Online Examination & Proctoring System

Abstract

The increasing adoption of online education has created a strong demand for secure and reliable examination systems, while simultaneously raising concerns regarding academic integrity in remote environments. This paper presents an enhanced AI-based Online Examination Proctoring System (OEPS) that incorporates role-based access control, biometric identity verification, and real¬time behavioral analysis to ensure fair and transparent online assessments. The system is designed using a full-stack architecture consisting of a React-based frontend, a Node.js and Express backend, Supabase database integration, and a FastAPI-based AI microservice. During student registration, facial data is captured and stored, which is later used for identity verification before theexamination begins. Real-time monitoring is performed using computer vision techniques through libraries such as face-api.js and MediaPipe to detect cheating behaviors including head movement, absence of face, and multiple face detection. All suspicious activities are logged as violations and made accessible to teachers via a dedicated dashboard for analysis. Additionally, the system supports exam creation, publishing, OCR-based question extraction using Tesseract.js, and performance analytics. The admin module provides complete system oversight by allowing accessto both teacher and student dashboards. Experimental observations indicate that the systemeffectively minimizes cheating, enhances examsecurity, andreduces reliance onmanualinvigilation, making it a scalable and efficient solution for modern digital education platforms

Introduction

The document presents an AI-based Online Examination Proctoring System (OEPS) designed to maintain academic integrity in remote exams, where traditional invigilation methods are insufficient and easily bypassed. It addresses issues like cheating, impersonation, and unauthorized assistance in online assessments using intelligent monitoring technologies.

The proposed system uses a full-stack architecture (React frontend, Node.js/Express backend, Supabase database, and FastAPI AI microservice) and supports three roles: Admin, Teacher, and Student. Students undergo facial enrollment and biometric verification before exams to prevent impersonation. During exams, real-time webcam monitoring uses computer vision tools (e.g., face detection, eye tracking, head pose estimation) to detect suspicious behavior.

A dedicated AI module flags violations such as looking away, multiple faces, or absence of the student, and logs them for teacher review. Teachers can create exams (including OCR-based question input), monitor attempts, and analyze results. The system also provides automated evaluation and detailed analytics after exams.

Evaluation results show that the system is scalable, responsive, and effective in detecting cheating behaviors, with reliable facial verification and real-time monitoring. However, performance can be affected by lighting, camera quality, and network conditions, and subtle cheating behaviors may still be difficult to detect. Overall, OEPS is presented as a secure, AI-driven, and scalable solution for modern online examination integrity.

Conclusion

In this paper, an AI-based Online Examination Proctoring System (OEPS) has been proposed and developed to address the challenges of maintaining academic integrity in remote assessment environments. The system integrates modern web technologies with artificial intelligence to provide a secure, scalable, and automated platform for conducting online examinations. By incorporating role-basedaccesscontrol,thesystemeffectivelymanagesinteractionsbetweenadministrators,teachers,and students, ensuring smooth operation and centralized monitoring. The implementationoffacialrecognitionfor identityverification, alongwithreal-timeproctoringusing computer vision techniques, significantly enhances the reliability of the examination process. The systemsuccessfullydetectsvariouscheatingbehaviorsand logsviolations, whichare madeavailable to teachers for further analysis. Additionally, features such as automated exam management, OCR-based question extraction, and performance analytics contribute to improved efficiency and usability. Theresultsdemonstratethattheproposed systemreducesthedependencyonmanualinvigilationwhile maintaining high levels ofsecurityand accuracy. Althoughcertain limitations exist, suchas sensitivity to environmental conditions and the need for more advanced detection models, the system provides a strong foundation for intelligent online examination platforms. Future work can focus on improving detectionaccuracyusingdeep learningtechniques,integrating multi-modalanalysis(audio and video), and implementing real-time alert systems for enhanced responsiveness. Overall, the OEPS systemhighlights the potentialofcombining artificialintelligence with full-stack development tocreatearobust andeffectivesolutionformoderndigitaleducation,ensuring fairness, transparency, and scalability in online assessments.

References

In this paper, an AI-based Online Examination Proctoring System (OEPS) has been proposed and developed to address the challenges of maintaining academic integrity in remote assessment environments. The system integrates modern web technologies with artificial intelligence to provide a secure, scalable, and automated platform for conducting online examinations. By incorporating role-basedaccesscontrol,thesystemeffectivelymanagesinteractionsbetweenadministrators,teachers,and students, ensuring smooth operation and centralized monitoring. The implementationoffacialrecognitionfor identityverification, alongwithreal-timeproctoringusing computer vision techniques, significantly enhances the reliability of the examination process. The systemsuccessfullydetectsvariouscheatingbehaviorsand logsviolations, whichare madeavailable to teachers for further analysis. Additionally, features such as automated exam management, OCR-based question extraction, and performance analytics contribute to improved efficiency and usability. Theresultsdemonstratethattheproposed systemreducesthedependencyonmanualinvigilationwhile maintaining high levels ofsecurityand accuracy. Althoughcertain limitations exist, suchas sensitivity to environmental conditions and the need for more advanced detection models, the system provides a strong foundation for intelligent online examination platforms. Future work can focus on improving detectionaccuracyusingdeep learningtechniques,integrating multi-modalanalysis(audio and video), and implementing real-time alert systems for enhanced responsiveness. Overall, the OEPS systemhighlights the potentialofcombining artificialintelligence with full-stack development tocreatearobust andeffectivesolutionformoderndigitaleducation,ensuring fairness, transparency, and scalability in online assessments.

Copyright

Copyright © 2026 Palash Poharkar, Sahil Maske, Parth Lavtawar, Pranay Bhure, Palvesh Dandre, Mr. Kunal Kanchankar. This is an open access article distributed under the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.