Spiking neural network target identification method

Abstract

The invention provides a spiking neural network target identification method. The loss of neuron membrane potential is effectively avoided by constructing the proportional attenuation characteristic of a spiking neuron model, the spiking emission rate of neurons is ensured, and the problem that neuron spiking emission is not activated enough is solved; the problem of neuron spiking emission rate over-activation is effectively reduced by constructing the self-adaptive threshold characteristics of the spiking neuron model.

Description

technical field [0001] The invention belongs to the field of artificial intelligence image processing, in particular to a pulse neural network target recognition method. Background technique [0002] In recent years, the spiking neural network (SNN) algorithm model based on biologically interpretable spiking neurons has become an indispensable research tool for the development of brain-like cognition, learning and memory mechanisms. SNN is a neural network based on sparse event triggering, which has the characteristics of hardware friendliness and energy saving, so the research on deep SNN algorithm is gradually favored by researchers. To train a traditional neural network, the transmission of information relies on high-precision floating-point numbers. In brain systems, information is transmitted in the form of action potentials, which in biological systems are also known as electrical impulses. Inspired by biological systems, SNNs process discrete spike sequence informat...

AI Technical Summary

Problems solved by technology

The existing technology cannot directly convert ANN to SNN equivalently, and the conversion process will cause performance loss in terms of accuracy of network target recognition methods

Images