An Application of Deep-Learning Techniques to Face Mask Detection During the COVID-19 Pandemic

摘要

COVID-19 is an infectious disease that is primarily spread by respiratory droplets of infected patients. Their droplets are generated while they speak, sneeze, or cough. In Thailand, wearing a face mask seems a must when you travel and go around to other places. The policy of wearing a face mask is applied in almost everywhere both indoor and outdoor, especially indoor places. You will be not allowed to enter the places without wearing a mask. In this paper, we developed a convolutional neural network and trained it with the data set of 848 images with resolution of 416x416 pixels. The accuracy rate is 81% based on the performance of the trained model. We implemented it successfully and installed the system (both hardware and software) in a real environment, i.e., in a front gate of a campus building during an open-house event.