following optimization strategies in designing a face recognition network: firstly, a global graph convolution module is designed to encode facial features as graph nodes, and a multi-scale feature enhancement residual module is designed to extract facial keypoint features in conjunction with the global graph convolution module. Secondly, after obtaining facial keypoints, they are constructed as a directed graph structure, and graph attention mechanisms are used to enhance the representation power of graph features. Finally, tensor computations are performed on the graph features of two faces, and the aggregated features are extracted and discriminated by a fully connected layer to determine whether the individuals' identities are the same. Through various experimental tests, the network designed in this paper achieves an AUC index of 85.65% for facial keypoint localization on the 300W public dataset and 88.92% on a self-built dataset. In terms of face recognition accuracy, the proposed network achieves an accuracy of 83.41% on the IBUG public dataset and 96.74% on a self-built dataset. Experimental results demonstrate that the network designed in this paper exhibits high detection and recognition accuracy for faces in surveillance videos.
Abstract
The construction of smart communities is a new method and important measure to ensure the security of residential areas. In order to solve the problem of low accuracy in face recognition caused by distorting facial features due to monitoring camera angles and other external factors, this paper proposes the following optimization strategies in designing a face recognition network: firstly, a global graph convolution module is designed to encode facial features as graph nodes, and a multi-scale feature enhancement residual module is designed to extract facial keypoint features in conjunction with the global graph convolution module. Secondly, after obtaining facial keypoints, they are constructed as a directed graph structure, and graph attention mechanisms are used to enhance the representation power of graph features. Finally, tensor computations are performed on the graph features of two faces, and the aggregated features are extracted and discriminated by a fully connected layer to determine whether the individuals' identities are the same. Through various experimental tests, the network designed in this paper achieves an AUC index of 85.65% for facial keypoint localization on the 300W public dataset and 88.92% on a self-built dataset. In terms of face recognition accuracy, the proposed network achieves an accuracy of 83.41% on the IBUG public dataset and 96.74% on a self-built dataset. Experimental results demonstrate that the network designed in this paper exhibits high detection and recognition accuracy for faces in surveillance videos.
참고문헌 (23)
- J. Zhang, "Research on Face Recognition Algorithm with Fusion Attention Mechanism", Foreign Electronic Measurement Technology, 42(02), pp.107~113 (2023)
- Y. H. Fan, "Low-light Image Enhancement Algorithm Driven by Face Recognition Task", Journal of Graphics, 43(06), pp.1170~1181 (2022)
- J. Feng, "Improving MobileNetv3 Face Recognition Algorithm in Classroom Scenarios", Foreign Electronic Measurement Technology, 41(10), pp.47~55 (2022)
- Boutros F, "Selfrestrained triplet loss for accurate masked face recognition", Pattern Recognition, 124, pp.108473~ (2022)[DOI]
- Chen Z, "A lightweight CNN-based algorithm and implementation on embedded system for real-time face recognition", Multimedia systems (2023)
- T. C. Zhu, "A comprehensive review of deep learning-based methods for face recognition", Modern Computer, 29(17), pp.36~40 (2023)
- L. Di, "Face alignment combined with shape constraints and Gaussian heatmap", International Journal of Machine Learning and Cybernetics, 14(12), pp.4311~4324 (2023)
- S. Y. Guan, "Face feature point localization algorithm with adaptive weighting of occlusion", Journal of Miniaturization of Microsystem, 44(12), pp.2773~2783 (2023)
- Y. Q. li, "Face alignment and reconstruction algorithm based on encoder-decoder network", Journal of Computer System Applications, 30(07), pp.184~189 (2021)
- L. M. Han, "Dynamic deformable transformer for end-to-end face alignment", IET Computer Vision, 17(8), pp.948~961 (2023)
- X. Q. Yang, "End-to-end face alignment algorithm based on deep convolutional neural network", Computer Engineering and Design, 40(09), pp.2666~2671 (2019)
- Y. H. Li, "Accurate face alignment algorithm based on K-means", Sensors and Microsystems, 40(03), pp.120~122 (2021)
- J. Q. Liu, "A comprehensive review of graph convolutional neural networks", Journal of Computer Engineering and Science, 45(08), pp.1472~1481 (2023)
- W. J. Li, "A comprehensive review of graph convolutional neural networks and their applications in image recognition", Journal of Computer Engineering and Applications, 59(22), pp.15~35 (2023)
- X. X. Zhang, "A comprehensive review of facial keypoint detection", Journal of Computer Engineering and Applications, pp.1~13 (2023)
- Jin H, "Pixel-in-Pixel Net: Towards Efficient Facial Landmark Detection in the Wild", International Journal of Computer Vision, 12(12), pp.129~ (2021)[DOI]
- Xu Z, "AnchorFace : An Anchor-based Facial Landmark Detector Across Large Poses", National Conference on Artificial Intelligence (2021)
- Schroff F, "FaceNet:A Unified Embedding for Face Recognition and Clustering", IEEE (2015)
- Wang H, "CosFace : Large Margin Cosine Loss for Deep Face Recognition", computer vision and pattern recognition, pp.5265~5274 (2018)