Face recognition by inverse fisher discriminant features

Xiao Sheng Zhuang; Dao Qing Dai; Pong Chi Yuen

Face recognition by inverse fisher discriminant features

Xiao Sheng Zhuang, Dao Qing Dai^*, Pong Chi Yuen

^*Corresponding author for this work

Department of Computer Science

Research output: Chapter in book/report/conference proceeding › Conference proceeding › peer-review

Abstract

For face recognition task the PCA plus LDA technique is a famous two-phrase framework to deal with high dimensional space and singular cases. In this paper, we examine the theory of this framework: (1) LDA can still fail even after PCA procedure. (2) Some small principal components that might be essential for classification are thrown away after PCA step. (3) The null space of the within-class scatter matrix S_w contains discriminative information for classification. To eliminate these deficiencies of the PCA plus LDA method we thus develop a new framework by introducing an inverse Fisher criterion and adding a constrain in PCA procedure so that the singularity phenomenon will not occur. Experiment results suggest that this new approach works well.

Original language	English
Title of host publication	Advances in Biometrics - International Conference, ICB 2006, Proceedings
Pages	92-98
Number of pages	7
Publication status	Published - 2006
Event	International Conference on Biometrics, ICB 2006 - Hong Kong, China Duration: 5 Jan 2006 → 7 Jan 2006

Publication series

Name	Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)
Volume	3832 LNCS
ISSN (Print)	0302-9743
ISSN (Electronic)	1611-3349

Conference

Conference	International Conference on Biometrics, ICB 2006
Country/Territory	China
City	Hong Kong
Period	5/01/06 → 7/01/06

Scopus Subject Areas

Theoretical Computer Science
General Computer Science

Cite this

@inproceedings{6f0f775c576b4f3f9bf0e523994b4f55,

title = "Face recognition by inverse fisher discriminant features",

abstract = "For face recognition task the PCA plus LDA technique is a famous two-phrase framework to deal with high dimensional space and singular cases. In this paper, we examine the theory of this framework: (1) LDA can still fail even after PCA procedure. (2) Some small principal components that might be essential for classification are thrown away after PCA step. (3) The null space of the within-class scatter matrix Sw contains discriminative information for classification. To eliminate these deficiencies of the PCA plus LDA method we thus develop a new framework by introducing an inverse Fisher criterion and adding a constrain in PCA procedure so that the singularity phenomenon will not occur. Experiment results suggest that this new approach works well.",

author = "Zhuang, {Xiao Sheng} and Dai, {Dao Qing} and Yuen, {Pong Chi}",

year = "2006",

language = "English",

isbn = "3540311114",

series = "Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)",

pages = "92--98",

booktitle = "Advances in Biometrics - International Conference, ICB 2006, Proceedings",

}

Zhuang, XS, Dai, DQ & Yuen, PC 2006, Face recognition by inverse fisher discriminant features. in Advances in Biometrics - International Conference, ICB 2006, Proceedings. Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics), vol. 3832 LNCS, pp. 92-98, International Conference on Biometrics, ICB 2006, Hong Kong, China, 5/01/06.

Face recognition by inverse fisher discriminant features. / Zhuang, Xiao Sheng; Dai, Dao Qing; Yuen, Pong Chi.
Advances in Biometrics - International Conference, ICB 2006, Proceedings. 2006. p. 92-98 (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics); Vol. 3832 LNCS).

Research output: Chapter in book/report/conference proceeding › Conference proceeding › peer-review

TY - GEN

T1 - Face recognition by inverse fisher discriminant features

AU - Zhuang, Xiao Sheng

AU - Dai, Dao Qing

AU - Yuen, Pong Chi

PY - 2006

Y1 - 2006

N2 - For face recognition task the PCA plus LDA technique is a famous two-phrase framework to deal with high dimensional space and singular cases. In this paper, we examine the theory of this framework: (1) LDA can still fail even after PCA procedure. (2) Some small principal components that might be essential for classification are thrown away after PCA step. (3) The null space of the within-class scatter matrix Sw contains discriminative information for classification. To eliminate these deficiencies of the PCA plus LDA method we thus develop a new framework by introducing an inverse Fisher criterion and adding a constrain in PCA procedure so that the singularity phenomenon will not occur. Experiment results suggest that this new approach works well.

AB - For face recognition task the PCA plus LDA technique is a famous two-phrase framework to deal with high dimensional space and singular cases. In this paper, we examine the theory of this framework: (1) LDA can still fail even after PCA procedure. (2) Some small principal components that might be essential for classification are thrown away after PCA step. (3) The null space of the within-class scatter matrix Sw contains discriminative information for classification. To eliminate these deficiencies of the PCA plus LDA method we thus develop a new framework by introducing an inverse Fisher criterion and adding a constrain in PCA procedure so that the singularity phenomenon will not occur. Experiment results suggest that this new approach works well.

UR - http://www.scopus.com/inward/record.url?scp=33744955667&partnerID=8YFLogxK

M3 - Conference proceeding

AN - SCOPUS:33744955667

SN - 3540311114

SN - 9783540311119

T3 - Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)

SP - 92

EP - 98

BT - Advances in Biometrics - International Conference, ICB 2006, Proceedings

T2 - International Conference on Biometrics, ICB 2006

Y2 - 5 January 2006 through 7 January 2006

ER -

Face recognition by inverse fisher discriminant features

Abstract

Publication series

Conference

Scopus Subject Areas

Other files and links

Fingerprint

Cite this