Independent component analysis of face images

Pong Chi YUEN, J. H. Lai

Research output: Chapter in book/report/conference proceedingConference proceedingpeer-review

5 Citations (Scopus)

Abstract

This paper addresses the problem of face recognition using independent component analysis. As the independent components (IC) are not orthogonal, to represent a face image using the determined ICs, the ICs have to be orthogonalized, where two methods, namely Gram-Schmit Method and Householder Transformation, are proposed. In addition, to find a better set of ICs for face recognition, an efficient IC selection algorithm is developed. Face images with different facial expressions, pose variations and small occlusions are selected to test the ICA face representation and the results are encouraging.

Original languageEnglish
Title of host publicationLecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)
EditorsTomaso Poggio, Seong-Whan Lee, Heinrich H. Bulthoff
PublisherSpringer Verlag
Pages545-553
Number of pages9
ISBN (Print)3540675604, 9783540675600
DOIs
Publication statusPublished - 2000
Event1st IEEE International Workshop on Biologically Motivated Computer Vision, BMCV 2000 - Seoul, Korea, Republic of
Duration: 15 May 200017 May 2000

Publication series

NameLecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)
Volume1811
ISSN (Print)0302-9743
ISSN (Electronic)1611-3349

Conference

Conference1st IEEE International Workshop on Biologically Motivated Computer Vision, BMCV 2000
Country/TerritoryKorea, Republic of
CitySeoul
Period15/05/0017/05/00

Scopus Subject Areas

  • Theoretical Computer Science
  • Computer Science(all)

Fingerprint

Dive into the research topics of 'Independent component analysis of face images'. Together they form a unique fingerprint.

Cite this