计算机工程与应用 ›› 2011, Vol. 47 ›› Issue (29): 161-163.

• 数据库、信号与信息处理 • 上一篇    下一篇

一种听觉掩蔽效应和维纳滤波的语音增强算法

李 宁,蒋建中,郭军利   

  1. 解放军信息工程大学,郑州 450002
  • 收稿日期:1900-01-01 修回日期:1900-01-01 出版日期:2011-10-11 发布日期:2011-10-11

Speech enhancement algorithm based on auditory masking effect and Wiener filter

LI Ning,JIANG Jianzhong,GUO Junli   

  1. PLA Information Engineering University,Zhengzhou 450002,China
  • Received:1900-01-01 Revised:1900-01-01 Online:2011-10-11 Published:2011-10-11

摘要: 将非平稳噪声估计算法以及基于听觉掩蔽效应得到的噪声被掩蔽概率应用于维纳滤波语音增强中,提出了一种听觉掩蔽效应和维纳滤波的语音增强方法。几种噪声背景下对语音增强的客观测试表明,提出的算法相比较于传统的维纳滤波语音增强算法而言不但可以提高语音信噪比,而且可以明显减少语音失真。

关键词: 语音增强, 非平稳噪声估计, 听觉掩蔽效应, 被掩蔽概率, 维纳滤波

Abstract: The non-stationary noise estimation algorithm and the noise masking probability which is based on auditory masking are used in the Wiener filter speech enhancement algorithm.This paper proposes a method based on auditory masking effect and the Wiener filter speech enhancement algorithm.Objective measure demonstrates that the proposed method not only enhances the SNR but also reduces the voice distortion significantly.

Key words: speech enhancement, non-stationary noise estimation, auditory masking effect, masking probability, Wiener filter