Design of an Efficient Distracted Driver Detection System: Deep Learning Approaches

Distracted driving is any activity that deviates an individual’s attention from driving. Some of these activities include talking to people in the vehicle, using hand-held devices such as mobile phones or tablets, eating drinking, and adjusting stereo navigation systems while To counter effects caused by distracted driving, many countries around world have imposed rules charged fines on drivers order ensure safe Owing technological advancement recent times, modern-day technologies computer vision, image processing, machine learning techniques can further support efforts governments prevent accidents In this paper, efficient driver detection scheme (DDDS) has been proposed two robust deep architectures, mainly visual geometric groups (VGG-16) residual networks (ResNet-50). The DDDS contains pre-processing module, augmentation techniques, classification modules based architectures. Both architectures are implemented, results compared terms performance indices, namely accuracy, logarithmic loss. two-dimensional (2D) dashboard images derived State-Farm dataset pre-processed used for training, testing, validation Accuracy 86.1% 87.92% achieved with VGG-16 ResNet-50 models, respectively, it observed found highly c4, c5, c7 categories dataset. obtained methodology existing literature be satisfactory. algorithms developed discussed instrumental reducing fatalities injuries due