Abstract:
To track the participation of students is mandatory in numerous educational
sectors. The manual administration of the participation sheets is difficult for
swarmed study halls. Face detection and identification frameworks, as one of the
sub-parts of computer vision, were initially intended for public reconnaissance.
In this study, a novel mechanism is proposed with a face detection system using
Convolutional neural networks (CNN), and Support vector machines (SVM)
techniques for tracking the attendance of students. Even though several automated
models of attendance are used in schools and universities, the proposed approach
uses the effectual machine learning and deep learning techniques to increase the
effectiveness of the approaches. In the proposed model, multiple cameras are used
to take the photo in 360 degrees (reduce the possibility of missing some faces of
students) and then the face detection process will be applied by applying a few
machine learning approaches to detect the faces of students. Lightning, context,
and pose variation will be extracted using Local Binary Patterns Histograms
(LBCH) algorithm, and then, the face recognition process will have conducted by
CNN and SVM algorithms. Finally, the attendance report will be generated by
the matching process via matching the captured images with stored images after the
duplicate removing process by applying the AdaBoost classification algorithm. We
believe that our proposed approach based on face detection by applying several
machine learning algorithms and deep learning algorithms can be used to track
student attendance and prevent fake attendance effectively.
