[1]那 天,宋晓宁,於东军.基于主元分析和线性判别分析降维的稀疏表示分类[J].南京理工大学学报(自然科学版),2018,42(03):286.[doi:10.14177/j.cnki.32-1397n.2018.42.03.005]
 Na Tian,Song Xiaoning,Yu Dongjun.Sparse representation-based classification based on principal componentanalysis and linear discriminant analysis dimensionality reduction[J].Journal of Nanjing University of Science and Technology,2018,42(03):286.[doi:10.14177/j.cnki.32-1397n.2018.42.03.005]
点击复制

基于主元分析和线性判别分析降维的稀疏表示分类()
分享到:

《南京理工大学学报》(自然科学版)[ISSN:1005-9830/CN:32-1397/N]

卷:
42卷
期数:
2018年03期
页码:
286
栏目:
出版日期:
2018-06-30

文章信息/Info

Title:
Sparse representation-based classification based on principal component analysis and linear discriminant analysis dimensionality reduction
文章编号:
1005-9830(2018)03-0286-06
作者:
那 天1宋晓宁1於东军2
1.江南大学 物联网工程学院,江苏 无锡 214122; 2.南京理工大学 计算机科学与工程学院,江苏 南京 210094
Author(s):
Na Tian1Song Xiaoning1Yu Dongjun2
1.School of Internet of Things Engineering,Jiangnan University,Wuxi 214122,China; 2.School of Computer Science and Engineering,Nanjing University of Science and Technology,Nanjing 210094,China
关键词:
主元分析 线性判别分析 降维 稀疏表示分类 人脸识别 协同表达分类
Keywords:
principal component analysis linear discriminant analysis dimensionality reduction sparse representation-based classification face recognition collaborative representation-based classification
分类号:
TP391
DOI:
10.14177/j.cnki.32-1397n.2018.42.03.005
摘要:
为解决传统的稀疏表示分类(SRC)算法在小样本人脸识别过程中的过大时间开销问题,该文提出2种基于降维的SRC算法。扩展主元分析(EPCA)算法利用PCA算法构造约束优化稀疏模型,对测试样本进行线性表示,通过比较测试样本和每类训练样本的重构PCA系数进行决策分类。EPCA+线性判别分析(EPCA+LDA)算法在EPCA算法的基础上增加LDA约束模型,提高重构样本的稀疏表示的鉴别性。将该文算法应用于AR和FERET人脸数据库,与扩展SRC(ESRC)、SRC、SRC_PCA、协同表达分类(CRC)算法相比,该文算法有较高的识别率和较低的时间复杂度。将EPCA算法和EPCA+LDA算法应用于FETET数据集,识别率分别为61.46%和59.17%,运行时间分别为383.02 s和220.62 s。
Abstract:
Two sparse representation-based classification(SRC)algorithms based on dimensionality reduction are proposed to solve the high time cost problem of traditional sparse representation-based classification methods for small sample data face recognition.In the extended principal component analysis(EPCA)algorithm,an optimization sparse model is achieved using the PCA algorithm,the test samples are represented linearly,and classification is performed by comparing the reconstructed PCA coefficient of the test samples with that of training samples.In the EPCA+linear discriminant analysis(EPCA+LDA)algorithm,a LDA constraint model is added to improve the identification of sparse representation of reconstructed samples.The experimental results of the AR and FERET database show that,compared with extended SRC(ESRC),SRC,SRC_PCA,collaborative representation-based classification(CRC)algorithm,the algorithms proposed here have higher recognition rates and lower time complexities.Especially on the FERET database,the EPCA algorithm and EPCA+LDA algorithm achieve 61.46% and 59.17% recognition rates,and 383.02 s and 220.62 s running times respectively.

参考文献/References:

[1] 徐勇,范自柱,张大鹏.基于稀疏算法的人脸识别[M].北京:国防工业出版,2014:10-22. [2]Jolliffe I T.Principal component analysis[M].2nd ed.New York,USA:Springer-Verlag New York,2005:41-64. [3]边肇祺,张学工.模式识别[M].北京:清华大学出版社,2007:87-91,222-226. [4]Belkin M,Niyogi P.Laplacian eigenmaps for dimensionality reduction and data representation[J].Neural Computation,2003,15(6):1373-1396. [5]He Xiaofei,Niyogi P.Locality preserving projections[J].Advances in Neural Information Processing Systems,2005,45(1):186-197. [6]任成娟,胡淑芳,刘崇文.子模式局部保持映射人脸识别算法[J].重庆理工大学学报(自然科学版),2011,25(6):84-89. Ren Chengjuan,Hu Shufang,Liu Chongwen.Subpattern locality preserving projections for face recognition[J].Joumal of Chongqing University of Techology(Natural Science),2011,25(6):84-89. [7]马瑞,王家,宋亦旭.基于局部线性嵌入(LLE)非线性降维的多流形学习[J].清华大学学报(自然科学版),2008,48(4):582-585. Ma Rui,Wang Jia,Song Yixu.Multi-manifold learning using locally linear embedding(LLE)nonlinear dimensionality reduction[J].Journal of Tsinghua University(Science and Technology),2008,48(4):582-585. [8]杨荣根,任明武,杨静宇.基于稀疏表示的人脸识别算法[J].计算机科学,2010,37(9):267-269. Yang Ronggen,Ren Mingwu,Yang Jingyu.Sparse representation based face recognition algorithm[J].Computer Science,2010,37(9):267-269. [9]Deng Weihong,Hu Jiani,Guo Jun.Extended SRC:Undersampled face recognition via intraclass variant dictionary[J].IEEE Transactions on Pattern Analysis & Machine Intelligence,2012,34(9):1864-1870. [10]Yang Meng,Zhang Lei,Zhang D,et al.Relaxed collaborative representation for pattern classification[EB/OL]. http://www4.comp.polyu.edu.hk/~cslzhang/paper/conf/RCR_CVPR12.pdf,2018-04-10. [11]Zhu Pengfei,Zuo Wangmeng,Zhang Lei,et al.Image set-based collaborative representation for face recognition[J].IEEE Transactions on Information Forensics & Security,2014,9(7):1120-1132. [12]Song Xiaoning,Feng Zhenhua,Hu Guosheng,et al.Half-face dictionary integration for representation-based classification[J].IEEE Transactions on Cybernetics,2017,41(1):142-152. [13]Martinez A M.The AR face database[R].Barcelona,Spain:Computer Vision Center,1998. [14]Phillips P J,Moon H,Rizvi S A,et al.The FERET evaluation methodology for face-recognition algorithms[J].IEEE Transactions on Pattern Analysis and Machine Intelligence,2000,22(10):1090-1104.

备注/Memo

备注/Memo:
收稿日期:2017-01-17 修回日期:2017-02-23 基金项目:国家重点研发计划(2017YFC1601800); 中国博士后科学基金(2018T110441); 江苏省自然科学基金(BK20161135); 江苏省“六大人才高峰”高层次人才项目(XYDXX-012) 作者简介:那天(1993-),女,硕士生,主要研究方向:人工智能与模式识别,E-mail:na_tiantian@163.com; 通讯作者:宋晓宁(1975-),男,博士,副教授,主要研究方向:人工智能与模式识别,E-mail:x.song@jiangnan.edu.cn。 引文格式:那天,宋晓宁,於东军.基于主元分析和线性判别分析降维的稀疏表示分类[J].南京理工大学学报,2018,42(3):286-291. 投稿网址:http://zrxuebao.njust.edu.cn
更新日期/Last Update: 2018-06-30