Efficient Face Mask Detection for Banking Information Systems

Summary

The creation of a real‐time face mask identification system utilizing the Single Shot Detector (SSD) algorithm and MobileNetV2 architecture is the main goal of this study. With the aim of addressing the challenges posed by the COVID‐19 pandemic, this study explored automated methods to monitor face mask compliance in public spaces. By combining deep learning techniques and OpenCV, the system enables the real‐time detection of CPU‐based devices, eliminating the need for GPUs. This study compared the performance and speed of different model architectures and determined MobileNetV2 as the most suitable choice for CPU‐based devices. The developed face mask detection model achieved a high accuracy score of 0.97, on the testing data. The system's applicability extends to edge devices, making it suitable for deployment in various public settings, such as banking information systems, parks, schools, hotels, and hospitals. This study contributes to the development of an efficient and accessible solution for monitoring face mask usage, aiding in the prevention of virus transmission during ongoing pandemics and future outbreaks.