三角稀疏回归分类器及其在稳健人脸识别中的应用

南京大学学报(自然科学版) ›› 2017, Vol. 53 ›› Issue (6): 1091–.

三角稀疏回归分类器及其在稳健人脸识别中的应用

郑　玮1，2，杨万扣3，邵　斐1*，赵　炜1

出版日期:2017-11-27 发布日期:2017-11-27
作者简介:1.金陵科技学院计算机工程学院，南京，211169；
2.南京理工大学计算机科学与技术学院，南京，210094；
3．东南大学自动化学院，南京，210096
基金资助:
基金项目：江苏省高等学校自然科学研究重大资助经费项目(17KJA520001)，江苏省研究生科研创新计划项目(KYCX17_0361)
收稿日期：2017－09－26
*通讯联系人，E－mail：shaofei@jit.edu.cn

Triangle sparse regression and its application in robust face recognition

Zheng Wei1，2，Yang Wankou3，Shao Fei1*，Zhao Wei1

Online:2017-11-27 Published:2017-11-27
About author:1. School of Computer Engineering，Jinling Institute of Technology，Nanjing，211169，China；
2．School of Computer Science and Engineering，Nanjing University of Science and Technology，Nanjing，210094，China；
3．School of Automation，Southeast University，Nanjing，210096，China

摘要/Abstract

摘要： 以稀疏表示为代表的回归分类方法对于高斯噪声具有较好的鲁棒性，但容易受到训练样本中离群点数据的影响导致欠拟合或过拟合．通过探索余弦函数对离群数据呈现的周期不敏感特性，使用余弦函数来刻画回归残差，并在复数域空间进行稀疏回归，提出了三角稀疏回归分类器(TSRC)模型．考虑到模型的非凸特性，普通的迭代算法难以获得全局最优解．因此，通过三角函数演算与核函数技巧将TSRC转化为一个凸优化问题，使用交替方向乘子法(ADMM)对模型进行求解，核函数的计算过程从核空间角度解释了模型对离群点鲁棒的本质原因，通过欧拉公式能够完全避开复数域的计算过程，从而起到加速的作用．在AR，Extend－YaleB及NUST－RF带有遮挡和光照变化的人脸识别数据集上进行了识别率与速度的实验，验证了所提出模型的有效性，在Extend－YaleB数据集上测试了所提出方法在不同尺度的训练样本下的运行效率，并与现阶段先进方法进行了对比．

Abstract: Regression－based classifiers(e.g.Sparse Representation Classifier)have shown the promising performances in several classification tasks with Gaussian noise.However，they are sensitive to the impacts from outlier points in the training data，and lead to underfitting or overfitting results.Exploring the insensitivity of trigonometry functions in one period，the cosine function is employed to measure the fitting residuals of regressions.Considering the non－convexity of the cosine function，directly optimizations are not able to achieve the global minimizer.Therefore，the proposed Triangle Sparse Regression Classifier(TSRC)is transformed to a convex problem by trigonometry and kernel tricks.Subsequently，Alternating Direction Method of Multipliers(ADMM)method is used to solve the proposed model.The provided calculations reveal the robustness of TSRC，and eliminate the complex numbers in optimizations，by virtue of the Euler－Equation.Hence the optimization is only operated in the real numbers field，and thus obtain the efficiency algorithm.Finally，experiments on AR，Extend－YaleB and NUST－RF datasets(with various occlusions and illumination changes)validate the accuracy of the proposed methods，and the running time is illustrated on the variations of samples size of Extend－YaleB dataset，compared with the state－of－the－art regression methods.

郑　玮1，2，杨万扣3，邵　斐1*，赵　炜1. 三角稀疏回归分类器及其在稳健人脸识别中的应用[J]. 南京大学学报(自然科学版), 2017, 53(6): 1091–.

Zheng Wei1，2，Yang Wankou3，Shao Fei1*，Zhao Wei1. Triangle sparse regression and its application in robust face recognition[J]. Journal of Nanjing University(Natural Sciences), 2017, 53(6): 1091–.

参考文献

[1]　周志华．机器学习．北京：清华大学出版社，2016，60－63.(Zhou Z H．Machine learning.Beijing：Tsinghua University Press，2016，60－63.)
[2] Naseem I，Togneri R，Bennamoun M．Linear regression for face recognition.IEEE Transactions on Pattern Analysis and Machine Intelligence，2010，32(11)：2106－2112.
[3] Wright J，Yang A Y，Ganesh A，et al.Robust face recognition via sparse representation.IEEE Transactions on Pattern Analysis and Machine Intelligence，2009，31(2)：210－227.
[4] Yang J，Luo L，Qian J J，et al.Nuclear norm based matrix regression with applications to face recognition with occlusion and illumination changes.IEEE Transactions on Pattern Analysis and Machine Intelligence，2017，39(1)：156－171.
[5] Mairal J，Bach F，Ponce J，et al.Supervised dictionary learning.arXiv：0809.3083，2009.
[6] Zhang Q，Li B X．Discriminative K－SVD for dictionary learning in face recognition.In：Proceedings of 2010 IEEE Computer Society Conference on Computer Vision and Pattern Recognition.San Francisco，CA，USA：IEEE，2010：2691－2698.
[7] Yang J C，Yu K，Gong Y H，et al.Linear spatial pyramid matching using sparse coding for image classification.In：Proceedings of 2009 IEEE Conference on Computer Vision and Pattern Recognition.Miami，FL，USA：IEEE，2009：1794－1801.　
[8] Elad M，Aharon M．Image denoising via sparse and redundant representations over learned dictionaries.IEEE Transactions on Image Processing，2006，15(12)：3736－3745.
[9] Guyon I，Elisseeff A．An introduction to variable and feature selection.Journal of Machine Learning Research，2003，3：1157－1182.
[10] Donoho D L．De－noising by soft－thresholding.IEEE Transactions on Information Theory，1995，41(3)：613－627.
[11] 陈宝林．最优化理论与算法．第2版．北京：清华大学出版社，2005，232－238.(Chen B L．Optimization theory and algorithm.The 2nd Editon.Beijing：Tsinghua University Press，2005，232－238.)
[12] Eckstein J．Augmented Lagrangian and alternating direction methods for convex optimization：A tutorial and some illustrative computational results.Rutgers University：RUTCOR Research Reports，2012，32.
[13] Lee H，Battle A，Raina R，et al.Efficient sparse coding algorithms.In：Advances in Neural Information Processing Systems 19，Proceedings of the 20th Annual Conference on Neural Information Processing Systems.Vancouver，Canada：NIPS，2006，801－808.
[14] Naseem I，Togneri R，Bennamoun M．Robust regression for face recognition.Pattern Recognition，2012，45(1)：104－118.
[15] Li X X，Dai D Q，Zhang X F，et al.Structured sparse error coding for face recognition with occlusion.IEEE Transactions on Image Processing，2013，22(5)：1889－1900.
[16] He R，Zheng W S，Hu B G，et al.A regularized correntropy framework for robust pattern recognition.Neural Computation，2011，23(8)：2074－2100.
[17] He R，Zheng W S，Hu B G．Maximum correntropy criterion for robust face recognition.IEEE Transactions on Pattern Analysis and Machine Intelligence，2011，33(8)：1561－1576.
[18] Liwicki S，Tzimiropoulos G，Zafeiriou S，et al.Euler principal component analysis.International Journal of Computer Vision，2013，101(3)：498－518.　
[19] Gao S，Tsang I，Chia L T．Kernal sparse representation for image classification and face recognition.In：The 11th European Conference on Computer Version(ECCV 2010)．Crete，Greece：Springer，2010：1－14.
[20] Boyd S，Parikh N，Chu E，et al.Distributed optimization and statistical learning via the alternating direction method of multipliers.Foundations and Trends? in Machine Learning，2011，3(1)：1－122.
[21] Lin T Y，Ma S Q，Zhang S Z．On the global linear convergence of the admm with multiblock variables.SIAM Journal on Optimization，2015，25(3)：1478－1497.
[22] Jia K，Chan T H，Ma Y．Robust and practical face recognition via structured sparsity.In：Fitzgibbon A，Lazebnik S，Perona P，et al.12th European Conference on Computer Version.Springer Berlin Heidelberg，2012：331－344.
[23] Martinez A M，Benavente R．The AR face database.CVC Technical Report #24，1998.
[24] Lee K C，Ho J，Kriegman D J．Acquiring linear subspaces for face recognition under variable lighting.IEEE Transactions on Pattern Analysis and Machine Intelligence，2005，27(5)：684－698.

Metrics

Viewed

Full text

Abstract

Cited

Shared

Discussed