AN INTELLIGENT SYSTEM FOR SOCIAL DISTANCE DETECTION USING DEEP LEARNING TECHNIQUES

Abstract - Social distancing and wearing mask properly are the most effective ways to reduce the infection in current pandemic situation. So, we are developing a model that automatically detects the Social Distance violation prescribed by WHO (which includes maintaining a minimum of 6 ft distance and wearing the face mask). The solution includes developing a model that predicts social distance violation for that, a camera is used as input for frame of video and YOLO-V3 objection detection model which detects people for calculating social distance by applying distance measurement formula and individual identification of face mask violation by notifying them. Index Terms -YOLOv3((You Only Look Once),MobilenetV2,FPS(Frame Per Second), CNN(Convolutional neural network).