Automatic blind source separation of speech sources in an auditory scene | IEEE Conference Publication | IEEE Xplore