Résumé
Coronavirus illness 2019 has had a major impact on the entire world over the past two to three years. One important approach for people's protection is to wear masks in public. Furthermore, putting on a mask properly Many public service providers demand that users only utilise the service while properly wearing masks. Only a small number of studies have examined face mask identification using image analysis, nevertheless. We suggest Face Mask, a highly accurate and practical face mask detector, in this study. The suggested Face Mask is a one-stage detector that combines a novel context attention module for detecting face masks with a feature pyramid network to fuse high-level semantic information with various feature maps. We also provide a brand-new cross-class object removal method to reject and predictions with a high intersection of union and low confidence. Additionally, we investigate the viability of integrating Face Mask with a portable or embedded neural network called MobileNet. By utilising1)Contactless temperature sensing,2)we create a fack mask detection alarm system to boost COVID-19 indoor safety.Infrared sensor and contactless temperature sensing subsystems rely on Arduino Uno, while computer vision algorithms are used for mask identification. © 2023 IEEE.