Automatic Attendance Management System Using AI and Deep Convolutional Neural Network

Abstract

Traditionally, the attendance of scholars has been a significant concern for colleges, and schools mainly need to pay valuable of your time, and it may be tedious to mark group actions manually. The present biometric group action system isn't automatic; it wastes heaps of your time, is tough to take care of, and needs a queue for scanning fingerprints to mark their group action. Within the epoch, everybody encompasses a smartphone and is connected via the web on every occasion. During this paper, group action watching will be done through smartphones offered to most school members. Some in-style object detection algorithms are back propagation neural network, region-based convolution network (RCNN), quicker RCNN, and single-shot detector. Our unified structure relies on the YOLO V3 (You solely look once) formula for face detection and Microsoft Azure victimization face API for face recognition (face database). The distinctive half maybe a camera put in within the room can take footage doubly one at the beginning and one at the tip to confirm students have attended complete category. YOLO V3 can first count the scholars in a picture, followed by distinguishing faces as well-known and unknown, generating spreadsheets individually, and distributing an email at the tip of the month to students, parents, and college. The designed system performs expeditiously in real-time implementation for enumeration and detection. Our entire system has tried to assemble high accuracy in face detection and performance.