The Smart AI Proctoring System is a real-time computer vision-based solution designed to monitor candidates during online exams or interviews and detect suspicious behavior automatically.
With the rise of remote assessments, ensuring integrity has become a major challenge. This project addresses that by leveraging AI and computer vision techniques to analyze live webcam feeds and identify anomalies such as multiple faces, absence, and abnormal movement patterns.
Online exams and remote interviews lack reliable supervision, making them vulnerable to:
- Cheating using external assistance
- Multiple people appearing in the frame
- Candidate leaving the screen
- Unmonitored suspicious activities
This project aims to provide an automated proctoring solution to enhance fairness and integrity.
The system uses real-time video processing to:
- Capture frames from webcam
- Detect faces using OpenCV
- Track presence and movement
- Identify suspicious scenarios
- Trigger alerts or flags
- 👤 Face Detection – Detects whether a candidate is present
- 👥 Multiple Person Detection – Flags if more than one person appears
- 🚫 Absence Detection – Detects if candidate leaves the frame
- 🎥 Real-Time Monitoring – Continuous webcam processing
⚠️ Suspicious Activity Detection – Identifies abnormal behavior- ⚡ Lightweight & Efficient – Optimized for real-time use
- Python
- OpenCV (Computer Vision)
- MediaPipe (Facial tracking & landmarks)
- NumPy
- Matplotlib
smart-ai-proctoring-system/
│── notebook.ipynb # Development & experimentation
│── main.py # Main executable script
│── requirements.txt # Project dependencies
│── README.md # Documentation
git clone https://github.com/AayushCodes-28/Smart-AI-proctoring-System.git
cd smart-ai-proctoring-system
pip install -r requirements.txt
python main.py
- Captures video input using webcam
- Processes frames using OpenCV
- Applies detection algorithms
- Tracks face count and presence
- Flags anomalies in real time
- 🧑🎓 Online examinations
- 💼 Remote job interviews
- 📜 Certification platforms
- 🏫 E-learning systems
- 👀 Eye gaze tracking for attention monitoring
- 🧑🤝🧑 Face recognition for identity verification
- 🔊 Audio-based anomaly detection
- 🌐 Web dashboard for live monitoring
- 🤖 Deep learning model integration (CNN/LSTM)
- Real-world AI application
- Combines Computer Vision + Behavioral Analysis
- Inspired by industry-level AI proctoring systems
- Scalable for production-level deployment
Aayush Kumar Singh
- B.Tech CSE @ Manipal Institute of Technology
- AI/ML Enthusiast
If you found this project useful, consider giving it a ⭐ on GitHub!