Recognizing sound signals through spiking neurons and spike-timing-dependent plasticity

Abstract

Spiking Neural Networks (SNNs) are regarded as brain-inspired neural networks. Most SNNs described spiking neurons with the leaky integrate-and-fire model, which does not incorporate biological properties of real neurons. In this paper, a model motivated by the human auditory pathway is proposed to explore the possible sound signals recognition mechanism based on the biological dynamic properties of Hodgkin-Huxley (HH) neurons and the spike-timing-dependent-plasticity (STDP) rule of synapses. The first mechanism is that HH neurons have the property of frequency selective response. They only respond to their characteristic frequencies in burst spike trains, which makes the recognition of sound intensity based on the dynamic neurons become possible. The second mechanism is that according to the STDP rule, a synaptic connection structure is formed, and the frequency and the intensity information of input signals are stored in the synaptic delay times. Finally, the neural networks recognize sound signals with spatiotemporal firing patterns.