Continuous visual speech recognition for multimodal fusion | IEEE Conference Publication | IEEE Xplore