Face Mask Detection in the Era of COVID-19: A CNN-Based Approach

Abstract

AbstractThe global epidemic of the coronavirus COVID-19 is wreaking havoc on the world’s health and according to the World Health Organization (WHO), using a face mask in crowded locations is among the most common security practices. An artificial neural network for face mask classification utilizing deep learning will be introduced in this research. As the outbreak of the COVID-19 pandemic, a remarkable development in the fields of object recognition and computer vision has been made in the identification of face masks. Many architectures and methods have been used to construct a variety of face recognition models. Face masks can be distinguished using the method proposed in this work, which makes use of deep learning, TensorFlow, Keras, and OpenCV. This approach may be evaluated for use in protection jobs due to the fact that it is quite inexpensive to execute. In fact, the GAN-generated face-masked datasets have been selected for evaluation purposes. Compared to other standard Convolutional Neural Network models, the proposed framework outscored them all, attaining a 99.73% accuracy rating.KeywordsCOVID-19Masked faceConvolutional neural networkData augmentationDeep learningTransfer learning