Visual Speech Recognition

This lecture overviews Visual Speech Recognition that has many applications in Human-centered Computing, Image and Video Analysis and Social Media Analytics. It covers the following topics in detail: Visual Speech Recognition: Visemes and Phonemes, Face detection, Landmark Localization, Lip reading, Speech reading beyond the lips. Audio-Visual Speech Recognition. Deep Audio-Visual Speech Recognition: Convolutional Neural Networks. Recurrent Neural Networks. Overlapped speech. Speaker targeted AVSR models. Visual Speech Recognition for mobile devices. Visual Speech Recognition DataSets. Experiments on each data set.