首页 | 本学科首页   官方微博 | 高级检索  
     

基于基音频能值的端点检测算法
引用本文:相征,朗朗,王静.基于基音频能值的端点检测算法[J].安徽机电学院学报,2008(3):50-53.
作者姓名:相征  朗朗  王静
作者单位:安徽工程科技学院安徽省电气传动与控制重点实验室,安徽芜湖241000
摘    要:根据语音的发声特点,基于分析语音信号中的基音频率和语音之间的关系,提出一种在强噪声环境下实现提高语音识别率的方法,并对基于基音频能值的端点检测算法和传统语音端点检测算法进行比较.实验结果表明该方法不仅能够有效提高语音段的检测率,同时还具有计算量小、实时性高、不受噪声影响即鲁棒性好等特点.

关 键 词:基音  频能值  语音识别

An endpointing decteting algorithm of speech based on energy of frequency at pitch
XIANG Zheng,LANG Lang,WANG Jing.An endpointing decteting algorithm of speech based on energy of frequency at pitch[J].Journal of Anhui Institute of Mechanical and Electrical Engineering,2008(3):50-53.
Authors:XIANG Zheng  LANG Lang  WANG Jing
Affiliation:(Anhui Province Key Laboratory of Electric and Control, Anhui University of Technology and Science, Wuhu 241000, China)
Abstract:According to the audio features, this paper ananlyses the relations between pitch and speech. On this basis, a method of improving speech recognition accuracy in the presence of strong noise is proposed. And in comparison with traditional Endpointing Algorithms, this method not only effectively improves speech recognition accuracy,but also is of small amount of computing, better real-time performance,strong robustness and so on.
Keywords:pitch  frequency energy  speech recognition
本文献已被 维普 等数据库收录!
设为首页 | 免责声明 | 关于勤云 | 加入收藏

Copyright©北京勤云科技发展有限公司  京ICP备09084417号