| 以DSP实现基于概率DP匹配算法的说话人识别系统 |
| |
| 引用本文： | 周洁,赵力,邹采荣.以DSP实现基于概率DP匹配算法的说话人识别系统[J].实验室研究与探索,2005,24(11):12-14. |
| |
| 作者姓名： | 周洁 赵力 邹采荣 |
| |
| 作者单位： | 1. 云南交通职业技术学院,基础部,云南,昆明,650101;东南大学,无线电工程系,江苏,南京,210096
2. 东南大学,无线电工程系,江苏,南京,210096 |
| |
| 基金项目： | 国家自然科学基金（编号：60472058）;教育部科学技术重点项目（编号：03082） |
| |
| 摘 要： | 提出了利用概率DP匹配算法进行说话人识别的的设想。并给出了运用TMS320C5416实现说话人自动识别系统的方案。该系统利用一种新的语音信号r阶的倒谱线性回归系数等参数构成识别的特征矢量集，运用提出的概率DP匹配算法进行与文本无关的说话人识别。实验结果表明该系统具有识别精度高、识别速度快，占用系统资源少等特点，是一种有效的说话人自动识别的实现方法。
|
| 关 键 词： | 说话人识别 概率DP匹配算法 数字信号处理 |
| 文章编号： | 1006-7167（2005）11-0012-03 |
| 收稿时间： | 2004-12-23 |
| 修稿时间： | 2004年12月23 |
Realization of Speaker Recognition System Based on Probability DP Matching Algorithm with DSP |
| |
| ZHOU Jie,ZHAO Li,ZOU Cai-rong.Realization of Speaker Recognition System Based on Probability DP Matching Algorithm with DSP[J].Laboratory Research and Exploration,2005,24(11):12-14. |
| |
| Authors: | ZHOU Jie ZHAO Li ZOU Cai-rong |
| |
| Institution: | 1.Yunnan Traffic Professional Technology Academy, Kunming 650101,China; 2 Dept. of Radio Eng., Southeast Univ., Nanjing 210096 |
| |
| Abstract: | A probability DP matching algorithm was proposed to realize speech speaker recognition and TMS320C5416 was used to realize auto speaker recognition system. In this system, the r-order cepstral linear regeression coefficients of audio signal is used as part of the feature parameter set of recognition and the text-independent speaker recognition is carried out using probability DP matching algorithm. The experimental result shows that this system has the feature as high recognition precision, fast recognition rate and using fewer system resourcers. The system is effective method to realize auto speaker recognition. |
| |
| Keywords: | speaker recognition probability DP matching algorithm digital signal processing(DSP) |
| 本文献已被 CNKI 维普 万方数据 等数据库收录！ |
