Face-Recognition Based Attendance System using One-Shot Learning

R L Priya,Shreeja Nanda,Maitraiyi Dandekar,Anmol Devnani,Riteshsingh Kadakoti
DOI: https://doi.org/10.1109/icccnt51525.2021.9579899
2021-07-06
Abstract:An attendance marking system is a part of every institute, college, corporation, be it big or small. With giant strides made in the domain of image processing, we have created a system that uses the latter to simplify the former. Face Recognition is growing increasingly popular over other biometric authentications because of its easy, non-invasive, and contactless approach. The main objective of the system is to detect faces and recognize them in a real-time environment, match them with the data present in the database and record their attendance. This is aimed to ease the hectic manual attendance process and make it more efficient. This further also solves the problem of authentication and proxies as biometrics are unique, and facial features used for Face Recognition being one of them, cannot be duplicated. The designed system uses OpenCV, dlib, Face Recognition libraries, and One-Shot Learning for face detection and recognition which requires only one image per person in the database, thus also being space-optimized as compared to traditional training-testing models.
What problem does this paper attempt to address?