This paper presents a new blind speech separation algorithm using beamforming technique that is capable of extracting each individual speech signal from a mixture of three speech sources in a room. | Source separation employing beamforming and SRP-PHAT localization in three-speaker room environments