Pattern recognition using a hierarchical neural network

Pau-Choo Chung, E. Liang Chen, Ching Tsorng Tsai

Research output: Contribution to conferencePaper

1 Citation (Scopus)

Abstract

A pattern recognition system based on hierarchical neural networks is proposed in this paper. The hierarchical system consists of two levels of networks: low-level for feature extraction and high-level for object recognition. The low-level is a Competitive Hopfield Neural Network (CHNN) which detects the dominant points of a target shape to be the pattern features, based on the minimization of a cost function. The CHNN is implemented by incorporating a winner-take-all strategy in the network. By imposing the winner-take-all rule, we are relieved from the suffering of deciding the suitable values of the weighting factors in the cost function. Furthermore, from the experimental results, we also find that the proposed CHNN performs very well in determining the dominant points of a target shape. After the features have been extracted, they are applied to the high-level multilayered network for object recognition. Because the multilayer network has high robustness to the pattern variations, the recognition system is found to posses high noise tolerance capability. Experimental results show that the system can recognize all the objects correctly when the percentage of noises is under 10%. Even when the percentage of noises reaches 40%, the recognition ratio is still over 90%.

Original languageEnglish
Pages3104-3109
Number of pages6
Publication statusPublished - 1994 Dec 1
EventProceedings of the 1994 IEEE International Conference on Neural Networks. Part 1 (of 7) - Orlando, FL, USA
Duration: 1994 Jun 271994 Jun 29

Other

OtherProceedings of the 1994 IEEE International Conference on Neural Networks. Part 1 (of 7)
CityOrlando, FL, USA
Period94-06-2794-06-29

Fingerprint

Hopfield neural networks
Pattern recognition
Object recognition
Neural networks
Cost functions
Pattern recognition systems
Hierarchical systems
Feature extraction
Multilayers

All Science Journal Classification (ASJC) codes

  • Software

Cite this

Chung, P-C., Chen, E. L., & Tsai, C. T. (1994). Pattern recognition using a hierarchical neural network. 3104-3109. Paper presented at Proceedings of the 1994 IEEE International Conference on Neural Networks. Part 1 (of 7), Orlando, FL, USA, .
Chung, Pau-Choo ; Chen, E. Liang ; Tsai, Ching Tsorng. / Pattern recognition using a hierarchical neural network. Paper presented at Proceedings of the 1994 IEEE International Conference on Neural Networks. Part 1 (of 7), Orlando, FL, USA, .6 p.
@conference{ce02a24a459c47b1b68ce361e7deeedc,
title = "Pattern recognition using a hierarchical neural network",
abstract = "A pattern recognition system based on hierarchical neural networks is proposed in this paper. The hierarchical system consists of two levels of networks: low-level for feature extraction and high-level for object recognition. The low-level is a Competitive Hopfield Neural Network (CHNN) which detects the dominant points of a target shape to be the pattern features, based on the minimization of a cost function. The CHNN is implemented by incorporating a winner-take-all strategy in the network. By imposing the winner-take-all rule, we are relieved from the suffering of deciding the suitable values of the weighting factors in the cost function. Furthermore, from the experimental results, we also find that the proposed CHNN performs very well in determining the dominant points of a target shape. After the features have been extracted, they are applied to the high-level multilayered network for object recognition. Because the multilayer network has high robustness to the pattern variations, the recognition system is found to posses high noise tolerance capability. Experimental results show that the system can recognize all the objects correctly when the percentage of noises is under 10{\%}. Even when the percentage of noises reaches 40{\%}, the recognition ratio is still over 90{\%}.",
author = "Pau-Choo Chung and Chen, {E. Liang} and Tsai, {Ching Tsorng}",
year = "1994",
month = "12",
day = "1",
language = "English",
pages = "3104--3109",
note = "Proceedings of the 1994 IEEE International Conference on Neural Networks. Part 1 (of 7) ; Conference date: 27-06-1994 Through 29-06-1994",

}

Chung, P-C, Chen, EL & Tsai, CT 1994, 'Pattern recognition using a hierarchical neural network' Paper presented at Proceedings of the 1994 IEEE International Conference on Neural Networks. Part 1 (of 7), Orlando, FL, USA, 94-06-27 - 94-06-29, pp. 3104-3109.

Pattern recognition using a hierarchical neural network. / Chung, Pau-Choo; Chen, E. Liang; Tsai, Ching Tsorng.

1994. 3104-3109 Paper presented at Proceedings of the 1994 IEEE International Conference on Neural Networks. Part 1 (of 7), Orlando, FL, USA, .

Research output: Contribution to conferencePaper

TY - CONF

T1 - Pattern recognition using a hierarchical neural network

AU - Chung, Pau-Choo

AU - Chen, E. Liang

AU - Tsai, Ching Tsorng

PY - 1994/12/1

Y1 - 1994/12/1

N2 - A pattern recognition system based on hierarchical neural networks is proposed in this paper. The hierarchical system consists of two levels of networks: low-level for feature extraction and high-level for object recognition. The low-level is a Competitive Hopfield Neural Network (CHNN) which detects the dominant points of a target shape to be the pattern features, based on the minimization of a cost function. The CHNN is implemented by incorporating a winner-take-all strategy in the network. By imposing the winner-take-all rule, we are relieved from the suffering of deciding the suitable values of the weighting factors in the cost function. Furthermore, from the experimental results, we also find that the proposed CHNN performs very well in determining the dominant points of a target shape. After the features have been extracted, they are applied to the high-level multilayered network for object recognition. Because the multilayer network has high robustness to the pattern variations, the recognition system is found to posses high noise tolerance capability. Experimental results show that the system can recognize all the objects correctly when the percentage of noises is under 10%. Even when the percentage of noises reaches 40%, the recognition ratio is still over 90%.

AB - A pattern recognition system based on hierarchical neural networks is proposed in this paper. The hierarchical system consists of two levels of networks: low-level for feature extraction and high-level for object recognition. The low-level is a Competitive Hopfield Neural Network (CHNN) which detects the dominant points of a target shape to be the pattern features, based on the minimization of a cost function. The CHNN is implemented by incorporating a winner-take-all strategy in the network. By imposing the winner-take-all rule, we are relieved from the suffering of deciding the suitable values of the weighting factors in the cost function. Furthermore, from the experimental results, we also find that the proposed CHNN performs very well in determining the dominant points of a target shape. After the features have been extracted, they are applied to the high-level multilayered network for object recognition. Because the multilayer network has high robustness to the pattern variations, the recognition system is found to posses high noise tolerance capability. Experimental results show that the system can recognize all the objects correctly when the percentage of noises is under 10%. Even when the percentage of noises reaches 40%, the recognition ratio is still over 90%.

UR - http://www.scopus.com/inward/record.url?scp=0028744192&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=0028744192&partnerID=8YFLogxK

M3 - Paper

AN - SCOPUS:0028744192

SP - 3104

EP - 3109

ER -

Chung P-C, Chen EL, Tsai CT. Pattern recognition using a hierarchical neural network. 1994. Paper presented at Proceedings of the 1994 IEEE International Conference on Neural Networks. Part 1 (of 7), Orlando, FL, USA, .