首页 | 本学科首页   官方微博 | 高级检索  
     检索      

基于小波变换的汉语三字词语音基频提取方法
引用本文:钟金宏,杨善林,魏方传.基于小波变换的汉语三字词语音基频提取方法[J].合肥工业大学学报(自然科学版),2002,25(3):331-334.
作者姓名:钟金宏  杨善林  魏方传
作者单位:合肥工业大学,管理学院,安徽,合肥,230009
摘    要:小波变换具有良好的时频局部分析能力 ,可利用它检测语音信号的基频。文章将其应用于具有实际连续语音特征的三字词语音的基频提取 ,发现对大多数三字词提取的基频中含有较多的错误点 ,甚至出现一堆乱点。文章分析了其产生的原因 ,给出了基于投票策略的改进方案 ,并验证了该方案的有效性。

关 键 词:三字词  小波变换  基频检测  投票策略
文章编号:1003-5060(2002)03-0331-04
修稿时间:2001年5月26日

Research on the pitch detection method of Chinese trisyllabic words using wavelet transform
ZHONG Jin-hong,YANG Shan-lin,WEI Fang-chuan.Research on the pitch detection method of Chinese trisyllabic words using wavelet transform[J].Journal of Hefei University of Technology(Natural Science),2002,25(3):331-334.
Authors:ZHONG Jin-hong  YANG Shan-lin  WEI Fang-chuan
Abstract:The wavelet transform is a perfect time-frequency local analysis tool, by which the pitch of speech signals can be detected. In this paper, the wavelet transform is applied to the pitch detection of trisyllabic words with the characteristics of natural continuous speeches, and a lot of error points are found among the pitch points detected in most of trisyllabic words. The causes of this phenomenon are analyzed, an improved algorithm is proposed based on the voting strategy, and its validity is proved by experiments.
Keywords:Chinese trisyllabic word  wavelet transform  pitch detection  voting strategy
本文献已被 CNKI 万方数据 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