Speaker identification in a multimodal interface | IEEE Conference Publication | IEEE Xplore