A method of joint compensation of additive and convolutive distortions for speaker-independent speech recognition | IEEE Journals & Magazine | IEEE Xplore