| 基于GMM-UBM的语言辨识算法研究 |
| |
| 引用本文： | 陈业仙,张歆奕,毛杰.基于GMM-UBM的语言辨识算法研究[J].五邑大学学报(自然科学版),2010,24(3):56-60. |
| |
| 作者姓名： | 陈业仙 张歆奕 毛杰 |
| |
| 作者单位： | 五邑大学,信息工程学院,广东,江门,529020 |
| |
| 摘 要： | 运用Matlab软件,以自己建立的语音数据库为基础,对与文本无关的基于GMM-UBM的语言辨识系统进行了测试,获得的平均识别率达74%,与传统GMM算法的测试对比,基于GMM-UBM的语言辨识算法能更好地改善语言辨识系统的性能.
|
| 关 键 词： | 语言辨识 高斯混合-全局背景模型 期望最大化 贝叶斯自适应算法 |
A Study of a Language Identification Algorithm Based on the GMM-UBM Model |
| |
| CHEN Ye-xian,ZHANG Xin-yi,MAO Jie.A Study of a Language Identification Algorithm Based on the GMM-UBM Model[J].Journal of Wuyi University(Natural Science Edition),2010,24(3):56-60. |
| |
| Authors: | CHEN Ye-xian ZHANG Xin-yi MAO Jie |
| |
| Institution: | (School of Information Engineering,Wuyi University,Jiangmen 529020,China) |
| |
| Abstract: | Language identification technology is a very important part of the speech recognition technology.In this paper,based on the practical application and a self-established voice database,a language identification system based on the GMM-UBM model and independent of the speaker is studied and compared with the traditional GMM methods.Experiment results show that this algorithm can effectively improve the performance of the language identification system and achieve an average recognition rate of 74%. |
| |
| Keywords: | language identification GMM-UBM EM Bayesian adaptive algorithm |
| 本文献已被 维普 万方数据 等数据库收录！ |
|
