A Real-Time Driver Drowsiness Detection Using Deep Convolutional Neural Network

A Real-Time Driver Drowsiness Detection Using Deep Convolutional Neural Network

Siddharth Singh1, Peeyush Kumar Pathak2

1,2Computer Science and Engineering, Goel Institute of Technology and Management, Lucknow.

---------------------------------------------------------------------***---------------------------------------------------------------------

Abstract - Continuous video stream analysis is used by driver monitoring systems to detect drowsy drivers. This technique counts the frequency of eye blinking and shutting by taking a photo or frame per second. If these patterns indicate fatigue, the system warns drivers to take a break or switch drivers. This method, which is a component of a bigger initiative to improve road safety using technology, combines immediate feedback with real-time monitoring.The system accurately identifies drowsiness and distinguishes between stages of drowsiness, with a robustness in driving situations and sensitivity to even the smallest tiredness indicators, outperforming the VGG16 model by 99%.

Key Words: Driver Drowsiness Detection, CNN, DDD system, Deep Learning, SVM.