eduzhai > Applied Sciences > Engineering >

Masked Face Recognition for Secure Authentication

  • king
  • (0) Download
  • 20210507
  • Save

... pages left unread,continue reading

Document pages: 8 pages

Abstract: With the recent world-wide COVID-19 pandemic, using face masks have become animportant part of our lives. People are encouraged to cover their faces when inpublic area to avoid the spread of infection. The use of these face masks hasraised a serious question on the accuracy of the facial recognition system usedfor tracking school office attendance and to unlock phones. Many organizationsuse facial recognition as a means of authentication and have already developedthe necessary datasets in-house to be able to deploy such a system.Unfortunately, masked faces make it difficult to be detected and recognized,thereby threatening to make the in-house datasets invalid and making suchfacial recognition systems inoperable. This paper addresses a methodology touse the current facial datasets by augmenting it with tools that enable maskedfaces to be recognized with low false-positive rates and high overall accuracy,without requiring the user dataset to be recreated by taking new pictures forauthentication. We present an open-source tool, MaskTheFace to mask faceseffectively creating a large dataset of masked faces. The dataset generatedwith this tool is then used towards training an effective facial recognitionsystem with target accuracy for masked faces. We report an increase of 38 inthe true positive rate for the Facenet system. We also test the accuracy ofre-trained system on a custom real-world dataset MFR2 and report similaraccuracy.

Please select stars to rate!

         

0 comments Sign in to leave a comment.

    Data loading, please wait...
×