资 源 简 介
语音增强算法的维纳滤波器,通过scalart 1996提出的实现。使用一个决策指导的方法跟踪的先验信噪比。请注意,该信号的第一个0.25sec用于模型的噪声,即,它是假定的演讲之后,0.25sec开始- Wiener Filter for Noisy Speech Enhancement, implemented as proposed by Scalart 1996. Uses a decision directed method to track the a priori SNR. Note that the first 0.25sec of the signal is used to model the noise, that is, it is assumed that speech starts after that 0.25sec