Design and Implementation of an AI exam invigilator

Abstract

The rapid growth of online and computer-based examinations has increased the demand for reliable, scalable, and intelligent examination monitoring systems. Traditional invigilation methods rely heavily on human supervisors, which can be inefficient, costly, and prone to human error. This project presents the design and implementation of an Artificial Intelligence (AI)-based Exam Invigilator system that automates the process of monitoring candidates during examinations. The proposed system utilizes advanced AI techniques such as computer vision, facial recognition, eye movement tracking, object detection, and audio analysis to detect suspicious activities and ensure examination integrity. A webcam continuously captures video streams of examinees, while machine learning algorithms analyze facial orientation, multiple person presence, unusual head movements, mobile phone usage, and absence from the screen. The system can generate real-time alerts and maintain logs of detected malpractice incidents for later review by administrators.