Computer Engineering and Applications ›› 2009, Vol. 45 ›› Issue (19): 150-152.DOI: 10.3778/j.issn.1002-8331.2009.19.046

• 图形、图像、模式识别 • Previous Articles     Next Articles

Noise-robust speech recognition based on wavelet network and RBF network

HOU Xue-mei   

  1. Department of Information & Control,Xi’an Institute of Post and Telecommunication,Xi’an 710121,China
  • Received:2008-04-16 Revised:2008-07-07 Online:2009-07-01 Published:2009-07-01
  • Contact: HOU Xue-mei

小波网络和RBF网络的抗噪语音识别

侯雪梅   

  1. 西安邮电学院 信息与控制系,西安 710121
  • 通讯作者: 侯雪梅

Abstract: To solve the problem that recognition rates of speech recognition systems decrease in the noisy environment presently,uses character possessing wavelet neural network which integrates the good time-field local property of wavelet transform,uses character possessing RBF neural network,which have best classification ability and recognize ability etc.This paper constructes a wavelet-RBF neural network structure using Morlet mother-wavelet as wavelet basis stead of activate function in the RBF network,adopts whole supervision algorithm and realizes a noise-robust speech recognition system based on wavelet network RBF net-work.The experiment results show that the system has better identify effect than RBF network especially has stronger Robust under noisy environment.

Key words: speech recognition, RBF neural network, wavelet neural network

摘要: 针对目前在噪音环境下语音识别系统性能较差的问题,利用小波神经网络融合了小波变换良好的时频局域化性质和RBF神经网络具有最佳分类能力和辨识能力等特性。构建了一个用小波基替代RBF网络中激活函数的小波-RBF神经网络结构,并采用全监督训练算法,实现了基于小波-RBF网络的抗噪语音识别系统。实验结果表明该系统比RBF网络具有更好的识别效果,尤其在噪声环境下,具有更强的鲁棒性。

关键词: 语音识别, RBF神经网络, 小波神经网络