Face Mask Detection Using VGG-16 Net for Low Computation Power Devices

Abstract

With the current ongoing pandemic, wearing masks are the newest style apparel on the market. It has become a written rule that anyone who ventures out of their house must wear a mask. It has become imperative that face mask detection is a field to research on. The paper aims at creating a low computation power requirement model, using VGG-16 net. In this paper, we employ transfer learning on VGG-16 net to reduce training time and achieve maximum results. The dataset has many classes and has elaborate confusion classes that make the model stand apart. The model is easily deployable and works with both images and live video feed.