| 基于改进噪声方差估计的小波阈值语音增强 |
| |
| 引用本文： | 李玉,于凤芹,范树凯.基于改进噪声方差估计的小波阈值语音增强[J].电声技术,2008,32(3):56-58. |
| |
| 作者姓名： | 李玉 于凤芹 范树凯 |
| |
| 作者单位： | 江南大学,通信与控制工程学院,江苏,无锡,214122 |
| |
| 摘 要： | 提出了一种估计语音噪声方差的改进方法,利用语音信号小波系数尺度间的相关性,去除了语音信号细节小波系数对噪声方差估计的干扰。仿真结果表明,方法能准确估计语音信号的噪声方差,尤其当语音噪声较小时,方法比传统的Donoho法更准确。
|
| 关 键 词： | 小波变换 噪声估计 语音增强 |
| 文章编号： | 1002-8684(2008)03-0056-03 |
| 修稿时间： | 2007年12月25 |
Speech Enhancement Based on Improved Noise Variance Estimation in Wavelet Domain |
| |
| LI Yu,YU Feng-qin,FAN Shu-kai.Speech Enhancement Based on Improved Noise Variance Estimation in Wavelet Domain[J].Audio Engineering,2008,32(3):56-58. |
| |
| Authors: | LI Yu YU Feng-qin FAN Shu-kai |
| |
| Abstract: | An improved method is proposed to estimate the noise variance of noise, which eliminates the detail coefficients of the speech using correlation of the speech wavelet coefficients between scales. The simulation shows that the proposed method can provide better results than the traditional Donoho method, especially when the noise variance is small. |
| |
| Keywords: | wavelet transform noise estimation speech enhancement |
| 本文献已被 CNKI 维普 万方数据 等数据库收录！ |
|
