Enhancing Academic Integrity: Automated Surveillance and Cheating Detection in University Exam Halls

Overview

This project aims to enhance academic integrity in university examination environments by developing an AI-driven automated cheating detection and surveillance system. Leveraging computer vision and deep learning techniques, the proposed framework analyzes exam hall surveillance footage in real time to detect, classify, and flag suspicious student behaviors indicative of potential cheating.

The system is designed to support human proctors by providing objective, scalable, and efficient monitoring, as well as automated video summarization for post-exam review. The framework will be trained and validated using both public datasets and anonymized real-world exam data from the University of Dubai, ensuring robustness and practical applicability. By integrating intelligent surveillance with assistive proctoring, the project seeks to modernize exam monitoring practices and strengthen trust in assessment processes.