基于行稀疏的局部约束矩阵回归模型的人脸识别研究
Research on a Row-Sparse Local Constraint Matrix Regression Model in Face Recognition
DOI: 10.12677/aam.2024.133092, PDF,   
作者: 尹梓锦*, 张凌聪：河北工业大学理学院，天津
关键词: 人脸识别；矩阵回归；遮挡图像；ADMM；Face Recognition； Matrix Regression； Occluded Images； ADMM
摘要: 人脸识别是图像处理领域中一个经典而重要的问题。关于人脸识别有许多研究课题。在这些课题中，处理连续遮挡或光照变化是目前最具挑战性的问题之一。针对无遮挡与大面积遮挡下的人脸识别，本文设计基于行稀疏的局部约束矩阵回归模型(RSLMR)。该模型对误差矩阵施加l2,1-范数约束，而不是对误差矩阵直接施加核范数约束。此外，RSLMR利用测试图像内部的相似性添加关于重建图像的正则化项，还利用距离信息对类进行加权。此外，RSLMR通过减小表示系数相邻值的差异，迫使训练样本对应的系数值相近，以此利用类内所有训练样本的协作关系。最后，采用交替方向乘子法(ADMM)来求解所提出的模型。在FERET、ORL与CMU_PIE人脸数据库上的实验结果展示了RSLMR的有效性。与其他回归模型相比，即使存在大面积的遮挡区域，RSLMR也能保持较高的识别率。
Abstract: Face recognition is a classical and important problem in the field of image processing. There are numerous research topics related to face recognition. Among these topics, handling continuous occlusion or lighting variations is currently one of the most challenging problems. In addressing the issue of face recognition with extensive occlusions or no occlusions, this study proposes a row-sparse local constraint matrix regression model (RSLMR). Instead of directly imposing a nuclear norm constraint on the error matrix, this model applies an l2,1-norm constraint. Additionally, RSLMR incorporates regularization terms based on the internal similarity of test images and weights classes using distance information. Furthermore, RSLMR encourages the coefficients corresponding to training samples to be similar by minimizing the differences between adjacent coefficient values, thereby leveraging the collaborative relationship among all training samples within a class. Finally, the proposed model is solved using the alternating direction method of multipliers (ADMM). Experimental results on the FERET, ORL, and CMU_PIE face databases demonstrate the effectiveness of RSLMR. Compared to other regression models, RSLMR maintains a high recognition rate even in the presence of large occluded regions.
文章引用：尹梓锦, 张凌聪. 基于行稀疏的局部约束矩阵回归模型的人脸识别研究[J]. 应用数学进展, 2024, 13(3): 981-990. https://doi.org/10.12677/aam.2024.133092

参考文献

[1] Zhao, W., Chellappa, R., Phillips, P.J., et al. (2003) Face Recognition: A Literature Survey. ACM Computing Surveys (CSUR), 35, 399-458. [Google Scholar] [CrossRef] 
[2] Yang, M., Huang, F. and Lv, X. (2019) A Feature Learning Approach for Face Recognition with Robustness to Noisy Label Based on Top-N Prediction. Neurocomputing, 330, 48-55. [Google Scholar] [CrossRef] 
[3] Zhao, Y., Liu, Y., Shen, C., et al. (2020) MobileFAN: Transferring Deep Hidden Representation for Face Alignment. Pattern Recognition, 100, 107-114. [Google Scholar] [CrossRef] 
[4] Ou, W., You, X., Tao, D., et al. (2014) Robust Face Recognition via Occlusion Dictionary Learning. Pattern Recognition, 47, 1559-1572. [Google Scholar] [CrossRef] 
[5] Lai, Z., Xu, Y., Chen, Q., et al. (2014) Multilinear Sparse Principal Component Analysis. IEEE Transactions on Neural Networks and Learning Systems, 25, 1942-1950. [Google Scholar] [CrossRef] 
[6] Wang, H., Wang, Y., Zhou, Z., et al. (2018) CosFace: Large Margin Cosine Loss for Deep Face Recognition. Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition, Salt Lake City, 18-23 June 2018, 5265-5274. [Google Scholar] [CrossRef] 
[7] Naseem, I., Togneri, R. and Bennamoun, M. (2010) Linear Regression for Face Recognition. IEEE Transactions on Pattern Analysis and Machine Intelligence, 32, 2106-2112. [Google Scholar] [CrossRef] 
[8] Wright, J., Yang, A.Y., Ganesh, A., et al. (2008) Robust Face Recognition via Sparse Representation. IEEE Transactions on Pattern Analysis and Machine Intelligence, 31, 210-227. [Google Scholar] [CrossRef] 
[9] Zhang, L., Yang, M. and Feng, X. (2011) Sparse Representation or Collaborative Representation: Which Helps Face Recognition? 2011 International Conference on Computer Vision, 471-478.
[10] Huang, J., Nie, F., Huang, H., et al. (2013) Supervised and Projected Sparse Coding for Image Classification. Proceedings of the AAAI Conference on Artificial Intelligence, 27, 438-444. [Google Scholar] [CrossRef] 
[11] Naseem, I., Togneri, R. and Bennamoun, M. (2012) Robust Regression for Face Recognition. Pattern Recognition, 45, 104-118. [Google Scholar] [CrossRef] 
[12] Yang, M., Zhang, L., Yang, J., et al. (2012) Regularized Robust Coding for Face Recognition. IEEE Transactions on Image Processing, 22, 1753-1766. [Google Scholar] [CrossRef] 
[13] Lai, J. and Jiang, X. (2016) Classwise Sparse and Collaborative Patch Representation for Face Recognition. IEEE Transactions on Image Processing, 25, 3261-3272. [Google Scholar] [CrossRef] 
[14] Yang, J., Luo, L., Qian, J., et al. (2016) Nuclear Norm Based Matrix Regression with Applications to Face Recognition with Occlusion and Illumination Changes. IEEE Transactions on Pattern Analysis and Machine Intelligence, 39, 156-171. [Google Scholar] [CrossRef] 
[15] Chen, Z., Wu, X.J. and Kittler, J. (2019) A Sparse Regularized Nuclear Norm Based Matrix Regression for Face Recognition with Contiguous Occlusion. Pattern Recognition Letters, 125, 494-499. [Google Scholar] [CrossRef] 
[16] Qian, J., Yang J., Xu, Y., et al. (2020) Image Decomposition Based Matrix Regression with Applications to Robust Face Recognition. Pattern Recognition, 102, 107204. [Google Scholar] [CrossRef] 
[17] Zhang, C., Li, H., Qian, Y., et al. (2020) Locality-Constrained Discriminative Matrix Regression for Robust Face Identification. IEEE Transactions on Neural Networks and Learning Systems, 33, 1254-1268. [Google Scholar] [CrossRef] 
[18] Li, Q., He, H., Lai, H., et al. (2022) Enhanced Nuclear Norm Based Matrix Regression for Occluded Face Recognition. Pattern Recognition, 126, 108585. [Google Scholar] [CrossRef] 

为你推荐