AI-Enabled Adaptive Online Proctoring with NeuroCognitive Monitoring

Abstract

The AI-Enabled Adaptive Online Proctoring with Neuro-Cognitive Monitoring system is designed to improve the security and reliability of online examinations by using artificial intelligence and computer vision techniques. The proposed framework continuously monitors candidates through a webcam and analyzes their behaviour during the examination without requiring constant human supervision. It detects important events such as face presence, head movement, facial expressions, multiple person detection, and the presence of prohibited objects like mobile phones or books. These behavioural cues are combined to identify suspicious activities and help maintain examination integrity. A Convolutional Neural Network (CNN) is employed for emotion recognition, while computer vision models are used for object detection and head pose analysis. The system provides real-time monitoring and generates alerts whenever unusual behaviour is observed. Experimental evaluation shows that the proposed approach delivers accurate detection with minimal false predictions, making it suitable for practical online assessment environments. Overall, the framework offers an intelligent, efficient, and cost-effective solution for secure remote examinations by reducing manual monitoring effort and enhancing the fairness and credibility of online assessments.