| 基于GMM的实时说话人识别系统 |
| |
| 引用本文： | 胡益平,蔡骏,洪青阳.基于GMM的实时说话人识别系统[J].电声技术,2007,31(6):53-55. |
| |
| 作者姓名： | 胡益平 蔡骏 洪青阳 |
| |
| 作者单位： | 厦门大学,计算机科学系,福建,厦门,361005 |
| |
| 基金项目： | 厦门大学985工程二期信息技术创新平台项目 |
| |
| 摘 要： | 介绍了一个基于GMM实时说话人识别系统的设计与实现,系统具有实时说话人辨认和实时说话人确认功能。在实验室条件下,对不同的高斯混合密度个数及采样率进行了测试,测试了模型的自适应性能。实验表明系统具有较好的识别准确率。
|
| 关 键 词： | 说话人识别 实时系统 高斯混合模型 |
| 文章编号： | 1002-8684（2007）06-0053-03 |
| 修稿时间： | 2007-02-13 |
A Real Time Speaker Recognition System Based on GMM |
| |
| Hu Yi-ping,CAI Jun,HONG Qing-yang.A Real Time Speaker Recognition System Based on GMM[J].Audio Engineering,2007,31(6):53-55. |
| |
| Authors: | Hu Yi-ping CAI Jun HONG Qing-yang |
| |
| Affiliation: | Department of Computer Science, Xiamen University, Xiamen Fujian 361005, China |
| |
| Abstract: | The design and implementation of a real-time speaker recognition system which is based on GMM(Gaussian Mixture Model) are presented. The system has the characteristics of real time speaker identification and real time speaker verification. In the lab environment, the performance of the system, as well as the model adaptation, has been fully tested with GMMs of different numbers of Gaussian mixtures and different sampling rates. The testing results show that the GMM-based system has a satisfactory correctness in performing speaker recognition. |
| |
| Keywords: | speaker recognition real-time systems GMM |
| 本文献已被 CNKI 维普 万方数据 等数据库收录！ |
|
