🎯 ExamWatch: Cheating Detection using YOLO11m Pose Estimation

ExamWatch is an AI-powered system that detects suspicious student behavior such as looking sideways or potential cheating during classroom or online examinations. Using YOLO11m Pose Estimation, it tracks head pose angles (yaw and pitch) to distinguish between focused and suspicious sideways glances, ensuring exam integrity through computer vision and deep learning.

---

🧠 Features

• 🎥 Real-time detection of student head movement and posture
• 🧍 Tracks yaw (left–right) and pitch (up–down) angles
• 🕵️ Detects focused and suspicious head orientations
• 📊 Produces annotated video output with visual alerts
• ⚡ Built with YOLO11m Pose, OpenCV, and cvzone

---

🧰 Tech Stack

Component	Description
Model	YOLO11m Pose (Ultralytics)
Framework	Python
Libraries	OpenCV, cvzone, NumPy, math
Environment	Google Colab / VS Code

---

⚙️ Installation

1️⃣ Clone the repository

git clone https://github.com/SyedaEmanSaleem/ExamWatch.git
cd ExamWatch

2️⃣ Install dependencies

pip install ultralytics opencv-python cvzone numpy

3️⃣ Download YOLO11m Pose Model

from ultralytics import YOLO
model = YOLO('yolo11m-pose.pt')

4️⃣ Run the project

python examwatch.py

---

🧩 How It Works

1. The YOLO11m Pose model detects students and extracts body keypoints (nose, eyes, shoulders).
2. The system calculates yaw (left–right) and pitch (up–down) angles for each detected student.
3. Based on these angles, it classifies head movement as:

Classification	Behavior	Indicator
🟢 Focused	Head facing forward/downward (normal behavior)	✅ Green
🔴 Looking Sideways	Head turned beyond threshold (potential cheating)	⚠️ Red

🎥 The annotated video output displays detections in real-time.

---

🖼️ Example Output

Focused	Looking Sideways
✅ Green	⚠️ Red

---

📦 Requirements

• Python ≥ 3.8
• OpenCV
• cvzone
• ultralytics
• numpy

---

🚀 Future Enhancements

• 👁️ Add eye gaze estimation for higher accuracy
• 🧠 Apply temporal filters to smooth pose variations
• 🌐 Enable real-time webcam monitoring
• 📈 Develop a dashboard for live analytics

---

🧑‍💻 Author

Syeda Eman Saleem GitHub • LinkedIn

---

🏷️ License

Released under the MIT License — free to use, modify, and distribute.

---

🔖 Hashtags

#AIProjects #ComputerVision #YOLO11 #DeepLearning #PoseEstimation #ExamProctoring #CheatingDetection #MachineLearning #OpenCV #Ultralytics #Python #ExamWatch