Live Facemask Detection System

In the current ongoing situation of the pandemic, it has become necessary for people to wear a mask in order to protect themselves from exposure of the wide spread Novel-CoronaVirus, however many people do not wear it. The aim of this paper is to depict a system created which detects whether a person has worn a mask or not. For achieving this aim, a dataset consisting of 18236 images of people wearing a mask and without a mask is created. Using the same dataset, 101 layers deep, ResNet-101 convolutional neural network is trained. Indeed, the algorithm step regarding mask detection accomplished an accuracy rate of 96.02%. Lastly, the model is deployed to the RaspberryPI board