Employing Laplacian-Gaussian densities for speech enhancement

A new efficient speech enhancement algorithm (SEA) is developed. Noisy speech is first decorrelated and then the clean speech components are estimated from the decorrelated noisy speech samples. The distributions of clean speech and noise are assumed to be Laplacian and Gaussian, respectively. The clean speech components are estimated either by maximum likelihood (ML) or minimum-mean-square-error (MMSE) estimators. These estimators require some statistical parameters that are adaptively extracted by the ML approach during the active speech or silence intervals, respectively. A voice activity detector (VAD) is employed to detect whether the speech is active or not. The simulation results show that this SEA performs as well as a recent high efficiency SEA that employs the Wiener filter. The complexity of this algorithm is very low compared with existing SEAs.