Masked Face Detection for Effective COVID-19 Containment: A Light Convolution Neural Network Based Model

COVID-19 a pandemic caused by novel coronavirus has been circulating across the world for a long time. COVID-19 has had an influence on almost every aspect of growth. The healthcare system is in a state of emergency. According to WHO, wearing a mask is mandatory to protect ourselves from the virus. In this paper, a framework has been proposed for limiting COVID-19 growth by identifying people who are not wearing a facial mask in a crowded places of a city with the help of CCTV. When a person without a mask is detected, the appropriate authority is notified via the city network. A light CNN model has been proposed to identify the presence or absence of mask from face images. The proposed CNN model has given high accuracy on publicly available masked face datasets namely LFW-s, SMFD and RMFD. It has outperformed ResNet-50 and EfficientNetB0 CNN models.