基于混沌神经网络的语音识别方法

Speech Recognition Method Based on Chaotic Neural Network

基于语音信号的时变特性，研究了神经网络语音识别的方法．把混沌特性引入到神经元，构造了一种新的多层混沌神经网络结构，同时推导了相应的学习算法．把这种混沌神经网络用于语音识别，并与常用的神经网络语音识别方法作了比较．实验结果表明，混沌神经网络方法的平均识别率要高于同等条件下常用神经网络方法的识别率．

Speech recognition using neural networks was investigated.Especially,chaotic dynamics was introduced to neurons,and a multilayer chaotic neural network(CNN) architecture was built.A learning algorithm was also derived to train weights of the network.We apply the CNN to speech recognition and compare the performance of the network with those of recurrent neural network and time delay neural network. The experimental results show that the CNN method outperforms the other neural network methods with respect to average recognition rate.