| 一种基于迭代多带谱相减的语音增强方法 |
| |
| 引用本文： | 陈奎成,童莹. 一种基于迭代多带谱相减的语音增强方法[J]. 西安邮电学院学报, 2005, 10(1): 15-18 |
| |
| 作者姓名： | 陈奎成 童莹 |
| |
| 作者单位： | 1. 南京邮电学院,通信工程系,江苏,南京,210003
2. 南京邮电学院,信息工程系,江苏,南京,210003 |
| |
| 摘 要： | 根据传统的谱相减语音增强算法,提出了一种抑制噪声的迭代多带谱相减语音增强方法,暨将迭代方法和多带谱相减方法相结合。同时该方法采取了一种不需要静音检测(VAD)的自适应方法来估计噪声。仿真结果表明, 该方法在较好的抑制背景噪声和残留音乐噪声的同时,保持较好的语音可懂度。
|
| 关 键 词： | 语音增强 自适应噪声估计 多带谱相减 |
| 文章编号： | 1007-3264（2005）01-0015-04 |
| 收稿时间： | 2004-05-13 |
| 修稿时间： | 2004-05-13 |
A speech enhancement method based on iterative multi-band spectral subtraction algorithm |
| |
| Chen KuiCheng;Tong Ying. A speech enhancement method based on iterative multi-band spectral subtraction algorithm[J]. Journal of Xi'an Institute of Posts and Telecommunications, 2005, 10(1): 15-18 |
| |
| Authors: | Chen KuiCheng Tong Ying |
| |
| Abstract: | According to the traditional spectral subtraction speech enhancement algorithm,a speech enhancement method based on iterative multi - band spectral subtraction algorithm has been put forward to suppress noise. It combines iterative spectral subtraction with multi - band spectral subtraction algorithm. At the same time, this method adopts adaptive noise estimation which does not need voice activity detector (VAD). The simulation results shows its excellent performance in background noise and music noise suppression while keeping very low audible distortion. |
| |
| Keywords: | speech enhancement adaptive noise estimation multi- band spectral subtraction |
| 本文献已被 CNKI 维普 万方数据 等数据库收录！ |
|
