Speech Command Recognition Based on Convolutional Spiking Neural Networks

Abstract

This article presents a new technique for speech recognition that combines Convolutional Neural Networks (CNNs) with Spiking Neural Networks (SNNs) to create an SNNCNN model. The model is tested on the Google Speech Command Dataset and achieves an accuracy of 72.03%, which is similar to the current state-of-the-art speech recognition methods. The study also compares the performance of the SNNCNN model with other SNN models that use Multi-Layer Perceptrons (MLPs) and traditional Artificial Neural Networks (ANNs). The results show that the CNN-based SNNs outperform both MLPs and ANNs, demonstrating the superiority of the proposed model. The approach presented in this study can potentially be applied to other speech recognition tasks and could lead to further improvements in the field.