| 大学生闽南语说话人识别语音库的设计 |
| |
| 引用本文： | 王静芸. 大学生闽南语说话人识别语音库的设计[J]. 厦门理工学院学报, 2009, 17(3): 79-83 |
| |
| 作者姓名： | 王静芸 |
| |
| 作者单位： | 厦门理工学院数理系,福建,厦门,361024 |
| |
| 基金项目： | 厦门市科技计划指导性项目 |
| |
| 摘 要： | 设计了一个包含120名在校大学生的闽南语说话人识别语音库．该语音库可以根据闽南语口音、说话人性别等条件来进行语音语料的查询．针对不同口音的说话人进行了一系列的GMM（Gaussian Mixture Speaker Model）说话识别实验，验证了特定口音模型比非特定口音模型的识别率一般要高10％左右．
|
| 关 键 词： | 闽南语语音库 说话人识别 GMM |
Design of Hokkien Speech Corpus Based on College Students for Speaker Recognition |
| |
| WANG Jing-yun. Design of Hokkien Speech Corpus Based on College Students for Speaker Recognition[J]. Journal of Xiamen University of Technology, 2009, 17(3): 79-83 |
| |
| Authors: | WANG Jing-yun |
| |
| Affiliation: | WANG Jing-yun ( Department of Mathematics and Physics, Xiamen University of Technology, Xiamen 361024, China) |
| |
| Abstract: | The paper designs a Hokkien speech corpus which contains 120 speakers from college Students. This corpus is convenient for us to inquire about voice materials by gender and accent. Experiments for different accents with speaker recognition using GMM show that the recognition rate of specific accent model is about 10% higher than non-specific accent model. |
| |
| Keywords: | GMM |
| 本文献已被 维普 万方数据 等数据库收录！ |
|
