Improvement of the histogram in the degenerate unmixing estimation technique algorithm

Abstract

A method of improving the histogram in the degenerate unmixing estimation technique (DUET) algorithm is proposed. The DUET algorithm is one of the methods of blind signal separation (BSS). The BSS framework is to retrieve source signals from mixtures of them without a priori information about the source signals and the mixing process. In the DUET algorithm, the histogram of both the direction-of-arrivals (DOAs) and the distances is formed from the mixtures which are observed using two sensors. And then, signal separation is achieved using time-frequency masking based on the histogram. Consequently, the capability for the DUET algorithm strongly depends on the performance of the histogram. In general, the histogram is degraded by the reverberation or the reflection of the source signals when the DUET algorithm is applied in the real environment. Our approach is to improve the histogram of the DUET algorithm. In our method, the phase component of the observed mixture at each frequency bin is modified by those at the neighboring frequency bins. The proposed method gives the sharper histogram in comparison with the conventional approach.