A Review on YOLO Algorithms for Social Distancing

Object detection is one of the appealing branch of the study in the fields of computer science. Object detection defines some popular object detection models in deep learning such as YOLO (You Only Look Once), Single Sort Detector, R-CNN etc. It can be applied to solve many important challenges in computer vision, human-computer interaction such as health-care monitoring, autonomous driving, anomaly detection etc. Social distancing terminology has been coined repeatedly in-light of COVID19. Social distancing is an inherent part of health-care systems. In general, nowadays, social distancing applications have become more important due to the transmission of various diseases like smallpox, Scabies, COVID19, etc. It happens due to not maintaining of specified distance between infected peoples. The paper proposes a review of real-time applications of YOLO based object detection in social distancing. The present work is useful to understand the overview of YOLO models that can be applied for social distancing. Additionally, comparisons between the various YOLO versions with respect to different performance metrics concerning speed, accuracy, and computational complexity. These results are vital for the selection of an optimal YOLO version for real-time social distancing application.