摘要
本文在简单介绍三种词组语音识别模型后,提出了采用离散隐马尔柯夫模型(DHMM)来构建语音识别系统,并在此基础上引入矢量量化(VQ)模块来保证识别率和识别速度。
This paper describes three kinds of phrases in a simple speech recognition model,the paper introduces the discrete hidden Markov model(DHMM)to build speech recognition systems,and on this basis,the introduction of vector quantization(VQ)module to ensure that the recognition rate and recognition speed,with good usability.
出处
《计算机光盘软件与应用》
2011年第14期144-144,共1页
Computer CD Software and Application
基金
绵阳师范学院科研项目(项目编号:MA2009012)部分研究成果.
关键词
矢量量化模块
语音识别
隐马尔柯夫模型
Vector quantization module
Speech recognition
Hidden Markov model