Detection of COVID-19 Protocols Violation in Real Time using Deep Learning

Abstract

COVID-19 pandemic has created a severe health emergency all over the globe since last couple of years and is still emerging in few countries. According to the World Health Organization (WHO), around 520 million cases and 6.2 million casualties due to COVID-19 have been reported till the writing of this manuscript, 19th May 2022. The COVID-19 protocols including wearing masks, following social distancing have been imposed in almost all the countries worldwide. It is a challenge to track the adherence of the COVID-19 protocols by the people in real time. This work proposes a model for the detection of COVID-19 protocols violation in real time. We have also created a web application which uses the proposed model to detect the adherence of COVID-19 protocols in real time. The proposed model is tested on a dataset comprises of 1376 images and has shown promising results even in complex environment.