有色噪声下基于Unscented粒子滤波的语音增强方法

Speech enhancement based unscented particle filter with non-gaussian noises

针对含有色噪声的语音,提出了一种基于Unscented粒子滤波的单通道语音增强方法。采用时变自回归模型(TVAR)对干净语音建模,通过Unscented粒子滤波器估计AR模型的参数并滤除有色噪声。与大多数常用的粒子滤波选择的建议分布不同,Unscented粒子滤波器采用Unscented卡尔曼滤波器生成粒子滤波的建议分布。由于在粒子的更新过程中考虑了最近的观测值,Unscented粒子滤波器能够在粒子数少于传统粒子滤波算法所需粒子数目的基础上改善估计的性能。仿真实验结果表明,在有色噪声背景下该算法具有良好的语音增强效果。

Considering speech signals with color noises, a novel speech enhancement technique is proposed based on unscented particle filter （UPF）. The technique models speech signals with time-varying autoregressive （TVAR） models. Unscented particle filter is applied to estimate the parameters of AR model and filter color noises. Instead of most popular choice of proposal distribution, Unscented particle filter uses an Unscented Kalman filter （UKF） to generate the importance proposal distribution. It allows the particle filter to incorporate the latest observations into a prior updating routine so as to improve estimation performance greatly with fewer particles. Simulation results demonstrate that the proposed algorithm possesses good performance with color noises.