| 基于小波包变换和MFCC的说话人识别特征参数 |
| |
| 引用本文： | 陈静,张飞云.基于小波包变换和MFCC的说话人识别特征参数[J].电声技术,2009,33(2):49-51. |
| |
| 作者姓名： | 陈静 张飞云 |
| |
| 作者单位： | 许昌学院,电气信息工程学院,河南,许昌,461000 |
| |
| 基金项目： | 河南省教育厅自然科学研究项目，许昌学院校内科研基金 |
| |
| 摘 要： | 研究了小波包变换及MFCC参数的提取，结合二者特点，提取了基于小波包变换和MFCC的新参数DWT-MFCC．并在基于16阶GMM系统上进行说话人识别实验。实验结果表明，相对于传统的MFCC参数，在相同的噪声环境下，DWT—MFCC参数具有更高的说话人识别率。
|
| 关 键 词： | 小波包变换 MFCC参数 GMM系统 说话人识别 |
Feature Extraction Based on Wavelet Packet-MFCC in Speaker Recognition |
| |
| CHEN Jing,ZHANG Fei-yun.Feature Extraction Based on Wavelet Packet-MFCC in Speaker Recognition[J].Audio Engineering,2009,33(2):49-51. |
| |
| Authors: | CHEN Jing ZHANG Fei-yun |
| |
| Affiliation: | Electrical Information Department;Xuchang College;Xuchang Henan 461000;China |
| |
| Abstract: | The wavelet packet transform and the extraction of the MFCC(Mel-Frequency Cepstral Coefficients) parameters are studied.Based on these research,a new parameters,DWT-MFCC,is extracted.Based on the GMM system the speaker recognition is evaluated.The results show that the DWT-MFCC has higher recognition accuracy in noisy circum stances compared with the MFCC parameters. |
| |
| Keywords: | wavelet packet transform MFCC parameters GMM system speaker recognition |
| 本文献已被 CNKI 维普 万方数据 等数据库收录！ |
