Gaussian decomposition for robust face recognition

Fumihiko Sakaue, Takeshi Shakunaga

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

Abstract

This paper discusses Gaussian decomposition of facial images for robust recognition. While it cannot sufficiently extract an effective component, it can decompose an image into two effective components, the filtered image and its residual. The Gaussian component represents rough information for a lighting condition and small individuality. The residual represents individuality and the other information including small noise. The two components complement each other and they are evaluated independently in the framework of eigenface method. The image decomposition can also collaborate with parallel partial projections for robust recognition.

Original languageEnglish
Title of host publicationComputer Vision - ACCV 2006 - 7th Asian Conference on Computer Vision, Proceedings
Pages110-119
Number of pages10
DOIs
Publication statusPublished - Jun 15 2006
Event7th Asian Conference on Computer Vision, ACCV 2006 - Hyderabad, India
Duration: Jan 13 2006Jan 16 2006

Publication series

NameLecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)
Volume3851 LNCS
ISSN (Print)0302-9743
ISSN (Electronic)1611-3349

Other

Other7th Asian Conference on Computer Vision, ACCV 2006
CountryIndia
CityHyderabad
Period1/13/061/16/06

ASJC Scopus subject areas

  • Theoretical Computer Science
  • Computer Science(all)

Fingerprint Dive into the research topics of 'Gaussian decomposition for robust face recognition'. Together they form a unique fingerprint.

  • Cite this

    Sakaue, F., & Shakunaga, T. (2006). Gaussian decomposition for robust face recognition. In Computer Vision - ACCV 2006 - 7th Asian Conference on Computer Vision, Proceedings (pp. 110-119). (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics); Vol. 3851 LNCS). https://doi.org/10.1007/11612032_12