A real time robust facial recognition model for masked face images using Machine learning model

Abstract

Almost everybody is wearing a coronavirus mask to prevent the spread of COVID19 effectively. Conventional facial recognition technologies are often virtually ineffective, like Community access monitoring, facial access control, front attendance, and facial security inspections at railway stations. Consequently, recognition of the current masked face recognition technology must be improved as a matter of urgency. The most advanced approaches to facial recognition are based on a deep knowledge based on multiple facial samples. However, the public is currently unable to view any masked facial recognition datasets. In this study, 3 types of masked face recognition data sets are offered, including Masked Face Data Sets (MFDDs) (SMFRD). RMFRD is considered as the largest masked face dataset in the world. These data sets are available to industry and academia to develop various applications on masked sides. The model developed, which masked multi-granularity, achieves 99% precision, and goes further than the industry's reporting results.