Authors: Professor Swati Bagade, Khushi Dilip Wable, Siddhi Virendra Galande, Shreya Vinod Galande, Chanchal Suresh Khandarkar
Abstract: Wearing face masks is a recommended preventive measure to curb the spread of infectious diseases, particularly SARS-CoV-2. Consequently, automated detection of mask usage, including proper placement and mask type, remains a key area of research. Coronaviruses, a vast family of viruses, have significantly impacted public health due to their high transmissibility. To safeguard public health, individuals are encouraged to practice social distancing, maintain hand hygiene, and most importantly, wear face masks. Mask usage has become widespread globally, with densely populated regions, such as India, facing heightened challenges in ensuring compliance. The developed model has undergone through training and validation using a real-world dataset and has been additionally tested on live video streams to assess its effectiveness. The system’s accuracy has been evaluated under various conditions, including different distances, positions, and multiple individuals within a single frame, ensuring reliable and consistent performance.
