| 基于VQ的声纹识别研究 |
| |
| 引用本文： | 张旺俏.基于VQ的声纹识别研究[J].中国科技信息,2007,28(7):124-125,127. |
| |
| 作者姓名： | 张旺俏 |
| |
| 作者单位： | 浙江工业大学,310014;浙江金华职业技术学院,321017 |
| |
| 摘 要： | 采用能够反映人对语音的感知特性的Mel频率倒谱系数（MFCC）作为语音的特征参数，研究了基于MFCC的VQ的识别方法，对单独使用MFCC与使用MFCC和AMFCC结合的识别率进行比较，实验结果表明通过对说话人的特征参数进行倒谱提升之后，MFCC和△MFCC结合能更好地区分不同说话人。
|
| 关 键 词： | △MFCC LBG算法 矢量量化 |
Study of Speaker Recognition System Based on VQ |
| |
| Zhang Wangqiao.Study of Speaker Recognition System Based on VQ[J].CHINA SCIENCE AND TECHNOLOGY INFORMATION,2007,28(7):124-125,127. |
| |
| Authors: | Zhang Wangqiao |
| |
| Abstract: | This paper adopts the Mel-Frequency Cepstrum Coefficients reflecting the characteristic of esthesia as the voice feature parameters, researches the recognition method based on Mel-Frequency Cepstrum Coefficient, compares the recognition rate between using the only MFCC and using MFCC and delta MFCC, Finally, the results of experiment indicate that the combination of MFCC and delta MFCC is better to distinguish different speakers, |
| |
| Keywords: | MFCC |
| 本文献已被 维普 万方数据 等数据库收录！ |
|
