Published Paper Details:

This project presents the design and implementation of an AI-driven Online Proctoring System aimed at enhancing fairness, transparency, and academic integrity in remote examination environments. With the rapid growth of digital learning platforms, educational institutions face significant challenges in monitoring student behavior during online assessments. The proposed system addresses these challenges by integrating advanced computer vision techniques, intelligent audio analysis, and an interactive web-based interface to enable continuous and automated exam supervision. Real-time facial landmark detection, gaze tracking, head-pose estimation, and posture monitoring are performed using MediaPipe and OpenCV to identify behavioral anomalies that may indicate inattentiveness or potential malpractice. In addition, an audio monitoring module analyzes ambient sound patterns to detect irregularities such as background conversations or the presence of multiple individuals. The system provides real-time visual alerts through an intuitive frontend interface and generates a post-examination distraction graph that summarizes detected events over time. Developed using a modular Flask-based architecture, the solution is scalable, cost-effective, and adaptable for integration with external databases and cloud services. The results demonstrate the effectiveness of the proposed system in supporting secure, efficient, and technology-enhanced remote examinations.