首页 | 官方网站   微博 | 高级检索  
     

基于DCT增强和改进谱熵的语音端点检测方法
引用本文:董胡,钱盛友,刘洋,谭乔来.基于DCT增强和改进谱熵的语音端点检测方法[J].通信技术,2007,40(11):364-365,368.
作者姓名:董胡  钱盛友  刘洋  谭乔来
作者单位:湖南师范大学,物理与信息科学学院,湖南,长沙,410081
摘    要:端点检测是语音信号处理过程中非常重要的一步,它的准确性直接影响到后面的学习及识别效果.提出了基于DCT(离散余弦变换)增强和改进谱熵的语音端点检测方法.首先,通过DCT(离散余弦变换)进行语音增强,然后再通过改进谱熵法判决去噪后语音的端点位置.仿真结果表明:此方法快速有效,具有较强的抗噪能力,特别适合低信噪比下的端点检测.

关 键 词:离散余弦变换  阈值  去噪  改进谱熵  端点检测
文章编号:1002-0802(2007)11-0364-02
收稿时间:2007-06-01
修稿时间:2007年6月1日

Speech Endpoint Detection Method Based on DCT Enhancement and Improved Spectral Entropy
DONG Hu,QIAN Sheng-you,Liu Yang,Tan Qiao-lai.Speech Endpoint Detection Method Based on DCT Enhancement and Improved Spectral Entropy[J].Communications Technology,2007,40(11):364-365,368.
Authors:DONG Hu  QIAN Sheng-you  Liu Yang  Tan Qiao-lai
Abstract:The endpoint detection is an important step in speech signal processing,which directly influences the afterward speech recognition.In the paper,a speech endpoint detection method based on DCT(Discrete Cosine Transform)enhancement and improved spectral entropy is proposed.The speech is first enhanced according to DCT(Discrete Cosine Transform),then endpoint location of denoising speech is detected by improved spectral entropy.The simulation results indicate that the proposed method can detect the endpoint exactly in the low SNR environments.
Keywords:discrete cosine transform threshold  denoising  improved spectral entropy  endpoint detection
本文献已被 CNKI 维普 万方数据 等数据库收录!
设为首页 | 免责声明 | 关于勤云 | 加入收藏

Copyright©北京勤云科技发展有限公司    京ICP备09084417号-23

京公网安备 11010802026262号