Speech enhancement using a mixture-maximum model | IEEE Journals & Magazine | IEEE Xplore