CMOS architecture of synchronous pulse-coupled neural network and its application to image processing

Abstract

This paper presents a compact architecture for CMOS implementation of a PCNN and its application to image processing. A computational style described in this article mimics a biological neural network using pulse stream signaling and analog summation ant} multiplication. Pulse-stream encoding technique utilizes pulse streams to carry information and control analog circuitry, while storing fUrther analog information on the time axis. The structural form of the pulse-coupled neuron is presented first, then its application to image processing and the synchronization effect between neighboring neurons are demonstrated.

Gabor-Like Image Filtering Using a Neural Microcircuit

“…Technical implementations of neural circuits aim to copy some of the image segmentation/analysis/decomposition properties exhibited by mammalian visual pathways. VLSIbased spiking neurons so far have mainly exploited the synchronization patterns of locally coupled spiking neurons using various synapse adaptation rules [2,3]. However, this type of network pattern is relegated to simple image analysis.…”

“…The ones analyzed in [6] are using excitatory and inhibitory synapses, but purely inhibitory microcircuits have also been found in mammals [7]. In both references, [6] and [8], it is postulated that extracting correlations between inputs is one of the major processing functions, with [8] presuming that inhibitory synapses are necessary to achieve the synchronous activity among neurons required for correlation detection [2,4]. This is in contrast with the work reported in [2,3,5], where purely excitatory synapses also achieve synchronous neuronal activity for correlated inputs.…”

“…In both references, [6] and [8], it is postulated that extracting correlations between inputs is one of the major processing functions, with [8] presuming that inhibitory synapses are necessary to achieve the synchronous activity among neurons required for correlation detection [2,4]. This is in contrast with the work reported in [2,3,5], where purely excitatory synapses also achieve synchronous neuronal activity for correlated inputs. So, while neural microcircuits consisting solely of excitatory synapses…”