基于奇异值分解的Gabor遮挡字典学习

doi:10.11896/j.issn.1002-137X.2018.06.049

计算机科学 ›› 2018, Vol. 45 ›› Issue (6): 275-283.doi: 10.11896/j.issn.1002-137X.2018.06.049

基于奇异值分解的Gabor遮挡字典学习

李小薪¹, 周元申¹, 周旋², 李晶晶¹, 刘志勇²

浙江工业大学计算机科学与技术学院杭州310032¹;
深圳职业技术学院工业中心深圳518055²

收稿日期:2017-03-24 出版日期:2018-06-15 发布日期:2018-07-24
作者简介:李小薪(1980－),男,博士,副教授,主要研究方向为图像处理与模式识别,E-mail:mordecai@163.com;周元申(1992－),男,硕士,主要研究方向为图像处理与模式识别,E-mail:ysczys@gmail.com;周旋(1982－),男,讲师,主要研究方向为复杂网络与非线性系统,E-mail:zhoux0428@szpt.edu.cn;李晶晶(1986－),女,硕士,主要研究方向为图像处理与模式识别,E-mail:jing186@126.com;刘志勇(1975－),男,副教授,CCF会员,主要研究方向为计算机视觉与模式识别,E-mail:liuzhiyong@szpt.edu.cn(通信作者)
基金资助:
本文受国家自然科学基金(61402411),浙江省自然科学基金(LY18F020031,LQ14C010001,LY18F020028),深圳市科技项目(JCYJ20150630114140642)资助

Gabor Occlusion Dictionary Learning via Singular Value Decomposition

LI Xiao-xin¹, ZHOU Yuan-shen¹, ZHOU Xuan², LI Jing-jing¹, LIU Zhi-yong²

College of Computer Science and Technology,Zhejiang University of Technology,Hangzhou 310032,China¹;
Industry Center,Shenzhen Polytechnic,Shenzhen 518055,China²

Received:2017-03-24 Online:2018-06-15 Published:2018-07-24

摘要/Abstract

摘要： 因遮挡、光照等变化因素所引发的协变量偏移问题是面向现实的人脸识别系统需要重点解决的问题。从字典编码的角度探讨了这一问题。通过对现有的结构化误差编码方法的回顾,指出几种主流的结构化误差编码方法都可以转化为训练字典与遮挡字典联合表示的形式,只需对不同的误差编码方法建立合适的遮挡字典即可。鉴于遮挡字典在结构化误差编码方法中的重要作用,针对一种重要的基于字典表示的误差校正方法——基于Gabor特征的鲁棒表示与分类方法(GRRC)展开研究,指出其基于K-SVD的遮挡字典学习方法的主要不足在于:计算代价较高、冗余性较强、缺乏针对自然遮挡的结构,并提出了一种基于奇异值分解(SVD)的Gabor遮挡字典学习方法。在Extended Yale B,UMBDB和AR 3个人脸数据库上的实验结果表明,相对于基于K-SVD字典学习方法的GRRC,基于SVD字典学习方法的GRRC在各种情形下都具有更好的时间性能和识别性能。

关键词: Gabor特征, K-SVD, 奇异值分解, 遮挡字典, 主成分分析

Abstract: Covariate shift incurred by occlusion and illumination variations is an important problem for real-world face recognition systems.This paper explored this problem from the perspective of dictionary coding.By reviewing several extant structured error coding methods,this paper indicated that these error coding methods can be rewritten as a linear system by combining training dictionary and well-designed occlusion dictionary.Due to the importance of occlusion dictionary in structured error coding,this paper studied the dictionary learning method,K-SVD (Singular Value Decomposition),which is used in the Gabor feature based robust representation and classification (GRRC) method,and has been paid great attentions in the field of error coding.The K-SVD learned occlusion dictionary is strongly redundant and lack of natural structures.In addition,K-SVD is time-consuming.This paper proposed an SVD-based occlusion dictionary learning method.It is simple,but generates a more compacted and structured occlusion dictionary.Experiments on three face datasets,including Extended Yale B,UMBDB and AR,demonstrates that the proposed SVD-based GRRC consis-tently outperforms the K-SVD-based GRRC in several challenging situations.

Key words: Gabor feature, K-SVD, Occlusion dictionary, PCA, Singular value decomposition

中图分类号:

TP391.4

李小薪, 周元申, 周旋, 李晶晶, 刘志勇. 基于奇异值分解的Gabor遮挡字典学习[J]. 计算机科学, 2018, 45(6): 275-283. https://doi.org/10.11896/j.issn.1002-137X.2018.06.049

LI Xiao-xin, ZHOU Yuan-shen, ZHOU Xuan, LI Jing-jing, LIU Zhi-yong. Gabor Occlusion Dictionary Learning via Singular Value Decomposition[J]. Computer Science, 2018, 45(6): 275-283. https://doi.org/10.11896/j.issn.1002-137X.2018.06.049

参考文献

[1]PARKHI O M,VEDALDI A,ZISSERMAN A.Deep Face Recognition [C]//Proceedings of British Machine Vision Confe-rence.London:BMVA Press,2015:41.
[2]SUN Y,WANG X,TANG X.Sparsifying Neural Network Connections for Face Recognition [C]//IEEE Conference on Computer Vision and Pattern Recognition.2016:4856-4864.
[3]IOFFE S,SZEGEDY C.Batch Normalization:Accelerating Deep Network Training by Reducing Internal Covariate Shift [C]//International Conference on Machine Learning.2015:448-456.
[4]ZHOU Z,WAGNER A,MOBAHI H,et al.Face Recognition with Contiguous Occlusion Using Markov Random Fields [C]//IEEE International Conference on Computer Vision.2009:1050-1057.
[5]HE R,ZHENG W,HU B.Maximum Correntropy Criterion for Robust Face Recognition [J].IEEE Transactions on PatternAnalysis and Machine Intelligence,2011,33(8):1561-1576.
[6]YANG M,ZHANG L,YANG J,et al.Regularized Robust Co-ding for Face Recognition [J].IEEE Transactions on Image Processing,2013,22(5):1753-1766.
[7]LI X,DAI D,ZHANG X,et al.Structured Sparse Error Coding for Face Recognition with Occlusion [J].IEEE Transactions on Image Processing,2013,22(5):1889-1900.
[8]YANG M,ZHANG L,SHIU S C,et al.Gabor Feature Based Robust Representation and Classification for Face Recognition with Gabor Occlusion Dictionary [J].Pattern Recognition,2013,46(7):1865-1878.
[9]YANG M,ZHANG L.Gabor Feature Based Sparse Representation for Face Recognition with Gabor Occlusion Dictionary [C]//European Conference on Computer Vision.Springer Berlin Heidelberg,2010:448-461.
[10]DENG W,HU J,GUO J.Extended SRC:Undersampled Face Recognition via Intra-Class Variant Dictionary [J].IEEE Transactions on Pattern Analysis and Machine Intelligence,2012,34(9):1864-1870.
[11]WEI X,LI C,HU Y.Robust Face Recognition under Varying Illumination and Occlusion Considering Structured Sparsity [C]//International Conference on Digital Image Computing Techniques and Applications.2012:1-7.
[12]WRIGHT J,YANG A Y,GANESH A,et al.Robust Face Recognition via Sparse Representation [J].IEEE Transactions on Pattern Analysis and Machine Intelligence,2009,31(2):210-227.
[13]JIA K,CHAN T H,MA Y.Robust and Practical Face Recognition via Structured Sparsity [C]//European Conference on Computer Vision.2012:331-344.
[14]YANG J,LUO L,QIAN J,et al.Nuclear Norm Based Matrix Regression with Applications to Face Recognition with Occlusion and Illumination Changes [J].IEEE Transactions on Pattern Analysis and Machine Intelligence,2017,39(1):156-171.
[15]HE R,ZHENG W S,TAN T,et al.Half-Quadratic-Based Iterative Minimization for Robust Sparse Representation [J].IEEE Transactions on Pattern Analysis and Machine Intelligence,2014,36(2):261-275.
[16]OU W,YOU X,TAO D,et al.Robust Face Recognition via Occlusion Dictionary Learning [J].Pattern Recognition,2014,47(4):1559-1572.
[17]AZEEM A,SHARIF M,RAZA M,et al.A Survey:Face Recognition Techniques under Partial Occlusion[J].International ArabJournal of Information Technology,2014,11(1):1-10.
[18]HASSABALLAH M,ALY S.Face Recognition:Challenges,Achievements and Future Directions [J].IET Computer Vision,2015,9(4):614-626.
[19]YANG M,FENG Z,SHIU S C K,et al.Fast and Robust Face Recognition via Coding Residual Map Learning Based Adaptive Masking [J].Pattern Recognition,2014,47(2):535-543.
[20]QIAN J,LUO L,YANG J,et al.Robust Nuclear Norm Regularized Regression for Face Recognition with Occlusion [J].Pattern Recognition,2015,48(10):3145-3159.
[21]WEN Y,LIU W,YANG M,et al.Structured Occlusion Coding for Robust Face Recognition [J].Neurocomputing,2016,178:11-24.
[22]AHARON M,ELAD M,BRUCKSTEIN A.K-SVD:An Algorithm for Designing Overcomplete Dictionaries for Sparse Representation [J].IEEE Transactions on Signal Processing,2006,54(11):4311-4322.
[23]HASTIE T,TIBSHIRANI R,FRIEDMAN J,et al.The Elements of Statistical Learning:Data Mining,Inference and Prediction [M].NewYork,USA:Springer,2005:83-85.
[24]TURK M,PENTLAND A.Eigenfaces for Recognition [J].Journal of Cognitive Neuroscience,1991,3(1):71-86.
[25]KIM W,SUH S,HWANG W,et al.SVD Face:Illumination-Invariant Face Representation [J].IEEE Signal Processing Letters,2014,21(11):1336-1340.
[26]CHAN T,JIA K,GAO S,et al.PCANet:A Simple Deep Learning Baseline for Image Classification? [J].IEEE Transactions on Image Processing,2015,24(12):5017-5032.
[27]ELHAMIFAR E,VIDAL R.Robust Classification Using Structured Sparse Representation [C]//IEEE Conference on Computer Vision and Pattern Recognition.2011:1873-1879.
[28]LEE K C,HO J,KRIEGMAN D J.Acquiring Linear Subspaces for Face Recognition under Variable Lighting [J].IEEE Tran-sactions on Pattern Analysis and Machine Intelligence,2005,27(5):684-698.
[29]GEORGHIADES A S,BELHUMEUR P N,KRIEGMAN D J.From Few to Many:Illumination Cone Models for Face Recognition under Variable Lighting and Pose [J].IEEE Transactions on Pattern Analysis and Machine Intelligence,2001,23(6):643-660.
[30]COLOMBO A,CUSANO C,SCHETTINI R.UMB-DB:A Database of Partially Occluded 3D Faces [C]//IEEE International Conference on Computer Vision Workshops.2011:2113-2119.
[31]MARTINEZ A M.The AR Face Database[D].Columbus,USA:Ohio State University,1998.
[32]EKENEL H,STIEFELHAGEN R.Why is Facial Occlusion a Challenging Problem? [C]//Proceedings of Advances in Biometrics.Springer Berlin,2009:299-308.

相关文章 15

[1]	李其烨, 邢红杰. 基于最大相关熵的KPCA异常检测方法 KPCA Based Novelty Detection Method Using Maximum Correntropy Criterion 计算机科学, 2022, 49(8): 267-272. https://doi.org/10.11896/jsjkx.210700175
[2]	阙华坤, 冯小峰, 刘盼龙, 郭文翀, 李健, 曾伟良, 范竞敏. Grassberger熵随机森林在窃电行为检测的应用 Application of Grassberger Entropy Random Forest to Power-stealing Behavior Detection 计算机科学, 2022, 49(6A): 790-794. https://doi.org/10.11896/jsjkx.210800032
[3]	任花, 牛少彰, 王茂森, 岳桢, 任如勇. 基于奇异值分解的同态可交换脆弱零水印研究 Homomorphic and Commutative Fragile Zero-watermarking Based on SVD 计算机科学, 2022, 49(3): 70-76. https://doi.org/10.11896/jsjkx.210800015
[4]	吴善杰, 王新. 基于AGA-DBSCAN优化的RBF神经网络构造煤厚度预测方法 Prediction of Tectonic Coal Thickness Based on AGA-DBSCAN Optimized RBF Neural Networks 计算机科学, 2021, 48(7): 308-315. https://doi.org/10.11896/jsjkx.200800110
[5]	胡昕彤, 沙朝锋, 刘艳君. 基于随机投影和主成分分析的网络嵌入后处理算法 Post-processing Network Embedding Algorithm with Random Projection and Principal Component Analysis 计算机科学, 2021, 48(5): 124-129. https://doi.org/10.11896/jsjkx.200500058
[6]	王艺皓, 丁洪伟, 李波, 保利勇, 张颖婕. 基于聚类与特征融合的蛋白质亚细胞定位预测 Prediction of Protein Subcellular Localization Based on Clustering and Feature Fusion 计算机科学, 2021, 48(3): 206-213. https://doi.org/10.11896/jsjkx.200200081
[7]	冯安然, 王旭仁, 汪秋云, 熊梦博. 基于PCA和随机树的数据库异常访问检测 Database Anomaly Access Detection Based on Principal Component Analysis and Random Tree 计算机科学, 2020, 47(9): 94-98. https://doi.org/10.11896/jsjkx.190800056
[8]	杨云铄, 桑庆兵. 无需学习的无参考彩色噪声图像质量评价方法 No-reference Color Noise Images Quality Assessment Without Learning 计算机科学, 2020, 47(10): 161-168. https://doi.org/10.11896/jsjkx.190900051
[9]	李桂会,李晋江,范辉. 自适应匹配追踪图像去噪算法 Image Denoising Algorithm Based on Adaptive Matching Pursuit 计算机科学, 2020, 47(1): 176-185. https://doi.org/10.11896/jsjkx.181202280
[10]	刘晴晴, 罗永龙, 汪逸飞, 郑孝遥, 陈文. 基于SVD填充的混合推荐算法 Hybrid Recommendation Algorithm Based on SVD Filling 计算机科学, 2019, 46(6A): 468-472.
[11]	史燕燕, 白静. 融合CFCC和Teager能量算子倒谱参数的语音识别 Speech Recognition Combining CFCC and Teager Energy Operators Cepstral Coefficients 计算机科学, 2019, 46(5): 286-289. https://doi.org/10.11896/j.issn.1002-137X.2019.05.044
[12]	杜秀丽, 左思铭, 邱少明. 基于图像灰度熵的自适应字典学习算法 Adaptive Dictionary Learning Algorithm Based on Image Gray Entropy 计算机科学, 2019, 46(5): 266-271. https://doi.org/10.11896/j.issn.1002-137X.2019.05.041
[13]	毛莺池,王静,陈小丽,徐淑芳,陈豪. 基于特征组合与CNN的大坝缺陷识别与分类方法 Dam Defect Recognition and Classification Based on Feature Combination and CNN 计算机科学, 2019, 46(3): 267-276. https://doi.org/10.11896/j.issn.1002-137X.2019.03.040
[14]	张明月, 王静. 基于深度学习的交互似然目标跟踪算法 Interactive Likelihood Target Tracking Algorithm Based on Deep Learning 计算机科学, 2019, 46(2): 279-285. https://doi.org/10.11896/j.issn.1002-137X.2019.02.043
[15]	高忠石, 苏旸, 柳玉东. 基于PCA-LSTM的入侵检测研究 Study on Intrusion Detection Based on PCA-LSTM 计算机科学, 2019, 46(11A): 473-476.

Metrics

Viewed

Full text

Abstract

Cited

Shared

Discussed

基于奇异值分解的Gabor遮挡字典学习

Gabor Occlusion Dictionary Learning via Singular Value Decomposition

PDF (PC)

摘要/Abstract

引用本文

使用本文

参考文献

相关文章 15

Metrics

本文评价

推荐阅读 0