| 基于主成分分析的汉语连续语音切分算法 |
| |
| 引用本文： | 张永锋,杨影,肖莹莹.基于主成分分析的汉语连续语音切分算法[J].应用声学,2011,30(5):366-369. |
| |
| 作者姓名： | 张永锋 杨影 肖莹莹 |
| |
| 作者单位： | 大连东软信息学院嵌入式系统工程系; |
| |
| 摘 要： | 连续语音切分是汉语连续语音识别中的一项重要技术。本文将通常用于图像处理的主成分分析(PCA)方法引入汉语连续语音切分领域,目的是通过PCA,降低语音帧之间的频谱相关性。再利用相邻语音帧特征构成的三维子空间的大小,实现连续语音切分,取得了不错的切分效果。
|
| 关 键 词： | 主成分分析 连续语音切分 |
A Chinese continuous speech segmentation algorithm based on principle component analyses |
| |
| ZHANG Yongfeng,YANG Ying and XIAO Yingying.A Chinese continuous speech segmentation algorithm based on principle component analyses[J].Applied Acoustics,2011,30(5):366-369. |
| |
| Authors: | ZHANG Yongfeng YANG Ying and XIAO Yingying |
| |
| Institution: | ZHANG Yongfeng YANG Ying XIAO Yingying(Dept.of Embedded System Engineering,Dalian Neusoft Institute of Information,Dalian 116023) |
| |
| Abstract: | Continuous speech segmentation is an important technology of Chinese continuous speech recognition.In this paper,the principal component analysis(PCA),which is commonly used for image processing,is introduced to the field of Chinese continuous speech segmentation to reduce the correlation among speech frames.Next,according to the size of three-dimensional sub-space which comes from adjacent speech frames,good speech segmentations are achieved. |
| |
| Keywords: | PCA Continuous speech segmentation |
| 本文献已被 CNKI 等数据库收录！ |
| 点击此处可从《应用声学》浏览原始摘要信息 |
| 点击此处可从《应用声学》下载免费的PDF全文 |
|
