Visual Speech Recognition using Model Data and Spatial Frequency Information