Blind Source Separation when Speech Signals Outnumber Sensors using a Sparseness - Mixing Matrix Estimation (SMME)

This paper focuses on underdetermined blind source separation (BSS) of three speech signals mixed in a real environment from measurements provided by two sensors. Underdetermined BSS is a problem that has not yet been intensely studied and so far no satisfying solution has been obtained. The major issue encountered in previous work relates to the occurrence of distortion, which affects a separated signal with loud musical noise. To overcome this problem, we propose combining sparseness with the use of an estimated mixing matrix. First, we use a geometrical approach to perform a preliminary separation and to detect when only one source is active. This information is then used to estimate the mixing matrix, which allows us to improve our separation. Experimental results show that this Sparseness - Mixing Matrix Estimation (SMME) provides separated signals of better quality (less distortion, less musical noise) than those extracted without using the estimated mixing matrix.