Odor detection using pulse coupled neural networks

G. Székely, Mary Lou Padgett, Gerry Dozier, T. A. Roppel

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

4 Citations (Scopus)

Abstract

Based on neural structure (not physiology) observed in clinical experiments, an odor image can be constructed for analysis with a cutting-edge image processing procedure termed Pulse Coupled Neural Networks Factoring (PCNNf). Enhancement of an odor image using PCNNf can significantly increase detection accuracy. Selection of the proper parameters for the implementation usually requires analysis by an expert familiar with the application targeted. Once suitable parameters have been selected, the PCNNf procedure is very robust, and can typically be used in a large number of situations similar to the original application. The purpose of this research is to advance the methodology for selecting parameters with reduced input from experts. The approach selected is use of a set of evolutionary algorithms (EAs) to find improved parameter sets and to establish automated procedures for setting bounds on parameters and Weight matrices for particular applications.

Original languageEnglish
Title of host publicationProceedings of the International Joint Conference on Neural Networks
PublisherIEEE
Pages317-321
Number of pages5
Volume1
Publication statusPublished - 1999
EventInternational Joint Conference on Neural Networks (IJCNN'99) - Washington, DC, USA
Duration: Jul 10 1999Jul 16 1999

Other

OtherInternational Joint Conference on Neural Networks (IJCNN'99)
CityWashington, DC, USA
Period7/10/997/16/99

Fingerprint

Odors
Neural networks
Physiology
Evolutionary algorithms
Image processing
Experiments

ASJC Scopus subject areas

  • Software

Cite this

Székely, G., Padgett, M. L., Dozier, G., & Roppel, T. A. (1999). Odor detection using pulse coupled neural networks. In Proceedings of the International Joint Conference on Neural Networks (Vol. 1, pp. 317-321). IEEE.

Odor detection using pulse coupled neural networks. / Székely, G.; Padgett, Mary Lou; Dozier, Gerry; Roppel, T. A.

Proceedings of the International Joint Conference on Neural Networks. Vol. 1 IEEE, 1999. p. 317-321.

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

Székely, G, Padgett, ML, Dozier, G & Roppel, TA 1999, Odor detection using pulse coupled neural networks. in Proceedings of the International Joint Conference on Neural Networks. vol. 1, IEEE, pp. 317-321, International Joint Conference on Neural Networks (IJCNN'99), Washington, DC, USA, 7/10/99.
Székely G, Padgett ML, Dozier G, Roppel TA. Odor detection using pulse coupled neural networks. In Proceedings of the International Joint Conference on Neural Networks. Vol. 1. IEEE. 1999. p. 317-321
Székely, G. ; Padgett, Mary Lou ; Dozier, Gerry ; Roppel, T. A. / Odor detection using pulse coupled neural networks. Proceedings of the International Joint Conference on Neural Networks. Vol. 1 IEEE, 1999. pp. 317-321
@inproceedings{1a7868b2c5ba4b64889ef006474ecc8c,
title = "Odor detection using pulse coupled neural networks",
abstract = "Based on neural structure (not physiology) observed in clinical experiments, an odor image can be constructed for analysis with a cutting-edge image processing procedure termed Pulse Coupled Neural Networks Factoring (PCNNf). Enhancement of an odor image using PCNNf can significantly increase detection accuracy. Selection of the proper parameters for the implementation usually requires analysis by an expert familiar with the application targeted. Once suitable parameters have been selected, the PCNNf procedure is very robust, and can typically be used in a large number of situations similar to the original application. The purpose of this research is to advance the methodology for selecting parameters with reduced input from experts. The approach selected is use of a set of evolutionary algorithms (EAs) to find improved parameter sets and to establish automated procedures for setting bounds on parameters and Weight matrices for particular applications.",
author = "G. Sz{\'e}kely and Padgett, {Mary Lou} and Gerry Dozier and Roppel, {T. A.}",
year = "1999",
language = "English",
volume = "1",
pages = "317--321",
booktitle = "Proceedings of the International Joint Conference on Neural Networks",
publisher = "IEEE",

}

TY - GEN

T1 - Odor detection using pulse coupled neural networks

AU - Székely, G.

AU - Padgett, Mary Lou

AU - Dozier, Gerry

AU - Roppel, T. A.

PY - 1999

Y1 - 1999

N2 - Based on neural structure (not physiology) observed in clinical experiments, an odor image can be constructed for analysis with a cutting-edge image processing procedure termed Pulse Coupled Neural Networks Factoring (PCNNf). Enhancement of an odor image using PCNNf can significantly increase detection accuracy. Selection of the proper parameters for the implementation usually requires analysis by an expert familiar with the application targeted. Once suitable parameters have been selected, the PCNNf procedure is very robust, and can typically be used in a large number of situations similar to the original application. The purpose of this research is to advance the methodology for selecting parameters with reduced input from experts. The approach selected is use of a set of evolutionary algorithms (EAs) to find improved parameter sets and to establish automated procedures for setting bounds on parameters and Weight matrices for particular applications.

AB - Based on neural structure (not physiology) observed in clinical experiments, an odor image can be constructed for analysis with a cutting-edge image processing procedure termed Pulse Coupled Neural Networks Factoring (PCNNf). Enhancement of an odor image using PCNNf can significantly increase detection accuracy. Selection of the proper parameters for the implementation usually requires analysis by an expert familiar with the application targeted. Once suitable parameters have been selected, the PCNNf procedure is very robust, and can typically be used in a large number of situations similar to the original application. The purpose of this research is to advance the methodology for selecting parameters with reduced input from experts. The approach selected is use of a set of evolutionary algorithms (EAs) to find improved parameter sets and to establish automated procedures for setting bounds on parameters and Weight matrices for particular applications.

UR - http://www.scopus.com/inward/record.url?scp=0033313108&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=0033313108&partnerID=8YFLogxK

M3 - Conference contribution

AN - SCOPUS:0033313108

VL - 1

SP - 317

EP - 321

BT - Proceedings of the International Joint Conference on Neural Networks

PB - IEEE

ER -