Direction of arrival estimation using matching pursuit and its application to source separation for convolved mixtures

In this paper, we describe a new source separation method in which uses spatial information derived from the direction of arrival (DOA) estimates of each direct and reflected sound issued. The method we propose has the following steps: (1) each DOA is estimated using matching pursuit and reoptimized after each new DOA is estimated, (2) using these DOA estimates, the mixing matrix is also estimated and the inverse of the mixing matrix is used to separate the mixture signals. In our experiments, we obtained a better signal separation with the new method than with the conventional frequency-domain independent component analysis (ICA)-based source separation method.