摘要
针对基本谱减法的噪声残留以及损伤语音的可懂度和自然度等问题,在基本谱减法原理基础上,提出了基于参数自适应的改进谱减法。引进谱减功率系数和谱减噪声系数,以增强后语音信号的信噪比为输出,以变化的谱减功率系数和谱减噪声系数为输入,自动控制选择最优参数,并且将其与带噪语音输入信噪比相匹配。对信号进行语音增强时,通过计算其输入信噪比,利用最优参数曲线得到最优参数值,然后结合平滑处理的过程实现语音增强最大化。实验表明,本文算法有效提高了基本谱减法的语音增强性能。
Focusing on the problem of the residual noise and the damage of speech intelligibility and naturalness produced by the basic spectral subtraction,an improved spectral subtraction method based on parameter adaptive was proposed. Spectral subtraction noise correction factor and power spectrum correction factor were introduced into the basic spectral subtraction. Using the enhanced signal＇s SNR as the output,the two parameters as the input,two optimal parameters were chosen automatically. What＇s more,the selected optimal parameters were matched to the input SNR. When performing the enhancement processing,the two optimal parameters were got by calculating the input SNR according to the optimal parametric curve. Then,smoothing process ＇s combination maximized the speech enhancement. Experiments show that the proposed algorithm can effectively improve the performance of speech enhancement based on spectral subtraction.
出处
《科学技术与工程》
北大核心
2016年第3期192-196,共5页
Science Technology and Engineering
基金
国家自然科学基金青年基金(61305101)
河北省自然科学基金(F2014202121
F2010000137)资助
关键词
语音增强
谱减法
参数自适应
最优参数
speech enhancement
spectral subtraction
parameter self-adaption
optimal parameters
