Hybrid implementation of neural nets using switched resistor technique

Author

El-Bakry, Hazem M. ; Abo-Elsoud, Mohy A.

Author_Institution

Fac. of Comput. Sci. & Inf., Mansoura Univ., Egypt

fYear

1998

fDate

24-26 Feb 1998

Abstract

A simple hybrid implementation technique for realizing ANNs is presented. The technique is used for realizing a network of two layers in order to make a classification between two characters T and C independent of position, rotation, and scaling. The programmability of adaptive CMOS synaptic weights is achieved by employing the switched resistor (SR) technique. Due to the exponential nature of the bipolar transistors, the sigmoid function is represented by using bipolar transistors. So, the proposed neuron is fully compatible with BiCMOS technology. This implementation technique can be used for different applications

Keywords

BiCMOS integrated circuits; feedforward neural nets; image classification; mixed analogue-digital integrated circuits; optical character recognition; switched networks; ANN; BiCMOS technology compatibility; adaptive CMOS synaptic weight; artificial neural networks; bipolar transistors; characters; classification; hybrid implementation; neural nets; neuron; programmability; sigmoid function; switched resistor technique; Application software; Artificial neural networks; Bipolar transistors; CMOS technology; Character recognition; Multi-layer neural network; Neural networks; Neurons; Resistors; Strontium;

fLanguage

English

Publisher

ieee

Conference_Titel

Radio Science Conference, 1998. NRSC '98. Proceedings of the Fifteenth National

Conference_Location

Cairo

Print_ISBN

0-7803-5121-5

Type

conf

DOI

10.1109/NRSC.1998.711492

Filename

711492