Universitas Dian Nuswantoro conducted research on the development of a face mask detection system for Covid19 monitoring. The activity addressed Covid19 as a pandemic disease that affected the world, including Indonesia. Participants identified in the activity context are Djuniadi, N. Iksan, A.F. Hastawan, A.F. Suni, A.A. Firdaus, L. Hidayati, D.K. Setiawan, E.D. Udayanti, and F. Prasetyanto.
The research focused on detecting whether people used masks through a deep learning method, with particular attention to object detection. The resulting face mask detection system can assist monitoring of compliance with mask use in public areas that have the potential to cause crowds. In this way, the system is directed toward supporting a safe environment. The research therefore relates to SDG 3 through its focus on a health-related response to Covid19 and to SDG 4 through the application and development of a deep learning method.
The system used the MobileNets V2 architecture. The source reports that this deep learning method was quite effective and produced an accuracy result of 0.99 in testing the classification training data at 10 epochs. The system was also implemented in real-time direct testing. The activity was positioned alongside efforts to prevent virus spread, provide information services concerning virus spread and emergency response, detect people suspected of or infected with the virus, and implement programs for preventing spread and vaccination for all elements of society in Indonesia.
Through this research activity, UDINUS presents a technology-based contribution to Covid19 monitoring and mask-use compliance, within the scope and findings reported in the source.