有色噪声分布式多通道语音增强算法研究

卡尔曼滤波是一种有效的语音增强技术,在以往,研究单通道语音增强的卡尔曼滤波算法比较多,多通道语音增强的卡尔曼滤波算法是比较少的。文章提出了一种有色噪声环境下分布式多通道时域语音增强的卡尔曼滤波算法。仿真结果表明,该算法优于传统的分布式多通道语音增强算法,实现了更高的降噪、更少的信号失真。

Kalman filter is an effective speech enhancement technology.In the past,there are many Kalman filtering algorithms for single-channel speech enhancement,but few Kalman filtering algorithms for multi-channel speech enhancement.In this paper,a Kalman filtering algorithm for distributed multi-channel time-domain speech enhancement in colored noise reduction environment is proposed.The simulation results show that this algorithm is better than the traditional distributed multi-channel speech enhancement algorithm,and achieves higher noise reduction and less signal distortion.