Skip to main navigation Skip to search Skip to main content

Optimization of face relevance maps with total classification error minimization

Research output: Chapter in Book/Report/Conference proceedingConference contributionpeer-review

2 Citations (Scopus)

Abstract

This paper presents a concept of optimizing parameters used for solving image identification tasks developed during research aimed at improving recognition of human face images. Effectiveness of closed-set identification is measured in a form of Total Classification Error (TCE) which can be expressed as a function of parameters used for calculating similarity between samples. TCE can be minimized for a defined training set in order to obtain optimal values of the parameters. This method was implemented to optimize face relevance maps applied to improve the Eigenfaces method for human face recognition. Results of the experiments presented in this paper confirm effectiveness of the developed approach.

Original languageEnglish
Title of host publicationImage Analysis and Recognition - 5th International Conference, ICIAR 2008, Proceedings
PublisherSpringer Verlag
Pages935-944
Number of pages10
ISBN (Print)3540698116, 9783540698111
DOIs
Publication statusPublished - 2008
Event5th International Conference on Image Analysis and Recognition, ICIAR 2008 - Povoa de Varzim, Portugal
Duration: 25 Jun 200827 Jun 2008

Publication series

NameLecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)
Volume5112 LNCS
ISSN (Print)0302-9743
ISSN (Electronic)1611-3349

Conference

Conference5th International Conference on Image Analysis and Recognition, ICIAR 2008
Country/TerritoryPortugal
CityPovoa de Varzim
Period25/06/0827/06/08

ASJC Scopus subject areas

  • Theoretical Computer Science
  • General Computer Science

Fingerprint

Dive into the research topics of 'Optimization of face relevance maps with total classification error minimization'. Together they form a unique fingerprint.

Cite this