Journal of Guangdong University of Technology ›› 2009, Vol. 26 ›› Issue (4): 65-69.

• Comprehensive Studies • Previous Articles     Next Articles

The Application of a Weighted Kernel Fisher Discriminant Analysis Applied in Face Recognition

  

  1. Faculty of Applied Mathematics,Guangdong University of Technology,Guangzhou 510006,China
  • Online:2009-12-01 Published:2009-12-01

HTML

PDF (PC)

2010

Abstract: A weighted kernel maximum scatter difference discriminating criterion is developed for extraction of nonlinear features.The proposed method can be used to extract nonlinear features for faces effectively anf to reconstruct between-class scatter matrix and within-class scatter matrix by weighted schemes.Therefore,it can modify the kernel maximum scatter difference discriminating criterions.Experiments performed on ORL and Yale face database verify the effectiveness of the proposed method.

Key words: face recognition； Kernel Fisher Linear Discriminant Analysis(KFDA)； maximum satter difference criterion；

[1] 成新民,蒋云良,胡文军,吴小红.  基于核的Fisher非线性最佳鉴别分析在人脸识别中的应用[J]. 中国图象图形学报. 2007(08)

[2] 孔锐,张冰.  基于核Fisher判决分析的高性能多类分类算法[J]. 计算机应用. 2005(06)

[3] 程云鹏主编.矩阵论[M]. 西北工业大学出版社, 1989

[4] Belhumeur P N,,Hespanha J P,Kriegnan D J.Eigenfaces vs Fisherfaces:recognition using special linear projection. IEEE Transactions on Pattern Analysis and Machine Intelligence . 1997

[5] Mika S,Ratch G,Weston J,et al.Fisher discriminantAnalysis with kernels. proceedings of IEEE International Workshop on Neural Networks for Signal proceedingsⅨ . 1999

[6] Volker Roth,Volker Steinhage.Nonlinear discriminant analysis using kernel functions. Proc of Neural Information Processing systems . 1999

[7] Li Hai-feng,Jing Tao,Zhang Ke-shu.Efficient and Robust Feature Extraction by Maximum MarginCriterion. IEEE Transactions on Neural Networks . 2006

[8] LIANG YI-XIONG,LI CHENG-RONG,GONG WEI-GUO,et al.Un-correlated linear discriminantanalysis based on weighted pairwise Fisher criterion. Pattern Recognition . 2007
No related articles found!
Viewed
Full text


Abstract

Cited
  Shared   
  Discussed   
No Suggested Reading articles found!
Copyright © 2017 Journal of Guangdong University of Technology
Add：729 East Dongfeng RD., Guangzhou PRC. Postcode: 510090
Tel：020-37626653,020-37626139,020-37626396
Email：xbzrb@gdut.edu.cn
Supported:Beijing Magtech