International Research Journal of Engineering and Technology (IRJET)
e-ISSN: 2395-0056
Volume: 12 Issue: 11 | Nov 2025
p-ISSN: 2395-0072
www.irjet.net
Design of Face Detection and Recognition System to Monitor Students During Online Examinations Using Machine Learning Algorithms R.Manjupriya1, Saran S2, Varshanth D3 Jotheshvar D4 1 R.Manjupriya, Department
of Artificial Intelligence and Data Science, SRM Valliammai Engineering college, Tamilnadu, India 2 Saran S, Department of Artificial Intelligence and Data Science, SRM Valliammai Engineering college, Tamilnadu, India 3 Varshanth D, Department of Artificial Intelligence and Data Science, SRM Valliammai Engineering college, Tamilnadu, India 4 Jotheshvar D, Department of Artificial Intelligence and Data Science, SRM Valliammai Engineering college, Tamilnadu, India ---------------------------------------------------------------------***----------------------------------------------------------------------Individuals are more inclined to trust online testing Abstract— The global pandemic has sped up the growth of platforms. [5], [6].
online education, which has turned real classes into online places to learn. One of the hardest things for schools is making sure that students are present and that students are honest on online tests. This research uses the K-Nearest Neighbor (KNN) algorithm to demonstrate how a tracking system based on face recognition and detection could function. During tests, built-in webcams use face recognition to make sure that students are who they say they are. This also takes notes. It is the Python Face Recognition Library that trains the machine learning-based face embeddings that are used in the system idea. When people sign up for an online test, these embeddings are compared to listed datasets to make sure they are who they say they are. A management interface, a student site, and a database-driven authentication system that works with both MySQL and Java Servlet are all part of the basic model. There are pros and cons to the system and the basics of algorithms are talked about in this paper. It also lays the groundwork for real-time use in educational tracking systems in the future.
Educational institutions around are endeavouring to enhance digital management systems and provide comprehensive evaluation of all students, regardless of their physical or virtual classroom settings. A tool called computer vision is being used for this. With machine learning, you can get accurate and flexible results when you do identification work. There are many kinds of algorithms. One of them is the KNN method. Good for jobs that need you to talk to other people. It works quickly and well. KNN gives test pictures a score based on how well they match up with pictures that know what they are. In terms of how far apart their feature vectors are, the picture that looks most like the test image gets the best score. [7]. The KNN method can be used with face embeddings made with Dlib or Python's Face Recognition module to make sure that students' names are spelled correctly even if the lighting or stance changes. It is the best choice for small, light school security systems because of this [8, 9].
Keywords— Face Recognition, K-Nearest Neighbour (KNN), Online Examination Monitoring, Attendance Automation, Machine Learning.
The main purpose of this study is to develop a face recognition and tracking system that can verify students' names, watch test-takers, and keep real-time records of attendance. Front-end web technologies, recognition reasoning based on machine learning, and database-driven proof have made it simple for teachers and students to use. Not like some other systems on the market, this one doesn't need to be linked to the cloud or watched by someone. This system, like some others on the market, doesn't need to be connected to the cloud or watched by a person. But it's not expensive, and school PCs can use it. Its goal is to make AIbased biometrics work better so that online tests are safer, faster, and easier to get into.
1 .INTRODUCTION In-person administration of these tests is currently unfeasible because to the pandemic. Online lessons and assessments are increasingly transforming educational institutions. Monitoring the children during online instruction and assessments is more challenging. Robust facial recognition technologies are the most effective solution to address this problem. Currently, digital technologies for monitoring assessments play a significant role in maintaining good standards in technology-based courses. As an increasing number of students enroll in online courses, the likelihood of scams and identity theft escalates. Utilizing another individual's account and password to verify their identity is unacceptable. Verifying each student's identity is straightforward, expedient, and secure.
© 2025, IRJET
|
Impact Factor value: 8.315
|
ISO 9001:2008 Certified Journal
|
Page 649