A Novel and Efficient Method to Detect the Face Coverings to Ensurethe Safety using Comparison Analysis
Ketan Rathor, Keyur Patil, Mandiga Sahasra Sai Tarun, Shashwat Nikam, Devanshi Patel, Sasanapuri Ranjit
Abstract
A face mask worn by a single person or a group of individuals can be identified using face mask detection, an AI-based system that examines a video stream. A confidence value is produced for each detection by our DeepSight programme. Either a person is labeled as "wearing a mask" or "not wearing a mask" depends on their classification. The World Health Organization (WHO) recommends that a face mask completely cover the face, including the chin and nose. Face Mask Detection System may be used with both CCTV cameras and already-installed USB or IP cameras. To distinguish between persons wearing masks and those who aren’t, it employs the most recent computer vision techniques. Compact model’s accuracy is really great, and it can tell if you’re wearing a mask properly or not (i.e.; mask covering the nose). Face mask detection data may be used to a wide range of settings and sectors, including corporate offices, shops, public transit, airports, and taxis. The main motive of the paper by the author is represented as solving the problem using cnn and transfer learning and also a comparison between both the approaches.