Presentation 2000/2/23
Color Recognition System of Optical Neural Network Based on a Novel Method for Self-Organization of the Learning Procedures.
Keishi Takahashi, Naozumi Kosaki, Takayuki Okudaira, Wakao Sasaki,
PDF Download Page PDF download Page Link
Abstract(in Japanese) (See Japanese page)
Abstract(in English) This paper treats a novel scheme that can efficiently organize by itself the learning procedures as back propagation method or genetic algorithms by monitoring the learning performance in a color recognition system of optical neural network. By this scheme, we can avoid local minimum problems, and the color recognition system makes the learning processes fast and good. As a consequence, more naturalized color recognition like human can be performed.
Keyword(in Japanese) (See Japanese page)
Keyword(in English) optical neural network / color recognition / genetic algorithm
Paper # OFT99-56,OPE99-144
Date of Issue

Conference Information
Committee OFT
Conference Date 2000/2/23(1days)
Place (in Japanese) (See Japanese page)
Place (in English)
Topics (in Japanese) (See Japanese page)
Topics (in English)
Chair
Vice Chair
Secretary
Assistant

Paper Information
Registration To Optical Fiber Technology (OFT)
Language JPN
Title (in Japanese) (See Japanese page)
Sub Title (in Japanese) (See Japanese page)
Title (in English) Color Recognition System of Optical Neural Network Based on a Novel Method for Self-Organization of the Learning Procedures.
Sub Title (in English)
Keyword(1) optical neural network
Keyword(2) color recognition
Keyword(3) genetic algorithm
1st Author's Name Keishi Takahashi
1st Author's Affiliation Faculty of Engineering, Doshisha University()
2nd Author's Name Naozumi Kosaki
2nd Author's Affiliation Faculty of Engineering, Doshisha University
3rd Author's Name Takayuki Okudaira
3rd Author's Affiliation Faculty of Engineering, Doshisha University
4th Author's Name Wakao Sasaki
4th Author's Affiliation Faculty of Engineering, Doshisha University
Date 2000/2/23
Paper # OFT99-56,OPE99-144
Volume (vol) vol.99
Number (no) 650
Page pp.pp.-
#Pages 6
Date of Issue