Abstract
Distracted driving is an act of driving while doing some other activity that distracts the mind away from the road. A prominent factor in the increase of a vehicle crash is Driver distraction. India is a major contributor to road crash transience. The road crash rate in India is 16 lives per hour. The driver's action detection has always attracted many researchers towards it. This paper focuses on recent developments in the area of the driver's movement detection. The proposed methodology used to identify the features present in a driver's distraction activity. The features that we extract will further feed into the CNN network for classification Different CNN models like ImageNet, AlexNet, VGG-16, VGG-19 and InceptionV3, Gaussian Mixture Mode, Multiple Scale Faster-RCNN and many more are used to classify driver’s behavior. The proposed method will use the best among the above given CNN models. Section IV compares …