A Study of Smart Presence System by using Machine Learning Approach

Abstract

In large organizations, the time and effort spent verifying whether or not employees have met their attendance quotas may become a substantial burden. In order to solve this issue, a presence mechanism that is both automated and efficient is being built. However, this technique relies heavily on verification. Smart Presence System deployments often make use of real-time facial recognition and identification technologies. In this investigation, we use two different kinds of algorithms. It uses the Convolutional Neural Network (CNN) technique as well as the Haar Cascade Classifier. The Haar Cascade Classifier technique was used during development of this function. We used the CNN algorithm to compare the results. When a user's face is detected, the system will generate a new spreadsheet for each day of the week. This real-time face recognition and identification technology is restricted to authorized corporate personnel only. Those who have not yet registered may have their information confirmed using a QR code verification mechanism. This system might make use of selected pieces of the user's data. Users with and without accounts may cohabit without incident within the parameters of this system. We got 99% accuracy after using the Haar Cascade Classifier.