ABSTRACT
The covid-19 pandemic has affected both the health and lifestyle of the people, not only this the global economy also affected badly. The virus spreads at a very high rate and people can easily be infected. So for that people have to take a vaccine that is provided by their government but only vaccination is not a complete solution to this virus. It doesn't gives a full guarantee to prevent people's lives even after the vaccination, so we have to defend ourselves from the spread of the viruses as much as possible. For that mask and social distancing are the main key factor that is also recommended by the government and the public health agencies. As people are not habitual of wearing the mask so many times people forget to wear a mask in public places which is one of the main reasons for the spreading of the virus. Sometimes we see that at crowded places like metro stations and malls and universities, two or three guards are always present, to check the thermal temperature and if people are wearing masks or not and telling people to maintain a social distance. So there are a lot of problems in this because the metro station is a crowded place there people have to make a queue for checking their temperature so it is somehow hard to maintain social distancing and if there is an infected person found in the queue then all other surrounded people will also become infected. So we decided to solve this problem by contributing to the public health sector by making a complete system that will check if people are wearing a mask on their face properly or not. It will also check the thermal temperature of the people through the cameras and checks if someone is not violating social distancing rule. This will prevent people from infected people and also save the time of people. Now a maximum of one guard is needed for monitoring. To make this project practically we are taking the help of machine learning and deep learning. We will be using face detection and recognition algorithms that will be detecting the faces of people. We are using python as programming language. For face detection, we are using YOLO v4(You look only once) which supports a Convolutional neural network. So our process flow will be like that first we import all libraries and after that, we will build a neural network and after that training will be done on a model and then testing the model. After that system will become ready to deploy on the cloud. © 2022 IEEE.