A survey of visual speech recognition using deep learning

Visual speech recognition is one of the revolutionary applications of AI. An exceptional studies and improvement are occurring in a good way to gain the quality of it and is slowly catching as much as audio-primarily based speech recognition. This provides the machine to understand the languages in a noisy environment and can also be used for applications related to speaking problems, medical problems like ALS and biometric securities. It makes machine possible to understand human speech using the visuals only. This can make the silent speech from theory to practical. In recent works visual speech recognition models have outperformed human lip-reading experts for the given vocabulary constraints.