🧠 Real-Time Mask Detection & Face Recognition System

---

📌 About the Project

A high-performance AI system combining Face Mask Detection and Face Recognition in real-time, designed to demonstrate production-grade computer vision expertise. The project uses:

• ✅ MobileNetV2 (PyTorch) for Mask Classification
• ✅ MTCNN (facenet-pytorch) for Deep Learning–based Face Detection
• ✅ DeepFace (ArcFace / Facenet backend) for Face Recognition
• ✅ Streamlit for Web Interface
• ✅ OpenCV for Real-time Webcam Integration

This system can:

• Detect whether a person is wearing a mask, not wearing a mask, or wearing incorrectly.
• Recognize known faces using a pretrained face recognition model.
• Run as a local app (Streamlit) or be extended as a backend API (FastAPI recommended).

---

🚀 Features

• 🎥 Real-time Mask Detection via Webcam (OpenCV + PyTorch)
• 🖼️ Upload Image Interface via Streamlit
• 👤 Deep Learning–based Face Detection (MTCNN)
• 🧑‍💼 Face Recognition using Pretrained Models (DeepFace)
• ⚡ GPU Acceleration Support
• 📈 High Accuracy & Low Latency

---

📁 Project Structure

project/
├── mask_detector.pt             # Trained MobileNetV2 model
├── face_db/                     # Folder containing known faces (e.g. Alice.jpg)
├── app.py                       # Streamlit app for image-based usage
├── realtime.py                  # OpenCV real-time webcam application
├── requirements.txt             # Python dependencies
└── README.md                    # Project documentation

---

⚙️ Setup Instructions

1. Clone the Repository:

git clone <repository_url>
cd project

2. Install Dependencies:

pip install -r requirements.txt

3. Place Known Faces:

• Add clear face images in the face_db/ folder.

4. Run Streamlit Web App:

streamlit run app.py

5. Run Real-time Webcam Application:

python realtime.py

---

🛠️ Requirements

• torch
• torchvision
• facenet-pytorch
• opencv-python
• opencv-python-headless
• numpy
• pillow
• matplotlib
• seaborn
• scikit-learn
• streamlit
• deepface
• tensorflow

---

🏆 Deployment Suggestions

• Use Streamlit for browser-based interface (local users).
• Wrap as FastAPI backend for production-ready API serving.
• Optionally Dockerize for easy deployment.

---

📖 How It Works

• MTCNN detects faces using a deep learning face detector.
• MobileNetV2 classifies each detected face as Mask / No Mask / Incorrect.
• DeepFace compares detected faces against your face database (face_db/) to recognize known identities.
• Results are drawn on images or real-time webcam frames.

---

📷 Example Output

• ✅ Mask Detection: With_Mask (97.4%)
• ✅ Face ID: Alice

---

📚 Future Enhancements

• Dockerfile for containerized deployment.
• FastAPI backend with REST endpoints.
• Hugging Face ViT backbone for further optimization.
• CI/CD Pipeline for automatic builds.

---

👤 Author

Syed Taha Jameel
AI Engineer | Computer Vision Specialist

---

📄 License

This project is licensed for educational and demonstration purposes.