Publications

Flexible multi-classifier architecture for face recognition systems

Alexandre Lemieux and Marc Parizeau

Abstract - This paper presents face recognition results obtained using a multi-classifier system (MCS) with Borda count voting. Experiments were conducted on complete sections of the FERET face database with 4 different algorithms: embedded HMM, DCT, EigenFaces andEigenObjects. Particular classifier ensembles yielded almost 6% of improvement over the individual techniques. In order to facilitate experiments on classifier combinations and decision rules comparison, a flexible MCS software architecture based on object oriented principles is also presented. It allows runtime modifications to the algorithms employed and dynamical selection of classifiers. This architecture can be applied to any pattern recognition problem.

Bibtex:

@article{Lemieux442,
author    = { Alexandre Lemieux and Marc Parizeau },
title     = { Flexible multi-classifier architecture for face recognition systems },
year      = { 2003 },
journal   = { Vision Interface },
keywords  = { EigenFaces, HMM, Face Recognition, Multi-classifier, DCT },
location  = { Halifax },
language  = { English }
}

Last modification: 2003/04/09 by alemieux

		Home \| About \| People \| Research \| Publications \| Events \| Profile
		©2002-. Computer Vision and Systems Laboratory. All rights reserved