PhonemeViz places the character(s) of the most recently spoken initial phoneme just to the side of a speaker’s lips. This design should enable a speechreader to focus on the speaker’s eyes and lip movements (as in traditional speechreading), while also monitoring changes in PhonemeViz’s state using their peripheral vision to help disambiguate confusing visemes. PhonemeViz PhonemeViz can be overlaid onto video or displayed on a transparent head mounted display to augment natural speechreading and enhance speechreading acquisition. In a pilot evaluation of a PhonemeViz prototype it enabled all participants to achieve 100% word recognition (showing successful viseme disambiguation)

Associated Publications


  • Dr David Flalta, University of Guelph