ABSTRACT
COVID-19 virus has hit Indonesia since early March 2020. One of the government's efforts to prevent the spread of COVID-19 is to do physical distancing to require people to wear masks when doing activities outside the home. One way to overcome this problem is by detecting mask users to be more obedient and obedient to the rules, then the identification process is carried out for mask users and those who do not use masks. The process is carried out using the Convolutional Neural Network method. CNN is known to be superior and does not require pre-processing so it saves more time. In terms of algorithmic competence, CNN is considered capable of carrying out the data detection process well. Of the 1376 datasets used, 30 epochs, accuracy = 0.988, recall = 0.990, precision = 0.987, and F1 = 0.988 with the required detection time for each image between 4 to 5 seconds. © 2022 IEEE.