Face Recognition Attendance System

Face detection system has mostly gained attention for its huge application and market potential, such as face identification and video surveillance system. Real-time face detection and recognition is not only part of the recognition system but it is also becoming of the major subject for research. So, there are numerous ways to solve face detection. This paper will give a new implementation in automatic attendance management systems that is extended with computer vision algorithms. Supervised learning using convolutional networks (CNNs) has huge assumption in face recognition than unsupervised learning. In our project, we like to help the gap between the success of CNNs for supervised learning and unsupervised learning. We bring a class of CNNs called deep convolutional generative adversarial networks (DCGANs) that contains selective architectural constraints and denote that they have a strong prospect for unsupervised learning. We trained various image datasets, and showed decisive evidence that deep convolutional adversarial matches a hierarchy of presentations from object to scenes in both the generator and individual. Also, we use the learned features that demonstrates their significance as widespread image representations. This latest technique pursue to be much efficient than traditional methods like calling the names in the class; moreover, this is discrete and does not interfere with the regular teaching process. This method promises to give accurate and specific results of the face with more detailed information which shows student activity and their attendance in a classroom. This paper will introduce how we can apply algorithms for face detection and recognition in image processing to improve a system that will identify and recognize frontal faces of students in a classroom.
Department Name
Electrical and Computer Engineering
North South University
Printed Thesis