In this work, we address the problem of estimating a speech signal corrupted by an additive noise when observations from two microphones are available. The extended spectral subtraction method (ESS), based on adaptive Wiener filtering is first introduced. The ESS method is implemented in the two channels and an adaptive system identification (ASI) based on the LMS algorithm is applied only on the reference channel. Finally we obtain an estimation of the speech signal by a simple combination of the two signals obtained from each branch.
展开▼
