Publication result detail

FACE RECOGNITION BY MEANS OF NEW ALGORITHMS

BOHÁČ, M.; LÝSEK, J.; MOTYČKA, A.; CEPL, M.

Original Title

FACE RECOGNITION BY MEANS OF NEW ALGORITHMS

English Title

FACE RECOGNITION BY MEANS OF NEW ALGORITHMS

Type

Paper in proceedings (conference paper)

Original Abstract

This contribution describes new methods of face recognition. The methods Principal Component Analysis (PCA) and Evolutionary Principal Component Analysis (EPCA) with genetic algorithm (GA) are described and implemented. The Radial Basis Function (RBF) neural network is used in the way of classifier and genetic algorithm is used for RBF neural network training in one case and for selecting eigenvectors from PCA method in other case. Evolutionary PCA with genetic algorithm outperform other methods tested in this project on the testing database

English abstract

This contribution describes new methods of face recognition. The methods Principal Component Analysis (PCA) and Evolutionary Principal Component Analysis (EPCA) with genetic algorithm (GA) are described and implemented. The Radial Basis Function (RBF) neural network is used in the way of classifier and genetic algorithm is used for RBF neural network training in one case and for selecting eigenvectors from PCA method in other case. Evolutionary PCA with genetic algorithm outperform other methods tested in this project on the testing database

Keywords

face recognition, principal component analysis, radial basis neural network, genetic algorithm

Key words in English

face recognition, principal component analysis, radial basis neural network, genetic algorithm

Authors

BOHÁČ, M.; LÝSEK, J.; MOTYČKA, A.; CEPL, M.

RIV year

2012

Released

08.06.2011

ISBN

978-80-214-4302-0

Book

Mendel 2011

Pages from

324

Pages to

329

Pages count

6

BibTex

@inproceedings{BUT74826,
  author="Martin {Boháč} and Jiří {Lýsek} and Arnošt {Motyčka} and Miroslav {Cepl}",
  title="FACE RECOGNITION BY MEANS OF NEW ALGORITHMS",
  booktitle="Mendel 2011",
  year="2011",
  pages="324--329",
  isbn="978-80-214-4302-0"
}