Two extensions to ensemble speaker and speaking environment modeling for robust automatic speech recognition | IEEE Conference Publication | IEEE Xplore