• 中国计算机学会会刊
  • 中国科技核心期刊
  • 中文核心期刊

J4 ›› 2008, Vol. 30 ›› Issue (4): 139-141.

• 论文 • 上一篇    下一篇

基于脉冲神经网络的语音识别方法的初步探究

曹平 陈盼 章文彬 张潮   

  • 出版日期:2008-04-01 发布日期:2010-05-19

  • Online:2008-04-01 Published:2010-05-19

摘要:

本文提出了一种基于脉冲神经网络SNN的语音识剐方法。该方法以H-H脉冲神经网络模型为基础,采用圆映射的脉冲编码理论将脉冲序列转化为符号序列;同时,以符号序列的距离函数作为语音识别的相似度函数,结合SNN计算能力和实时性强等优势,对语音识别问题进行了初步探讨和研究。实验结果表明,本文提出的特别的研究方法是可行的,
具有深入研究的价值。

关键词: 脉冲神经网络 语音识别 H&mdash, H方程 圆映射

Abstract:

In this paper, a method of speech recognition based on Spiking Neural Networks(SNNs) is presented. Based on the H-H spiking neural model, this method can transform a sequence of spikes to a symbolic sequence with the circle map spiking coding. Meanwhile we use the distance function of the symbolic s equence as the similarity function of speech recognition, consider the merits of SNN's strong computing capability and real-timeness, and perform an in nitial study of the speech recognition problem. The experiment shows that this method is feasible and deserves an in-depth research.

Key words: spiking neural networks, speech recognition, H-H equation, circle map