Secure Online Assessment Platform with AI-based Proctoring

The system used in this project is an AI proctored (Online Proctoring) application which allows for high supervision of online examinations. It promotes MCQ-type examinations for tutors to design exam tests of various study areas. The proctoring model uses videocapturing technology to watch students — not just that they are focused and have their heads down, but the angles of their head in relation to the screen (for example any deviation from straight on is suspect: left, right, up, or down). A screenshot of the students using from that screen is obtained at 5 seconds after each other and put in opposition to a list (placed inside Excel) for speedy assessment of scholar behaviour later. Once the examination is completed, all video tapes will be kept in DB then we make it polished for viewing with using YOLOv8 algorithm. Additionally, for accounts security the authors have implemented two factor hardware authentication and password hashing mechanism to protect exam. A great initiative to practice for real-time monitoring, AI-assisted behavioral assessment, and security measures all put together in a single platform can be online secure assessments.