Detection of Driver Drowsiness Using Multi-task CNN Framework

Abstract

Driving may be a skill that needs our full attention to safety control the vehicle and reply to event happening on the road. Distracted driving becomes a bigger threat per annum and has been the leading causes of accidents for the past decades. There are 3 forms of distractions (i) eyes off the road (visual), (ii) mind off the road (cognitive), (iii) fatigue or drowsiness while driving. in step with the study released by National Highway Traffic Safety Administration(NHTSA) and also the Virginia Tech Transportation Institute(VTTI), 80 percent of collision and 65 percent of near collision involve some kind of driver distraction. To deal with these challenges, we introduce a system to observe the motive force in terms of fatigue, distraction and activities. This method deals with automatic driver activity detection supported visual information and MTCNN. We propose an algorithm to trace and analyze drivers eyes PERCOLS(percentage of eyelid closure), heartbeat rate.