Volume 9 Issue 3 ( September 2022 )


DDNet- A Deep Learning Approach to Detect Driver Distraction and Drowsiness

Prachi Panwar, Prachi Roshan, Rajat Singh, Monika Rai, Asha Rani Mishra, Sansar Singh Chauhan


Road accidents are the main cause of death among the human population. Distracted and drowsy driving takes thousands of lives every year around the world. Subsequently, to forestall such mishaps and save lives, there is a requirement for a system that detects both distraction and drowsiness for both day and night time. In this paper, we present a deep learning convolutional model to detect distraction and drowsiness during driving. The proposed model performs real-time video processing for monitoring the activities of drivers during driving. The model produces an alert in case of any careless driving or inappropriate behaviour of the driver with the minimum response time. For this purpose dataset for training as well as for testing were prepared. For training the model, we have used CNN model. The proposed model was able to achieve 99.95% accuracy on test dataset.

Keywords: Deep Learning, Convolutional Neural Network, Classification, ReLu, Validation loss, Real-time video processing, Drowsiness and Distracted Driving Detection System