Face recognition based on ICA combined with FLD

Juneho Yi, Jongsun Kim, Jongmoo Choi, Junghyun Han, Eunseok Lee

Research output: Contribution to journalArticle

10 Citations (Scopus)

Abstract

Recently in face recognition, as opposed to our expectation, the performance of an ICA (Independent Component Analysis) method combined with LDA (Linear Discriminant Analysis) was reported as lower than an ICA only based method. This research points out that (ICA+LDA) methods have not got a fair comparison for evaluating its recognition performance. In order to incorporate class specific information into ICA, we have employed FLD (Fisher Linear Discriminant) and have proposed our (ICA+FLD) method. In the experimental results, we report that our (ICA+FLD) method has better performance than ICA only based methods as well as other representative methods such as Eigenface and Fisherface methods.

Original languageEnglish
Pages (from-to)10-18
Number of pages9
JournalLecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)
Volume2359 LNCS
Publication statusPublished - 2002 Dec 1
Externally publishedYes

Fingerprint

Independent component analysis
Independent Component Analysis
Face recognition
Face Recognition
Discriminant
Discriminant analysis
Discriminant Analysis
Eigenface
Combined Method
Experimental Results

ASJC Scopus subject areas

  • Computer Science(all)
  • Theoretical Computer Science

Cite this

Face recognition based on ICA combined with FLD. / Yi, Juneho; Kim, Jongsun; Choi, Jongmoo; Han, Junghyun; Lee, Eunseok.

In: Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics), Vol. 2359 LNCS, 01.12.2002, p. 10-18.

Research output: Contribution to journalArticle

@article{dfbb2fb4c0cb4fcc8a13f5be576e1474,
title = "Face recognition based on ICA combined with FLD",
abstract = "Recently in face recognition, as opposed to our expectation, the performance of an ICA (Independent Component Analysis) method combined with LDA (Linear Discriminant Analysis) was reported as lower than an ICA only based method. This research points out that (ICA+LDA) methods have not got a fair comparison for evaluating its recognition performance. In order to incorporate class specific information into ICA, we have employed FLD (Fisher Linear Discriminant) and have proposed our (ICA+FLD) method. In the experimental results, we report that our (ICA+FLD) method has better performance than ICA only based methods as well as other representative methods such as Eigenface and Fisherface methods.",
author = "Juneho Yi and Jongsun Kim and Jongmoo Choi and Junghyun Han and Eunseok Lee",
year = "2002",
month = "12",
day = "1",
language = "English",
volume = "2359 LNCS",
pages = "10--18",
journal = "Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)",
issn = "0302-9743",
publisher = "Springer Verlag",

}

TY - JOUR

T1 - Face recognition based on ICA combined with FLD

AU - Yi, Juneho

AU - Kim, Jongsun

AU - Choi, Jongmoo

AU - Han, Junghyun

AU - Lee, Eunseok

PY - 2002/12/1

Y1 - 2002/12/1

N2 - Recently in face recognition, as opposed to our expectation, the performance of an ICA (Independent Component Analysis) method combined with LDA (Linear Discriminant Analysis) was reported as lower than an ICA only based method. This research points out that (ICA+LDA) methods have not got a fair comparison for evaluating its recognition performance. In order to incorporate class specific information into ICA, we have employed FLD (Fisher Linear Discriminant) and have proposed our (ICA+FLD) method. In the experimental results, we report that our (ICA+FLD) method has better performance than ICA only based methods as well as other representative methods such as Eigenface and Fisherface methods.

AB - Recently in face recognition, as opposed to our expectation, the performance of an ICA (Independent Component Analysis) method combined with LDA (Linear Discriminant Analysis) was reported as lower than an ICA only based method. This research points out that (ICA+LDA) methods have not got a fair comparison for evaluating its recognition performance. In order to incorporate class specific information into ICA, we have employed FLD (Fisher Linear Discriminant) and have proposed our (ICA+FLD) method. In the experimental results, we report that our (ICA+FLD) method has better performance than ICA only based methods as well as other representative methods such as Eigenface and Fisherface methods.

UR - http://www.scopus.com/inward/record.url?scp=57949099774&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=57949099774&partnerID=8YFLogxK

M3 - Article

AN - SCOPUS:57949099774

VL - 2359 LNCS

SP - 10

EP - 18

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

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

SN - 0302-9743

ER -