| 基于三音子模型连续语音声调识别方法 |
| |
| 引用本文： | 魏瑞莹,粱维谦.基于三音子模型连续语音声调识别方法[J].电声技术,2011,35(8):34-37. |
| |
| 作者姓名： | 魏瑞莹 粱维谦 |
| |
| 作者单位： | 1. 清华大学 微纳电子学系,北京,100084
2. 清华大学 电子工程系,北京,100084 |
| |
| 摘 要： | 作为汉语语音识别的重要组成部分,声调识别具有关键的作用.提出了一种新的基于前后文相关的模型识别方法用以提高汉语连续语音中的识别率.首先介绍用于声调识别的基因轨迹的提取和处理,然后提出6种特征来描述基因轨迹的变化趋势并给出具体的计算公式,利用这些特征并考虑连续语音中前后音节的相关性对基因轨迹造成的变化而建立细分的声调模型...
|
| 关 键 词： | 声调识别 基因轨迹 特征提取 三音子模型 决策树 |
Multi-feature Based Tri-phone Model for Tone Recognition of Chinese Continuous Speech |
| |
| WEI Ruiying,LIANG Weiqian.Multi-feature Based Tri-phone Model for Tone Recognition of Chinese Continuous Speech[J].Audio Engineering,2011,35(8):34-37. |
| |
| Authors: | WEI Ruiying LIANG Weiqian |
| |
| Affiliation: | WEI Ruiyinga,LIANG Weiqianb(a.Department of Microelectronic,b.Department of Electronic Engineering,Tsinghua University,Beijing 100084,China) |
| |
| Abstract: | As an important part of the recognition of Chinese speech,the tone recognition is also a gut issue.In this paper,a new kind of context-dependent tone model is proposed.Firstly,the Fundamental Frequency(F0) extraction is introduced,and then six kinds of features which contain important information about the discrimination between each type of the tone are proposed.Based on these features and the consideration of the syllable correlations in continuous speech,tri-phone tone models are built.Finally,the decisi... |
| |
| Keywords: | tone recognition fundamental frequency feature extraction tri-phone model decision tree |
| 本文献已被 CNKI 维普 万方数据 等数据库收录！ |
|
