改进的多窗谱谱减法语音增强研究  

speech enhancement based on improvement multi-window spectral subtraction

作　　者：姚远[1] 李亚伟 白天皓 Yao Yuan;Li Yawei;Bai Tianhao(College of Physical Science and Technology,Central China Normal University,Wuhan Hubei,430079)

机构地区：[1]华中师范大学物理科学与技术学院

出　　处：《电子测试》2019年第14期40-43,共4页Electronic Test

摘　　要：背景噪声是通信系统的干扰噪声来源之一,相关研究者越来越重视该问题。语音增强能够降低噪声影响,从而提高语音可懂度[1]。在多窗谱谱减法的基础之上,本文将帧间平滑以及先验信噪比估计结合起来,提出了一种改进多窗谱的语音增强算法。仿真结果显示本文所提出的算法既提高了语音信噪比,又改善了语音可懂度。Background noise is one of the sources of interference noise in communication systems,and researchers are paying more and more attention to this problem. Speech enhancement reduces noise effects and improves speech intelligibility [1]. Based on the multi-window spectral subtraction method, this paper combines inter-frame smoothing and a priori SNR estimation, and proposes a speech enhancement algorithm that improves multi-window spectrum. The simulation results show that the proposed algorithm not only improves the speech signal to noise ratio, but also improves the speech intelligibility.

关 键 词：语音增强 多窗谱谱减法 帧间平滑 先验SNR估计 

分 类 号：TN9[电子电信—信息与通信工程]