Improving technology in the field of education
will greatly help teachers and students in various ways, one of
which is attendance. An attendance system generally leaks
security and verification which may lead to fraud activity. In this
paper, the authors design an attendance system that utilizes
multiple verifications using card tapping and face recognition
which is also accompanied by an anti-video spoofing system. The
designed system is implemented as a web application with
various algorithms such as Convolutional Neural Network
(CNN), and Deep Metric Learning (DML) along with the
application of the PN532 sensor and the use of ESP8266 for
tapping the card. Our experiments show that the proposed
system performs well, achieving up to 87.50% system accuracy.
27
Mar