1.College of Computer Science and Technology,Nanjing University of Aeronautics Astronautics,Nanjing 210016 2.Information Engineering College,Yangzhou University,Yangzhou 225009 3.College of Science,Nanjing University of Aeronautics Astronautics,Nanjing 210016 4.National Key Laboratory for Novel Software Technology,Nanjing University,Nanjing 210093
Abstract Canonical Correlation Analysis (CCA) has following two deficiencies in performing classification task: CCA can not directly optimize them but their components, though combined features are the input of the classifier; CCA can not utilize any class information at all, though facing classification task. To overcome these deficiencies, a supervised dimension reduction method named combined-feature-discriminability enhanced canonical correlation analysis (CECCA) is proposed. CECCA is developed through incorporating discriminant analysis of combined features into CCA. Consequently, it optimizes the combined feature correlation and discriminability simultaneously and thus makes the extracted features more suitable for classification. The experimental results on artificial dataset, multiple feature database and facial databases show that the proposed method is effective.
[1] Hotelling H.Relations between Two Sets of Variates.Biometrika,1936,28(3/4): 321-377 [2] Anderson T W.An Introduction to Multivariate Statistical Analysis.3rd Edition.Hoboken,USA: Wiley,2003 [3] Johnson R A,Wichern D W.Applied Multivariate Statistical Analysis.6th Edition.Cambridge,USA: Prentice Hall,2007 [4] Liu Yanyan,Liu Xiuping,Su Zhixun.A New Fuzzy Approach for Handling Class Labels in Canonical Correlation Analysis.Neurocomputing,2008,71(7/8/9): 1735-1740 [5] Ogura T.A Variable Selection Method in Principal Canonical Correlation Analysis.Computational Statistics and Data Analysis,2010,54(3): 1117-1123 [6] Sun Quansen,Zeng Shenggen,Liu Yan,et al.A New Method of Feature Fusion and Its Application in Image Recognition.Pattern Recognition,2005,38(12): 2437-2448 [7] Hardoon D R,Szedmak S,Taylor J S.Canonical Correlation Analysis: An Overview with Application to Learning Method.Neural Computation,2004,16(12): 2639-2664 [8] Correa N M,Eichele T,Adali T,et al.Multi-Set Canonical Correlation Analysis for the Fusion of Concurrent Single Trial ERP and Functional MRI.Neuroimage,2010,50(4): 1438-1445 [9] Paskaleva B,Hayat M M,Wang Zhipeng,et al.Canonical Correlation Feature Selection for Sensors with Overlapping Bands: Theory and Application.IEEE Trans on Geoscience and Remote Sensing,2008,46(10): 3346-3358 [10] Fu Yun,Huang T S.Image Classification Using Correlation Tensor Analysis.IEEE Trans on Image Processing,2008,17(2): 226-234 [11] Huang Hua,He Huiting,Fan Xin,et al.Super-Resolution of Human Face Image Using Canonical Correlation Analysis.Pattern Recognition,2010,43(7): 2532-2543 [12] Sargin M E,Erzin E,Yemez Y,et al.Multimodal Speaker Identification Using Canonical Correlation Analysis // Proc of the IEEE International Conference on Acoustics,Speech and Signal Processing.Toulouse,France,2006: 613-616 [13] Sargin M E,Yemez Y,Erzin E,et al.Audiovisual Synchronization and Fusion Using Canonical Correlation Analysis.IEEE Trans on Multimedia,2007,9(7): 1396-1403 [14] Kim T K,Cipolla R.Canonical Correlation Analysis of Video Volume Tensors for Action Categorization and Detection.IEEE Trans on Pattern Analysis and Machine Intelligence,2009,31(8): 1415-1428 [15] Li Yiou,Adali T,Wang Wei,et al.Joint Blind Source Separation by Multiset Canonical Correlation Analysis.IEEE Trans on Signal Processing,2009,57(10): 3918-3929 [16] Pan Hao,Liang Zhipei,Huang T S.Exploiting the Dependencies in Information Fusion // Proc of the IEEE Conference on Computer Vision and Pattern Recognition.Fort Collins,USA,1999,Ⅱ: 407-412 [17] Sun Tingkai,Chen Songcan,Yang Jingyu,et al.A Novel Method of Combined Feature Extraction for Recognition // Proc of the IEEE Conferences on Data Mining.Pisa,Italy,2008: 1043-1048 [18] Peng Yan,Zhang Daoqiang,Zhang Jianchun.A New Canonical Correlation Analysis Algorithm with Local Discrimination.Neural Processing Letters,2009,31(1): 1-15 [19] Via J,Santamaria I,Pérez J.A Learning Algorithm for Adaptive Canonical Correlation Analysis of Several Data Sets.Neural Networks,2007,20(1): 139-152 [20] Fisher R A.The Use of Multiple Measurement in Taxonomic Problems.Annals of Eugenics,1937,7(2): 179-188 [21] Wang Huan,Yan Shuicheng,Xu Dong,et al.Trace Ratio vs.Ratio Trace for Dimensionality Reduction // Proc of the IEEE Conference on Computer Vision and Pattern Recognition.Minneapolis,USA,2007: 108-115 [22] Yan Shuicheng,Xu Dong,Zhang Benyu.Graph Embedding and Extensions: A General Framework for Dimensionality Reduction.IEEE Trans on Pattern Analysis and Machine Intelligence,2007,29(1): 40-51 [23] Belhumeur P N,Hespanha J P,Kriegman D J.Eigenfaces vs.Fisherfaces: Recognition Using Class Specific Linear Projection.IEEE Trans on Pattern Analysis and Machine Intelligence,1997,19(7): 711-720 [24] Martinez A M,Kak A C.PCA Versus LDA.IEEE Trans on Pattern Analysis and Machine Intelligence,2001,23(2): 228-233 [25] Cai Deng,He Xiaofei,Han Jiawei.Semi-Supervised Discriminant Analysis // Proc of the IEEE International Conference on Computer Vision.Rio de Janeiro,Brazil,2007: 1-7
2012, Vol. 25 
