AI/ML

ExamX Protocol - AI Exam Integrity System (Proof of Concept)

ExamX Protocol is a cutting-edge examination system that leverages advanced AI technologies to ensure academic integrity during online exams. The system combines computer vision with natural language processing to monitor student behavior in real-time, detecting potential cheating attempts through gaze tracking and speech analysis. Built entirely in Python, it represents a breakthrough in educational technology.

Sep 2025
University Malaysia Perlis (UniMAP)
3 months

Project Screenshots

Project Overview

ExamX Protocol is a cutting-edge examination system that leverages advanced AI technologies to ensure academic integrity during online exams. The system combines computer vision with natural language processing to monitor student behavior in real-time, detecting potential cheating attempts through gaze tracking and speech analysis. Built entirely in Python, it represents a breakthrough in educational technology.

The Challenge

Creating an accurate AI system that could detect cheating behaviors without false positives, while maintaining smooth exam experience and real-time processing capabilities.

The Solution

Developed custom TensorFlow models for gaze detection, integrated OpenAI Whisper for speech recognition, and created a proprietary algorithm for integrity scoring. Used PyQt6 for the exam interface and Firebase for real-time data storage.

Key Features

  • Real-time gaze tracking and analysis
  • Speech detection and conversation monitoring
  • Custom integrity percentage calculation algorithm
  • PyQt6-based secure exam interface
  • Firebase integration for data storage
  • Automatic Google Sheets reporting
  • Fine-tuned with real UniMAP student data
  • 100% Python implementation

Project Statistics

3+
Duration (months)
9+
Technologies
AI/ML
Category

Interested in a Similar Project?

I'm always excited to work on new challenges and bring innovative solutions to life. Let's discuss how we can work together!