Speech recognition using voice-characteristic-dependent acoustic models | IEEE Conference Publication | IEEE Xplore