Blind extraction of a dominant source from mixtures of many sources using ICA and time-frequency masking

The paper presents a method for enhancing a target source of interest and suppressing other interference sources. The target source is assumed to be close to sensors, to have dominant power at these sensors, and to have non-Gaussianity. The enhancement is performed blindly, i.e., without knowing the total number of sources or information about each source, such as position and active time. We consider a general case where the number of sources is larger than the number of sensors. We employ a two-stage process where independent component analysis (ICA) is first employed in each frequency bin and time-frequency masking is then used to improve the performance further. We propose a new sophisticated method for selecting the target source frequency components, and also a new criterion for specifying time-frequency masks. Experimental results for simulated cocktail party situations in a room (reverberation time was 130 ms) are presented to show the effectiveness and characteristics of the proposed method.