Coronavirus disease 2019 (COVID-19) was first identified in December 2019, Wuhan, China. It shortly became a global pandemic and affects the daily life of every individual. Currently, more than 150 million people have been infected. Even in UW-Madison, over six thousand students and eight hundred employees have been infected. Although COVID-19 vaccination is readily available through University Health Services, it is still vital for everyone to take precautionary measures. Wearing a mask can significantly reduce the chances of getting infected or spreading the virus. The school has also released guidance for face-covering on campus. Meanwhile, the effectiveness of face-covering would drop if it is not worn properly. Therefore, it is important to detect whether an individual on campus has a face covering and identify whether this person is wearing it correctly.

There are many companies or researchers that design and implement this kind of face mask detection system. In general, most of them achieved the goal of detecting whether people wear masks or not via developing a deep learning model. However, few of them can tell whether a person wears a mask properly. As a result, we plan to design a totally new face-covering detection system.