边界监督多重集典型相关分析
|
摘要
多重集典型相关分析(multiset canonical correlation analysis,MCCA)仅仅考虑了多组数据间的相关性信息,不能有效地反映样本数据的几何结构与鉴别信息,因此为了解决这个问题,首先在LDA思想的启发下,构建了监督多重集典型相关分析(supervised multiset canonical correlation analysis,SMCC)的理论框架,并以此为基础,结合边界Fisher分析(marginal fisher analysis,MFA),提出了边界监督多重集典型相关分析(marginal SMCC,MSMCC).该算法的基本思想是在最大化数据相关性的同时,还要最大化组内数据的类间离散度以及最小化组内数据的类内离散度.在人脸图像与目标数据库上的实验结果验证了所提算法的有效性.
Due to multiset canonical correlation analysis(MCCA)fails to discover the intrinsic geometrical and discriminating structure of multiple data spaces in real world applications.In oder to solve this problem,we first construct supervised multiset canonical correlation analysis(SMCC)by using the ideas from LDA.Based on marginal fisher analysis(MFA),we then propose marginal supervised multiset canonical correlation analysis(MSMCC).It can not only express the correlation among multiple feature vectors,but also effectively depict the data between the geometry and discriminative structure.Extensive experiments on both face image databases and object databases demonstrate the effectiveness of the proposed method.
Due to multiset canonical correlation analysis(MCCA)fails to discover the intrinsic geometrical and discriminating structure of multiple data spaces in real world applications.In oder to solve this problem,we first construct supervised multiset canonical correlation analysis(SMCC)by using the ideas from LDA.Based on marginal fisher analysis(MFA),we then propose marginal supervised multiset canonical correlation analysis(MSMCC).It can not only express the correlation among multiple feature vectors,but also effectively depict the data between the geometry and discriminative structure.Extensive experiments on both face image databases and object databases demonstrate the effectiveness of the proposed method.
引文
[1]Hotelling H.Relations between two sets of variates[J].Biometrika,1936,28(3/4):321-377.
[2]Sun Q S,Zeng S G,Liu Y,et al.A new method of feature fusion and its application in image recognition[J].Pattern Recognition,2005,38(12):2437-2448.
[3]孙权森,曾生根,王平安.典型相关分析的理论及其在特征融合中的应用[J].计算机学报,2005,28(9):1524-1533.
[4]Carlos A,Johan A K S.A regularized kernel CCA contrast function for ICA[J].Neural Networks,2008,21(2/3):170-181.
[5]许颖春,范丽亚.基于L1范数的非线性TSVR[J].聊城大学学报(自然科学版),2017,30(3):6-11.
[6]夏文静,陈耿,范丽亚.八种最小二乘SVM型学习算法的优势比较[J].聊城大学学报(自然科学版),2016,29(2):33-41.
[7]Sun Q S,Liu Z D,Heng P A,et al.A theorem on the generalized canonical projective vectors[J].Pattern Recognition,2005,38(3):449-452.
[8]Sun T K,Chen S C,Yang J Y,et al.A supervised combined feature extraction method for recognition[C].//Proceedings of IEEE International Conference on Data Mining,Pisa,Italy,2008.
[9]Yang J,Sun Q S.A novel generalized fuzzy canonical correlation analysis framework for feature fusion and recognition[J].Neural Processing Letters,2017,46:521-536.
[10]Yang J,Sun Q S,Yuan Y H.Feature extraction using fractional-order embedding direct linear discriminant analysis[J].Neural Processing Letters,2018,48(3):1583-1595.
[11]Kim T K,Kittler J,Cipolla R.Discriminative learning and recognition of image set classes using canonical correlations[J].IEEE Transactions on Pattern Analysis and Machine Intelligence,2007,29(6):1005-1018.
[12]Li Y O,AdalT,Wang W,et al.Joint blind source separation by multiset canonical correlation analysis[J].IEEE Transactions on Signal Processing,2009,57(10):3918-3929.
[13]Thompson B,Cartmill J,Azimi-Sadjadi M R,et al.A multichannel canonical correlation analysis feature extraction with application to buried underwater target classification[C].//International Joint Conference on Neural Networks,2006.
[14]Ji H K,Shen X B,Sun Q S,et al.Sparse Discrimination based Multiset Canonical Correlation Analysis for Multi-Feature Fusion and Recognition[C].//British Machine Vision Conference,2015.
[15]Ji H K,Sun Q S,Yuan Y H,et al.C2DMCP:View-consistent collaborative discriminative multiset correlation projection for data representation[J].Journal of Visual Communication and Image Representation,2016,40:393-405.
[16]Nielsen A A.Multiset canonical correlations analysis and multispectral,truly multitemporal remote sensing data[J].IEEE Transactions on Image Processing,2002,11(3):293-305.
[17]Correa N M,Eichele T,Adal T,et al.Multi-set canonical correlation analysis for the fusion of concurrent single trial ERP and functional MR[J].NeuroImage,2010,50(4):1438-1445.
[18]LiEmail Y O,Eichele T,Calhoun V D,et al.Group Study of Simulated Driving fMRI Data by Multiset Canonical Correlation Analysis[J].Journal of Signal Processing Systems,2012,68(1):31-48.
[19]Deleus F,Hulle M M V.Functional connectivity analysis of fMRI data based on regularized multiset canonical correlation analysis[J].Journal of Neuroscience Methods,2011,197(1):143-157.
[20]Gao L,Qi L,Chen E Q,et al.Discriminative Multiple Canonical Correlation Analysis for Multi-Feature Information Fusion[C].//2012IEEE Internation Symposium on Multimedia,2012.
[21]Lee G,Singanamalli A.Supervised Multi-View Canonical Correlation Analysis(sMVCCA):Integrating histologic and proteomic features for predicting recurrent prostate cancer[J].IEEE Transactions on Medical Imaging,2015,34(1):284-297.
[22]Yan S,Xu D,Zhang B,et al.Graph embedding and extensions:ageneral framework for dimensionality reduction[J].IEEE Transactions on Pattern Analysis and Machine Intelligence,2007,29(1):40-51.
[23]Chung F R K.Spectral graph theory[D].American:Mathematical Society,1997.
[24]Hou S,Sun Q,Xia D.Feature fusion using multiple component analysis[J].Neural Processing Letters,2011,34(3):259-275.
[25]Yuan Y H,Sun Q S,Zhou Q,et al.A novel multiset integrated canonical correlation analysis framework and its application in feature fusion[J].Pattern Recognition,2011,44(5):1031-1040.
[26]Nene S A,Nayar S K,Murase H.Columbia object image library(COIL-20)[R].Technical Report CUCS-005-96,1996.
[27]Nene S A,Nayar S K,Murase H.Columbia object image library(COIL-100)[R].Technical Report CUCS-006-96,1996.
[2]Sun Q S,Zeng S G,Liu Y,et al.A new method of feature fusion and its application in image recognition[J].Pattern Recognition,2005,38(12):2437-2448.
[3]孙权森,曾生根,王平安.典型相关分析的理论及其在特征融合中的应用[J].计算机学报,2005,28(9):1524-1533.
[4]Carlos A,Johan A K S.A regularized kernel CCA contrast function for ICA[J].Neural Networks,2008,21(2/3):170-181.
[5]许颖春,范丽亚.基于L1范数的非线性TSVR[J].聊城大学学报(自然科学版),2017,30(3):6-11.
[6]夏文静,陈耿,范丽亚.八种最小二乘SVM型学习算法的优势比较[J].聊城大学学报(自然科学版),2016,29(2):33-41.
[7]Sun Q S,Liu Z D,Heng P A,et al.A theorem on the generalized canonical projective vectors[J].Pattern Recognition,2005,38(3):449-452.
[8]Sun T K,Chen S C,Yang J Y,et al.A supervised combined feature extraction method for recognition[C].//Proceedings of IEEE International Conference on Data Mining,Pisa,Italy,2008.
[9]Yang J,Sun Q S.A novel generalized fuzzy canonical correlation analysis framework for feature fusion and recognition[J].Neural Processing Letters,2017,46:521-536.
[10]Yang J,Sun Q S,Yuan Y H.Feature extraction using fractional-order embedding direct linear discriminant analysis[J].Neural Processing Letters,2018,48(3):1583-1595.
[11]Kim T K,Kittler J,Cipolla R.Discriminative learning and recognition of image set classes using canonical correlations[J].IEEE Transactions on Pattern Analysis and Machine Intelligence,2007,29(6):1005-1018.
[12]Li Y O,AdalT,Wang W,et al.Joint blind source separation by multiset canonical correlation analysis[J].IEEE Transactions on Signal Processing,2009,57(10):3918-3929.
[13]Thompson B,Cartmill J,Azimi-Sadjadi M R,et al.A multichannel canonical correlation analysis feature extraction with application to buried underwater target classification[C].//International Joint Conference on Neural Networks,2006.
[14]Ji H K,Shen X B,Sun Q S,et al.Sparse Discrimination based Multiset Canonical Correlation Analysis for Multi-Feature Fusion and Recognition[C].//British Machine Vision Conference,2015.
[15]Ji H K,Sun Q S,Yuan Y H,et al.C2DMCP:View-consistent collaborative discriminative multiset correlation projection for data representation[J].Journal of Visual Communication and Image Representation,2016,40:393-405.
[16]Nielsen A A.Multiset canonical correlations analysis and multispectral,truly multitemporal remote sensing data[J].IEEE Transactions on Image Processing,2002,11(3):293-305.
[17]Correa N M,Eichele T,Adal T,et al.Multi-set canonical correlation analysis for the fusion of concurrent single trial ERP and functional MR[J].NeuroImage,2010,50(4):1438-1445.
[18]LiEmail Y O,Eichele T,Calhoun V D,et al.Group Study of Simulated Driving fMRI Data by Multiset Canonical Correlation Analysis[J].Journal of Signal Processing Systems,2012,68(1):31-48.
[19]Deleus F,Hulle M M V.Functional connectivity analysis of fMRI data based on regularized multiset canonical correlation analysis[J].Journal of Neuroscience Methods,2011,197(1):143-157.
[20]Gao L,Qi L,Chen E Q,et al.Discriminative Multiple Canonical Correlation Analysis for Multi-Feature Information Fusion[C].//2012IEEE Internation Symposium on Multimedia,2012.
[21]Lee G,Singanamalli A.Supervised Multi-View Canonical Correlation Analysis(sMVCCA):Integrating histologic and proteomic features for predicting recurrent prostate cancer[J].IEEE Transactions on Medical Imaging,2015,34(1):284-297.
[22]Yan S,Xu D,Zhang B,et al.Graph embedding and extensions:ageneral framework for dimensionality reduction[J].IEEE Transactions on Pattern Analysis and Machine Intelligence,2007,29(1):40-51.
[23]Chung F R K.Spectral graph theory[D].American:Mathematical Society,1997.
[24]Hou S,Sun Q,Xia D.Feature fusion using multiple component analysis[J].Neural Processing Letters,2011,34(3):259-275.
[25]Yuan Y H,Sun Q S,Zhou Q,et al.A novel multiset integrated canonical correlation analysis framework and its application in feature fusion[J].Pattern Recognition,2011,44(5):1031-1040.
[26]Nene S A,Nayar S K,Murase H.Columbia object image library(COIL-20)[R].Technical Report CUCS-005-96,1996.
[27]Nene S A,Nayar S K,Murase H.Columbia object image library(COIL-100)[R].Technical Report CUCS-006-96,1996.