计算机科学 ›› 2008, Vol. 35 ›› Issue (7): 129-132.

• • 上一篇    下一篇

基于动静态组合特征参数的语音识别

王旭 韩志艳 王健 薛丽芳   

  1. 东北大学信息科学与工程学院,沈阳110004
  • 出版日期:2018-11-16 发布日期:2018-11-16
  • 基金资助:
    国家自然科学基金重大资助项目(50477015).

WANG Xu HAN Zhi-yan WANG Jian XUE Li-fang (School of Information Science & Engineering, Northeastern University, Shenyang 110004, China)   

  • Online:2018-11-16 Published:2018-11-16

摘要: 基于语音信号的时变特性,本文提出了动静态特征参数结合的语音信号识别方法,首先在特征参数提取中引入了小波包变换,借助MFCC(Mel-Frequency Cepstrum Coefficient)参数的提取方法,用小波包变换代替傅立叶变换和Mel滤波器组,提取了新的静态特征参数DWPTMFCC(Discrete Wavelet Packet Transfori TlMel-Frequency Coefficient),然后把它与一阶DWPTMFCC差分参数相结合成一个向量,作为一帧语音信号的参数,通过试验和

关键词: 语音识别 特征参数 小波包变换 混沌神经网络

Abstract: In this paper,we propose a new speech recognition method of dynamic and static feature integration. Wavelet packet transformation method is introduced to feature parameters in virtue of MFCC(Mel-Frequency Cepstrum Coefficient), then combined difference fe

Key words: Speech recognition, Feature parameter, Wavelet packet transformation, Chaotic neural network

No related articles found!
Viewed
Full text


Abstract

Cited

  Shared   
  Discussed   
No Suggested Reading articles found!