Yosi Kristian, A.P. Mo Raymond, F.X. Ferdinandus, Gunawan, Arya Tandy Hermawan, Ilham Ari Elbaith Zaeni, Jehad Abdelhamid Hammad
The World Health Organization (WHO) pronounced a new type of coronavirus called COVID-19 as a global pandemic in March 2020. Wearing a mask and maintaining a distance were required to prevent the spread of the virus. However, some people still did not care about the protocol. Therefore, this work aimed to provide a deep learning-based system to monitor the implementation of physical distancing and face mask usage. YOLOv5 was used to estimate the physical distance between people. This chapter implemented RetinaFaceNet as the face object detection and MnasNet as the face mask usage classifier. The transfer learning approach boosted the overall efficiency of the model. The result was promising, with the overall system achieving an F1 score of 0.852. © 2024 selection and editorial matter, Aji Prasetya Wibawa.
Institut Sains dan Teknologi Terpadu Surabaya, Surabaya, Indonesia; Universitas Negeri Malang, Malang, Indonesia; Al-Quds Open University, Abu Dis, Palestine