Lip Reading System Using Deep Learning

Deep learning is one of the emerging technologies which is becoming popular nowadays due to its versatility, ability to handle larger data and ability to increase effectiveness by training more data. By using deep learning we can solve many real life as well as research field problems. Lip reading is the process of predicting text from a human's lip, mouth and facial movements. It is one of the essential parts of communication for the people who cannot speak. It removes communication gaps between deaf and dumb people, even though it is not prevalent as an audio based form of communication. This study explores the creation of effective solution using deep learning technology that can extract spatial and temporal features from lip movements from small video clips. Using deep learning techniques and LSTM, providing an optimal solution is the main focus of this study.