The rise of online education has increased the need for secure and fair exam systems. Many existing online proctoring tools face issues like identity fraud, tab-switching, false alerts, and privacy risks. This project introduces AssessMentor, a Python-based AI-powered platform built to enhance user trust and uphold academic integrity. Using FastAPI with PostgreSQL and aiortc (WebRTC), it enables real-time monitoring, while OpenCV and MediaPipe handle face detection and track suspicious behavior. A role-based model (Student, Staff, Admin, Master Admin) ensures transparent exam management. The system’s AI accuracy will be tested through empirical and survey-based research, focusing on detection precision, response time, and user feedback. Technical issues and privacy concerns are studied to improve reliability and usability. Compared with tools like ProctorU and Mettl, AssessMentor aims to deliver a scalable, explainable, and trustworthy online proctoring solution for modern education.
Dinkar et al. (Sat,) studied this question.