Design and Implementation of an Intelligent WebcamBased Proctoring System for Behavioural Monitoring

Abstract

The growing need for intelligent monitoring systems has led to the development of automated proctoring solutions capable of analyzing human behavior in real time. This project presents a deep learning–based webcam proctoring system designed to detect facial emotions and ensure attentive behavior during online sessions. The system provides a simple interface where users can initiate execution through a batch file, load a pre-trained Convolutional Neural Network (CNN) model, and activate real-time monitoring using a webcam. Once the model is generated and loaded, the system achieves high accuracy (approximately 97%) in recognizing facial expressions. It then enables a webcam-based proctoring module that continuously captures video input and analyzes facial features to identify different emotional states and activities. The system is capable of detecting multiple conditions such as attentiveness, distraction, and various facial expressions, providing visual feedback on the screen. To ensure accurate emotion detection, the system relies on efficient processing speed and precise facial feature extraction. It highlights all possible detections in real time, allowing users to observe system performance under different conditions. The implementation emphasizes the importance of system responsiveness and clear facial visibility for achieving reliable results. Overall, this project offers an effective and automated solution for real-time monitoring using deep learning techniques. It reduces manual supervision, enhances accuracy in behavioral analysis, and demonstrates the practical application of CNN models in intelligent proctoring systems.