| | |
Summary: IMPROVED TRINARY ASSOCIATIVE MEMORY FOR CHARACTER
RECOGNITION
Farid Ahmed ,A. A. S. Awwal and C. L. Philip Chen
Wright State University
Computer Science & Engineering Department
Dayton, OH 45435.
Abstract
Character recognition by trinary associative mem-
ory (TAM) neural network model is investigated.
Three different inner product thresholding schemes,
namely, zero plus average thresholding, arithmetic
mean thresholding and a novel adaptive thresh-
olding are examined. It is shown that the novel
threshold prescription with the permanent inhibi-
tion scheme enhances the convergence as well as
storage capacity of the inner product associative
memory.
1 Introduction
Retrieval of information from partial representation
is an important task of pattern recognition which
|
