2D-Motion Detection using SNNs with Graphene-Insulator-Graphene Memristive Synapses

TL;DR

This paper presents a biologically plausible, energy-efficient 2D motion detection system using spiking neural networks with graphene-insulator-graphene memristive synapses, demonstrating accurate detection through detailed simulations.

Contribution

It introduces a novel 2D motion detection architecture employing CMOS neurons and ultra-low power RRAM synapses, validated by transistor-level simulations.

Findings

Accurately detects complex 2D object motions

Uses ultra-low power memristive synapses

Validated by detailed transistor-level simulations

Abstract

The event-driven nature of spiking neural networks makes them biologically plausible and more energy-efficient than artificial neural networks. In this work, we demonstrate motion detection of an object in a two-dimensional visual field. The network architecture presented here is biologically plausible and uses CMOS analog leaky integrate-and-fire neurons and ultra-low power multi-layer RRAM synapses. Detailed transistorlevel SPICE simulations show that the proposed structure can accurately and reliably detect complex motions of an object in a two-dimensional visual field.