| 基于GMM说话人分类的说话人识别方法研究 |
| |
| 引用本文： | 赵振东,张静,李圆,胡喜梅. 基于GMM说话人分类的说话人识别方法研究[J]. 通信技术, 2009, 42(10): 192-193 |
| |
| 作者姓名： | 赵振东 张静 李圆 胡喜梅 |
| |
| 作者单位： | 华北电力大学电子与通信工程系,河北,保定,071003 |
| |
| 摘 要： | 提出了基于高斯混合模型（GMM）说话人分类的分级说话人识别系统，同时将小波神经网络（WNN）引入到子识别系统中。分别对未分级说话人识别系统和分级说话人识别系统进行了比较。仿真实验结果表明，分级网络在保证正确识别率的同时，不仅改善了网络训练速度，亦大大提高了识别响应速度。
|
| 关 键 词： | 说话人识别 小波神经网络 高斯混合模型 |
Speaker Recognition Research Based on GMM Speaker Clustering Technology |
| |
| ZHAO Zhen-dong,ZHANG Jing,LI Yuan,HU Xi-mei. Speaker Recognition Research Based on GMM Speaker Clustering Technology[J]. Communications Technology, 2009, 42(10): 192-193 |
| |
| Authors: | ZHAO Zhen-dong ZHANG Jing LI Yuan HU Xi-mei |
| |
| Affiliation: | Department of Electronic and Communication Engineering, North China Electric Power University, Baoding Hebei 071003, China) |
| |
| Abstract: | Hierarchical speaker recognition system based on Gauss Mixed Model （GMM） speaker clustering technology is proposed, and the model of wavelet neural network （WNN） is introduced into the sub-recognition system. The hierarchical speaker recognition system is compared with the non-hierarchical system. The simulation results show that the hierarchical speaker recognition system is improved in network training speed and recognition response speed. |
| |
| Keywords: | speaker recognition wavelet neural network： Gauss mixed model |
| 本文献已被 维普 万方数据 等数据库收录！ |
