MMSE

代码说明：

  In this letter, we derive a minimum mean squared error (MMSE) optimal estimator for clean speech spectral am-plitudes, which we apply in single channel speech enhancement. As opposed to state-of-the-art estimators, theoptimal estimator is derived for a given clean speech spectral phase. We show that the phase contains additional information that can be exploited to distinguish outliers in the noise from the target signal.