| 语音识别中DTW改进算法的研究 |
| |
| 引用本文： | 胡金平,陈若珠,李战明. 语音识别中DTW改进算法的研究[J]. 微型机与应用, 2011, 30(3): 30-32 |
| |
| 作者姓名： | 胡金平 陈若珠 李战明 |
| |
| 作者单位： | 兰州理工大学电气工程与信息工程学院,甘肃兰州,730050 |
| |
| 摘 要： | 动态时间规整DTW是语音识别中的一种经典算法。对此算法提出了一种改进的端点检测算法,特征提取采用了Mel频率倒谱系数MFCC,并采用计算量相对较小的改进的动态时间规整算法实现语音参数模板匹配,能够实现孤立词、特定人、小词汇量的语音识别,并用Matlab进行了算法仿真。试验结果表明,改进后的算法能够有效地提高系统对语音的识别率。
|
| 关 键 词： | 语音识别 端点检测 Mel倒谱参数 动态时间规整 |
Discussion of improved DTW algorithm in speech recognition |
| |
| Hu Jinping,Chen Ruozhu,Li Zhanming. Discussion of improved DTW algorithm in speech recognition[J]. Microcomputer & its Applications, 2011, 30(3): 30-32 |
| |
| Authors: | Hu Jinping Chen Ruozhu Li Zhanming |
| |
| Affiliation: | Hu Jinping,Chen Ruozhu,Li Zhanming (College of Electric and Information Engineering,Lanzhou University of Science and Technology,Lanzhou 730050,China) |
| |
| Abstract: | Dynamic time warping is a kind of classical programming in speech recognition.It adopts the improved endpoint detection algorithm and Mel frequence cepstrum coeffiients to catch speech characteristic parameters and introduces dynamic time wrapping arithmetic to realize speech pattern matching.It is proved that this article designs a small vocabulary,isolated word speech recognition system,arithmetic of speech recognize simulate with Matlab software,the results show that the modified algorithm can provide a ... |
| |
| Keywords: | speech recognition endpoint detection MFCC DTW |
| 本文献已被 CNKI 万方数据 等数据库收录！ |
|
