Nonspecific Speech Recognition based on HMM/LVQ Hybrid Network

摘要：

A novel method of speech recognition,which is based on HMM/LVQ1-LVQ2, is proposed in this paper.First, the MFCC, △MFCC and △△MFCC extraction algorithms are introduced,then these coefficients are normalized by HMM-based Viterbi method,after that,the normalized feature sequences are obtained.The recognition is first to learn coarsely by using LVQ1 and then to learn finely by LVQ2. Finally the result of the simulation is given, which shows that the proposed algorithm can be used to improve the recognition rates effec tively,in comparison with HMM used alone or LVQ1-LVQ2 hybrid network recognition, especially for nonspecific speech.

关键词： speech recognition MFCC HMM normalization Viterbi algorithm LVQ

作者: Liang Shuling Wang Chaoli Du Jiaming

作者单位: School of Optical-electrical and Computer Engineering University of Shanghai for Science and Technol School of Optical-electrical and Computer Engineering University of Shanghai for Science and Technol

会议类型: 国际会议

会议名称: 2009 Second International Conference on Intelligent Computation Technology and Automation(2009 第二届IEEE智能计算与自动化国际会议 ICICTA 2009)

会议地点: 长沙

会议语种:英文

页码: 645-648

在线出版日期: 2009-10-10（万方平台首次上网日期，不代表论文的发表时间）