A Review on Drowsiness Prediction System using Deep Learning Approaches

Driver inattentiveness is a crucial cause for many accidents associated with vehicle crashes. Therefore, it is mandatory to develop a warning system. By considering multiple features like eye position, head gestures, hand gestures, mouth, vehicle-based parameters like lane deviation, steering wheel, brake control, EEG, ECG, etc. the drowsiness pattern among the operators can be assessed. Deep learning is the fastest emerging artificial intelligence technologies in the machine learning domain and used in solving image processing tasks. This paper summarizes on the most popular machine and deep learning algorithm utilized by the researchers to identify the in-alertness. In addition, datasets, sensors used, features, advantages and weaknesses of different techniques have been addressed.