Kernel eigenfaces vs. kernel fisherfaces: Face recognition using kernel methods

Research output: Chapter in Book/Report/Conference proceedingConference contribution

542 Citations (Scopus)

Abstract

Principal Component A nalysis and Fisher Linear Discriminant methods have demonstrated their success in fac edete ction, r ecognition and atrcking. The repr e- sentations in these subspac e methods are based on sec- ond order statistics of the image set, and do not address higher order statistic al dep endencies such as the rela- tionships among three or more pixels. R ecently Higher Order Statistics and Independent Component Analysis (ICA) have been used as informative repr esentations for visual recognition. In this paper, we investigate the use of Kernel Principal Component Analysis and Ker- nel Fisher Linear Discriminant for learning low dimen- sional repr esentations for fac e recognition, which we call Kernel Eigenface and Kernel Fisherface methods. While Eigenface and Fisherface methods aim to find proje ction directions b ased on second order corr elation of samples, Kernel Eigenface and Kernel Fisherface methods provide gener alizationswhich take higher or- der correlations into account. We compare the perfor- mance of kernel methods with classical algorithms such as Eigenface, Fisherface, ICA, and Support Vector Ma- chine (SVM) within the context of appearance-based face recognition problem using two data sets where im- ages vary in pose, scale, lighting and expr ession. Ex- perimental results show that kernel methods provide better r epr esentations and achieve lower error rates for face recognition.

Original languageEnglish
Title of host publicationProceedings - 5th IEEE International Conference on Automatic Face Gesture Recognition, FGR 2002
PublisherIEEE Computer Society
Pages215-220
Number of pages6
ISBN (Print)0769516025, 9780769516028
DOIs
Publication statusPublished - 2002 Jan 1
Event5th IEEE International Conference on Automatic Face Gesture Recognition, FGR 2002 - Washington, DC, United States
Duration: 2002 May 202002 May 21

Publication series

NameProceedings - 5th IEEE International Conference on Automatic Face Gesture Recognition, FGR 2002

Conference

Conference5th IEEE International Conference on Automatic Face Gesture Recognition, FGR 2002
CountryUnited States
CityWashington, DC
Period02/5/2002/5/21

Fingerprint

Higher order statistics
Independent component analysis
Face recognition
Principal component analysis
Support vector machines
Lighting
Pixels
Statistics

All Science Journal Classification (ASJC) codes

  • Computer Vision and Pattern Recognition

Cite this

Yang, M. H. (2002). Kernel eigenfaces vs. kernel fisherfaces: Face recognition using kernel methods. In Proceedings - 5th IEEE International Conference on Automatic Face Gesture Recognition, FGR 2002 (pp. 215-220). [1004157] (Proceedings - 5th IEEE International Conference on Automatic Face Gesture Recognition, FGR 2002). IEEE Computer Society. https://doi.org/10.1109/AFGR.2002.1004157
Yang, Ming Hsuan. / Kernel eigenfaces vs. kernel fisherfaces : Face recognition using kernel methods. Proceedings - 5th IEEE International Conference on Automatic Face Gesture Recognition, FGR 2002. IEEE Computer Society, 2002. pp. 215-220 (Proceedings - 5th IEEE International Conference on Automatic Face Gesture Recognition, FGR 2002).
@inproceedings{522e26ef7c9c451f9602352f1b757f01,
title = "Kernel eigenfaces vs. kernel fisherfaces: Face recognition using kernel methods",
abstract = "Principal Component A nalysis and Fisher Linear Discriminant methods have demonstrated their success in fac edete ction, r ecognition and atrcking. The repr e- sentations in these subspac e methods are based on sec- ond order statistics of the image set, and do not address higher order statistic al dep endencies such as the rela- tionships among three or more pixels. R ecently Higher Order Statistics and Independent Component Analysis (ICA) have been used as informative repr esentations for visual recognition. In this paper, we investigate the use of Kernel Principal Component Analysis and Ker- nel Fisher Linear Discriminant for learning low dimen- sional repr esentations for fac e recognition, which we call Kernel Eigenface and Kernel Fisherface methods. While Eigenface and Fisherface methods aim to find proje ction directions b ased on second order corr elation of samples, Kernel Eigenface and Kernel Fisherface methods provide gener alizationswhich take higher or- der correlations into account. We compare the perfor- mance of kernel methods with classical algorithms such as Eigenface, Fisherface, ICA, and Support Vector Ma- chine (SVM) within the context of appearance-based face recognition problem using two data sets where im- ages vary in pose, scale, lighting and expr ession. Ex- perimental results show that kernel methods provide better r epr esentations and achieve lower error rates for face recognition.",
author = "Yang, {Ming Hsuan}",
year = "2002",
month = "1",
day = "1",
doi = "10.1109/AFGR.2002.1004157",
language = "English",
isbn = "0769516025",
series = "Proceedings - 5th IEEE International Conference on Automatic Face Gesture Recognition, FGR 2002",
publisher = "IEEE Computer Society",
pages = "215--220",
booktitle = "Proceedings - 5th IEEE International Conference on Automatic Face Gesture Recognition, FGR 2002",
address = "United States",

}

Yang, MH 2002, Kernel eigenfaces vs. kernel fisherfaces: Face recognition using kernel methods. in Proceedings - 5th IEEE International Conference on Automatic Face Gesture Recognition, FGR 2002., 1004157, Proceedings - 5th IEEE International Conference on Automatic Face Gesture Recognition, FGR 2002, IEEE Computer Society, pp. 215-220, 5th IEEE International Conference on Automatic Face Gesture Recognition, FGR 2002, Washington, DC, United States, 02/5/20. https://doi.org/10.1109/AFGR.2002.1004157

Kernel eigenfaces vs. kernel fisherfaces : Face recognition using kernel methods. / Yang, Ming Hsuan.

Proceedings - 5th IEEE International Conference on Automatic Face Gesture Recognition, FGR 2002. IEEE Computer Society, 2002. p. 215-220 1004157 (Proceedings - 5th IEEE International Conference on Automatic Face Gesture Recognition, FGR 2002).

Research output: Chapter in Book/Report/Conference proceedingConference contribution

TY - GEN

T1 - Kernel eigenfaces vs. kernel fisherfaces

T2 - Face recognition using kernel methods

AU - Yang, Ming Hsuan

PY - 2002/1/1

Y1 - 2002/1/1

N2 - Principal Component A nalysis and Fisher Linear Discriminant methods have demonstrated their success in fac edete ction, r ecognition and atrcking. The repr e- sentations in these subspac e methods are based on sec- ond order statistics of the image set, and do not address higher order statistic al dep endencies such as the rela- tionships among three or more pixels. R ecently Higher Order Statistics and Independent Component Analysis (ICA) have been used as informative repr esentations for visual recognition. In this paper, we investigate the use of Kernel Principal Component Analysis and Ker- nel Fisher Linear Discriminant for learning low dimen- sional repr esentations for fac e recognition, which we call Kernel Eigenface and Kernel Fisherface methods. While Eigenface and Fisherface methods aim to find proje ction directions b ased on second order corr elation of samples, Kernel Eigenface and Kernel Fisherface methods provide gener alizationswhich take higher or- der correlations into account. We compare the perfor- mance of kernel methods with classical algorithms such as Eigenface, Fisherface, ICA, and Support Vector Ma- chine (SVM) within the context of appearance-based face recognition problem using two data sets where im- ages vary in pose, scale, lighting and expr ession. Ex- perimental results show that kernel methods provide better r epr esentations and achieve lower error rates for face recognition.

AB - Principal Component A nalysis and Fisher Linear Discriminant methods have demonstrated their success in fac edete ction, r ecognition and atrcking. The repr e- sentations in these subspac e methods are based on sec- ond order statistics of the image set, and do not address higher order statistic al dep endencies such as the rela- tionships among three or more pixels. R ecently Higher Order Statistics and Independent Component Analysis (ICA) have been used as informative repr esentations for visual recognition. In this paper, we investigate the use of Kernel Principal Component Analysis and Ker- nel Fisher Linear Discriminant for learning low dimen- sional repr esentations for fac e recognition, which we call Kernel Eigenface and Kernel Fisherface methods. While Eigenface and Fisherface methods aim to find proje ction directions b ased on second order corr elation of samples, Kernel Eigenface and Kernel Fisherface methods provide gener alizationswhich take higher or- der correlations into account. We compare the perfor- mance of kernel methods with classical algorithms such as Eigenface, Fisherface, ICA, and Support Vector Ma- chine (SVM) within the context of appearance-based face recognition problem using two data sets where im- ages vary in pose, scale, lighting and expr ession. Ex- perimental results show that kernel methods provide better r epr esentations and achieve lower error rates for face recognition.

UR - http://www.scopus.com/inward/record.url?scp=84905388890&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=84905388890&partnerID=8YFLogxK

U2 - 10.1109/AFGR.2002.1004157

DO - 10.1109/AFGR.2002.1004157

M3 - Conference contribution

AN - SCOPUS:84905388890

SN - 0769516025

SN - 9780769516028

T3 - Proceedings - 5th IEEE International Conference on Automatic Face Gesture Recognition, FGR 2002

SP - 215

EP - 220

BT - Proceedings - 5th IEEE International Conference on Automatic Face Gesture Recognition, FGR 2002

PB - IEEE Computer Society

ER -

Yang MH. Kernel eigenfaces vs. kernel fisherfaces: Face recognition using kernel methods. In Proceedings - 5th IEEE International Conference on Automatic Face Gesture Recognition, FGR 2002. IEEE Computer Society. 2002. p. 215-220. 1004157. (Proceedings - 5th IEEE International Conference on Automatic Face Gesture Recognition, FGR 2002). https://doi.org/10.1109/AFGR.2002.1004157