Face Mask Detection in Public Places Using Small CNN Models

ABSTRACT

The spread of coronavirus can be prevented among the people in a crowded place by making face mask mandatory so that the droplets from the mouth and nose would not reach the masses nearby. The negligence of some people, i.e., by not wearing the mask, causes the spread of this pandemic. Therefore, persons who do not wear masks should be tracked at the entrance to public venues such as malls, institutions, and banks. The mechanism proposed warns if the individual is wearing or not wearing the mask. The proposed system is built in a small CNN model to integrate any low-end devices with minimal cost. The small CNN model like ShuffleNet and Mobilenetv2 are evaluated in Transfer Learning and Deep Learning but the Deep Learning model has better performance than the Transfer Learning. Again, the Deep Learning approach, i.e., mobilenetv2 plus Support Vector Machine achieved 99.5% accuracy, 99.01% sensitivity, 100% precision, 100% FPR, 99.51% F1 score, 99.01% MCC, and 99.01% kappa coefficient. © 2022, The Author(s), under exclusive license to Springer Nature Singapore Pte Ltd.