计算机工程与应用 ›› 2007, Vol. 43 ›› Issue (22): 113-114.

• 网络、通信与安全 • 上一篇    下一篇

一种基于快速噪声估计的MMSE语音增强算法

陈照平1,马建芬2,张雪英1   

  1. 1.太原理工大学 信息工程学院,太原 030024
    2.太原理工大学 计算机与软件学院,太原 030024
  • 收稿日期:1900-01-01 修回日期:1900-01-01 出版日期:2007-08-01 发布日期:2007-08-01
  • 通讯作者: 陈照平

Minimum-Mean-Square-Error algorithm based on fast noise-estimation

CHEN Zhao-ping1,MA Jian-fen2,ZHANG Xue-ying1   

  1. 1.College of Information Engineering,Taiyuan University of Technology,Taiyuan 030024,China
    2.College of Computer and Software,Taiyuan University of Technology,Taiyuan 030024,China
  • Received:1900-01-01 Revised:1900-01-01 Online:2007-08-01 Published:2007-08-01
  • Contact: CHEN Zhao-ping

摘要: 提出一种基于快速噪声估计的MMSE语音增强算法,实验表明这种算法比起谱相减法和基于语音短时对数谱的最小均方误差(MMSE-LSA)算法能更显著地提高算法的客观性能,在非平稳噪声环境中能快速估计出变化的噪声功率谱。

关键词: 噪声估计, 最小均方误差, 语音增强

Abstract: A fast noise-estimation algorithm used in the speech enhancement algorithm of MMSE is studied.Experiments results prove that this algorithm can dramatically improve objective performance compared with spectrum subtraction algorithm and MMSE-LSA algorithm,and it can estimate the noise spectrum quickly in the non-stationary noise environment.

Key words: noise estimation, Minimum-Mean-Square-Error(MMSE), speech enhancement