Real-Time Driver Drowsiness Detection Using Hybrid CNN-LSTM Model with Facial Feature and Behavioral Analysis

Driver Drowsiness is a leading cause of road accidents. It is responsible for around 20 percent of fatal accidents all around the world. Early detection of driver tiredness is crucial in preventing road accidents and enhancing road safety. There are traditional detection methods like self-reporting, physiological and vehicle-based measures but they are either unreliable or intrusive which limits its effectiveness. In this paper, we have developed a deep learning-based technique to correctly detect driver drowsiness in real time using a non-intrusive method. The proposed approach consists of physiological monitoring and computer vision approach to analyze facial features, head poses, blink duration, blink rate, rate of eye closure and yawning using the camera within the vehicles. These features are processed by deep learning models which are compatible with computer vision. CNN model along with LSTM is used in finding temporal patterns and trained on diverse dataset to predict the drowsiness levels. Results indicate hybrid model CNN-LSTM is effective in detecting drivers drowsiness.