Malpractice Detection in Online Assessments Using Eye Gaze Tracking and Object Detection

Abstract

Internet-based client authentication protocols must be strengthened to reduce attacks and security vulnerabilities that threaten the performance of apps in fast internet distribution and cloud computing. Due to a multitude of benefits such as effectiveness, convenience, simplicity, and usability, distance and digital training (called e-learning) seems to have become the mainstream in skills and retraining. Secondly, because just like the COVID-19 pandemic's physical isolation rules, online learning has now become the exclusive possibility. Due to the lack of physical existence, however, online systems are a major issue in monitoring attendees and students over sessions, particularly during tests. It is necessary to establish technological tools that deliver survey clearly for monitoring unfair, unethical, and unauthorized behavior in classes and examinations. In this dissertation, we develop a modern online proctoring system based on machine learning.