Real-world application of face mask detection system using YOLOv6

The COVID-19 pandemic has drastically reshaped the human lifestyle and has placed immense importance on our health, safety, and sanitation practices. Among the various safety protocols assigned by the World Health Organisation (WHO) for the same, the usage of face masks to prevent the spread of the virus from an infected person to a healthy person has been of prime significance. To enable efficient execution of the WHO protocol, this case study proposes creating a real-time detection model built explicitly for capturing an audience to alert people who are not following COVID prevention protocols. The proposed case study utilises the state-of-the-art (SOTA) YOLOv6 algorithm along with different iterations of the YOLO algorithm, such as YOLOv4, and YOLOv5, for representing the variation in training performance among various iterations of YOLO. Further, it discusses and analyses the effectiveness of using a real-time detector for face mask detection. This study aims to decrease the risk of a healthy person being affected by the COVID-19 virus by keeping a check on a designated crowd and contributes towards the prevention of the further spread of the virus by crowd monitoring and control methods. The real-time implementation of the proposed case study reports a positive impact, with a 36% increment in people following the standard COVID-19 protocol of wearing masks in public places.