Spatially smooth subspace face recognition using LOG and DOG penalties

Wangmeng Zuo, Lei Liu, Kuanquan Wang, Dapeng Zhang

Research output: Chapter in book / Conference proceedingConference article published in proceeding or bookAcademic researchpeer-review

6 Citations (Scopus)

Abstract

Subspace face recognition methods have been widely investigated in the last few decades. Since the pixels of an image are spatially correlated and facial images are generally considered to be spatially smoothing, several spatially smooth subspace methods have been proposed for face recognition. In this paper, we first survey the progress and problems in current spatially smooth subspace face recognition methods. Using the penalized subspace learning framework, we then proposed two novel penalty functions, Laplacian of Gaussian (LOG) and Derivative of Gaussian (DOG), for subspace face recognition. LOG and DOG penalties introduce a scale parameter, and thus are more flexible in controlling the degree of smoothness. Experimental results indicate that the proposed methods are effective for face recognition, and achieve higher recognition accuracy than the original subspace methods.
Original languageEnglish
Title of host publicationAdvances in Neural Networks - ISNN 2009 - 6th International Symposium on Neural Networks, ISNN 2009, Proceedings
Pages439-448
Number of pages10
EditionPART 3
DOIs
Publication statusPublished - 10 Sep 2009
Event6th International Symposium on Neural Networks, ISNN 2009 - Wuhan, China
Duration: 26 May 200929 May 2009

Publication series

NameLecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)
NumberPART 3
Volume5553 LNCS
ISSN (Print)0302-9743
ISSN (Electronic)1611-3349

Conference

Conference6th International Symposium on Neural Networks, ISNN 2009
Country/TerritoryChina
CityWuhan
Period26/05/0929/05/09

Keywords

  • Derivative of Gaussian
  • Face recognition
  • Laplacian of Gaussian
  • Regularization
  • Subspace analysis

ASJC Scopus subject areas

  • Theoretical Computer Science
  • Computer Science(all)

Cite this