Open AccessA Novel Approach for Facial Attendance:AttendXNet
Author(s) -
Ankur Singh Bist
Publication year - 2020
Publication title -
aptisi transactions on technopreneurship
Language(s) - English
Resource type - Journals
eISSN - 2656-8888
pISSN - 2655-8807
DOI - 10.34306/att.v2i2.86
Subject(s) - attendance , computer science , artificial intelligence , facial recognition system , deep learning , face (sociological concept) , residual neural network , machine learning , spoofing attack , pattern recognition (psychology) , computer security , social science , sociology , economics , economic growth
Recent advancements in the area of facial recognition and verification introduced thepossibility of facial attendance for various use cases. In this paper we present a system namedas AttendXNet. Our method uses the ResNet and Multi-layer feed forward network to achievethe state of art results. Extensive analysis of various deep learning and machine learningtechniques is described. Face anti-spoofing is a major challenge in facial attendance.Extended-MobileNet is used to resolve the same issue. We also introduced the end to endpipeline to implement an attendance system for various use cases.