基于高斯混合模型的汉语方言辨识系统

计算机工程与应用 ›› 2007, Vol. 43 ›› Issue (3): 204-204.

基于高斯混合模型的汉语方言辨识系统

顾明亮马勇

徐州师范大学物理系徐州师范大学物理系

收稿日期:2006-02-07 修回日期:1900-01-01 出版日期:2007-01-21 发布日期:2007-01-21
通讯作者: 顾明亮

GMM-based Chinese Dialect identification System

Gu Mingliang1，2 Ma Yong1

Received:2006-02-07 Revised:1900-01-01 Online:2007-01-21 Published:2007-01-21

摘要/Abstract

摘要： 本文建立了一个基于高斯混合模型的汉语方言辨识系统，并给出了模型参数的估计方法，论文讨论了特征参数和高斯混合数对系统辨识的影响，实验结果表明，系统对同一省内的三种不同方言的辨识率平均可以达到84.17%。

关键词: 高斯混合模型、汉语方言辨识、语音识别

Abstract: This paper sets up a GMM-based Chinese dialect identification system and presents the approach for estimation of model parameters. The paper also discusses the influence of the feature and the number of Gaussian mixture component. The experimental results showed that the system can arrived at average identification rate of 84.17% on three different dialects in the same province.

Key words: GMM, Chinese Dialect identification, speech recognition

顾明亮马勇. 基于高斯混合模型的汉语方言辨识系统[J]. 计算机工程与应用, 2007, 43(3): 204-204.

Gu Mingliang1，2 Ma Yong1 . GMM-based Chinese Dialect identification System[J]. Computer Engineering and Applications, 2007, 43(3): 204-204.