首页 | 本学科首页   官方微博 | 高级检索  
     

基于短时对数谱估计MMSE的语音增强算法研究
引用本文:袁伟军,刘珩.基于短时对数谱估计MMSE的语音增强算法研究[J].电声技术,2007,31(10):59-62,65.
作者姓名:袁伟军  刘珩
作者单位:中国农业大学,信息与电气工程学院,北京,100083;中国农业大学,信息与电气工程学院,北京,100083
摘    要:结合语音激活检测(VAD)技术对短时对数谱估计最小均方误差(LSA-MMSE)语音增强算法进行了改进。通过实验表明,LSA-MMSE增强算法在消除背景噪声、增加语音清晰度和提高语音自然度等方面比谱减法更加有效。

关 键 词:语音增强  短时谱估计  最小均方误差
文章编号:1002-8684(2007)10-0059-04
修稿时间:2007-06-06

Study of Speech Enhancement Based on MMSE Short Time Log-spectral Amplitude Estimator
YUAN Wei-jun,LIU Heng.Study of Speech Enhancement Based on MMSE Short Time Log-spectral Amplitude Estimator[J].Audio Engineering,2007,31(10):59-62,65.
Authors:YUAN Wei-jun  LIU Heng
Affiliation:College of Information and Engineering, China Agricultural University, Beijing 100083, China
Abstract:To improve the performance of communication system and clear the noise especially the background noise in real environment in noisy speech,a method of LSA-MMSE(Log-Spectral Amplitude-Minimum Mean Square Error)combined with advanced VAD speech enhancement is proposed.Compared with the spectral subtraction,results show that the method of LSA-MMSE combined with VAD speech enhancement is more effective than the spectral subtraction in terms of noise suppression and speech signal distortion.
Keywords:speech enhancement  short-time spectrum analysis  MMSE
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