【作者】 张程； 潘翔；

【Author】 ZHANG Cheng;PAN Xiang;Department of Information Science and Electronic Engineering,Zhejiang University;

【机构】 浙江大学信息与电子工程学院；

【摘要】 在户外环境下,风噪声是语音信号的主要干扰源之一。风噪声具有非平稳性和不可预测性,因此,抑制风噪声具有挑战性。本文在分析语音信号的谐波特性的基础上,设计。相干调制梳状滤波器算法,用于抑制风噪声。结合基音频率估计、语音有声无声检测、风噪声检测、自动增益控制模块搭建了语音增强系统。利用不同信噪比的实验数据,对语音增强处理框架进行检验和评价。处理结果表明受污染的语音在可理解性上有所提高。更多还原

【Abstract】 In outdoor environments,wind noise is one of the main interference sources for speech.Due to the non-stationary and unpredictability of wind noise,noise suppression is a challenging task.On the base of analyzing the harmonic nature of speech signal,this paper presents design of coherent modulation comb filter(CMCF) algorithm for suppressing wind noise.A speech enhancement system is built on combination of CMCF,pitch frequency estimation,voiced/unvoiced detection,wind noise detection,and an automatic gain control module.The speech enhancement processing framework is tested and evaluated using the experimental data with different signal-to-noise ratios.The processing results have shown that there is an increase in the intelligibility of contaminated speech.更多还原