Classification of Mask Use during a Pandemic using the CNN Algorithm with Voice Notifications

Abstract

Various technologies were created to prevent the threat of the Covid-19 virus, which has spread in many countries including Indonesia. One of them is the use of masks in public places. With this in mind, this study aims to detect facial objects. Based on the Kaggle website, the object used for research is a human face in 2D form. This research consists of two stages, namely creating and testing a model. The model is a system that detects and classifies faces with masks, inappropriate masks and without masks. Then the model is tested for its accuracy. The result of thirty trials, the model has an accuracy of 99% which is tested using a webcam in real time. This model has a sound indicator which is a notification to faces using the Convolutional Neural Network (CNN) algorithm method.