计算机工程与应用 ›› 2007, Vol. 43 ›› Issue (27): 140-141.

• 网络、通信与安全 • 上一篇    下一篇

基于听觉掩蔽效应的MMSE语音增强算法

阔永红,陈 健,杨昌方   

  1. 西安电子科技大学 通信工程学院,西安 710071
  • 收稿日期:1900-01-01 修回日期:1900-01-01 出版日期:2007-09-21 发布日期:2007-09-21
  • 通讯作者: 阔永红

MMSE speech enhancement algorithm based on masking properties

KUO Yong-hong,CHEN Jian,YANG Chang-fang   

  1. School of Telecommunications Engineering,Xidian University,Xi’an 710071,China
  • Received:1900-01-01 Revised:1900-01-01 Online:2007-09-21 Published:2007-09-21
  • Contact: KUO Yong-hong

摘要: 针对MMSE语音增强算法低信噪比时产生较大的语音畸变的缺点,提出了一种结合人耳听觉掩蔽效应的MMSE语音增强算法。该算法利用掩蔽阈值来调整MMSE算法中的增益值,使得增强后的语音信号残留噪声和语音畸变较小。通过计算机仿真对增强前后语音信号的信噪比分析以及主观试听表明:改进的MMSE语音增强算法不仅提高了语音信号的信噪比,而且减少了语音畸变,提高了语音的可懂度。

关键词: 语音增强, MMSE, 听觉掩蔽, 噪声残差消除

Abstract: This paper focuses on the serious problems of the residual noise and the speech aberration brought by the MMSE method for speech enhancement when SNR is low.The proposed algorithm incorporates the masking properties of the ears to adjust the gain for MMSE.The experiment result by objective measure shows that the SNR is improved after speech enhancement.It is shown by informal listening test that the residual noise is reduced significantly with the increasing of the speech intelligibility and the decreasing of the speech aberration.

Key words: speech enhancement, Minimum Mean Square Error(MMSE), masking properties, residual noise removal