Real Time Face Mask Detection System using Transfer Learning with Machine Learning Method in the Era of Covid-19 Pandemic

Asif, S.; Wenhui, Y.; Tao, Y.; Jinhai, S.; Amjad, K.

The rapid development of computer vision has attracted more attention to the global epidemic Covid-19 to enable human-computer interaction and improve public health services. Due to the rapid spread of the (Covid-19), various countries are facing a major health crisis. According to the World Health Organization (WHO) an effective way to protect people from Covid-19 is to wear medical masks in public areas. It is very difficult to manually monitor people in public places and detect the face mask in the video. which is mainly because the mask itself acts as an obstruction to the face detection algorithm, because there are no face signs in the mask area. Therefore, automatic face mask detection system helps authorities to identify people who may be susceptible to infections disease. This research aims to use deep learning to automatically detect face masks in videos. The proposed framework consists of two components. The first component is designed for face detection and tracking using OpenCV and machine learning, and in the second component, these facial frames are then processed into our proposed deep transfer learning model MobileNetV2 to identify the mask area. The proposed framework was tested on different videos and images using the smartphone camera. The purpose is to achieve high-precision real-time detection and classification. The model achieved 99.2% accuracy during training and 99.8% validation accuracy. which is better than other recently proposed methods. Experimental results show that the work proposed in this paper can effectively recognize face masks with multiple targets and provide effective personnel surveillance. This research is useful for controlling the spread of the virus and preventing exposure to the virus. © 2021 IEEE.