Transputer-based implementation of distributed associative memories

Fabio Ancona, Stefano Rovetta, Rodolfo Zunino


Parallel Computing 23 (1997) 1479-1491
(c) 1997 Elsevier Science B.V.



ABSTRACT

The Distributed Associative Memory (DAM) model operates as a basic
device inside an image-classification system, whose accuracy is
increased by the restoration ability of the associative mechanism.
The paper descibes a methodology for implementing the DAM model on
transputer-based hierarchical parallel architectures. First, a
theoretical analysis reformulates the basic memory model to fit a
practical implementation; then, the parallel system's efficiency is
defined analytically. Experimental results on a significant testbed
confirm the classifier's effectiveness and the consistency of
theoretical predictions.


KEYWORDS

Distributed associative memory; Image-classification system;
Transputer-based system; Parallel architecture; Experimental results