takes in a sequence of lip images, and predicts the phonemes being said. - View it on GitHub