摘要
语音增强技术是语音信号处理中的重要课题之一,谱减法是目前语音增强处理中常用的方法。针对传统谱减法残余音乐噪声过强,清音部分损失严重的缺点,引入了一种利用语音信号的短时平均幅度差特征并结合短时平均幅度的语音检测算法,对传统的谱减法进行了改进。根据采集的真实航空噪声数据,将传统谱减法和改进的谱减法结果进行了比较分析。仿真结果表明,改进形式的谱减法可以有效降低音乐噪声,提高信噪比和可懂度,可以满足语音增强的要求。
The speech enhancement technique is an important topic in speech signal processing. At present, spectral subtraction method is an useful algorithm for speech enhancement. Aiming at reducing musical noise and the loss of unvoiced speech in the spectral subtraction method, a reliable voice active detection algorithm which utilizes the characteristics of short - time AMDF （ Average Magnitude Difference Function ） and short - time AM （ Average Magnitude） to detect the speech endpoint is introduced. And the spectral subtraction is improved. According to collecting real data of aviation noise, the outcomes of classical spectral subtraction and modified algorithm are compared. The result shows that the modified algorithm can eliminate musical noise effectively and improve the intelligibility and SNR of speech signal, and can meet the demand of speech enhancement.
出处
《计算机仿真》
CSCD
北大核心
2009年第6期363-366,共4页
Computer Simulation
基金
国家自然科学基金(60672139)
关键词
谱减法
语音增强
端点检测
可懂度
Spectral subtraction method
Speech enhancement
Endpoint detection
Intelligibility
