Face Mask Detection Using Optimized CNN

Abstract

COVID-19 has had a rapid impact on people's lives, affecting global trade and transportation. Protecting against COVID-19 by wearing a face mask has become the new normal. Many public service providers will need clients to wear masks to access their services in the near future. As a result, in today's culture, face mask detection is essential. This study proposes attaining the aim by utilizing some basic platforms such as Machine Learning packages such as TensorFlow, Keras, and OpenCV libraries. The goal of this project is to reliably detect the face in an image and then determine whether or not the individual is wearing a mask. In addition, the model can detect the existence of a mask in real time. The mask detection dataset was compiled using Internet resources, and a Google form was constructed to collect photographs with and without masks. We examine optimum parameter values for the Sequential Convolutional Neural Network model in order to correctly detect the presence of masks without causing over-fitting. On camera or in real time, we want to see if a person wearing a face mask is actually wearing one.