Face Mask Detection Streamlit App 😷

A modern, interactive web application for detecting face masks in images using deep learning with VGG19 architecture.

🚀 Features

• Real-time Mask Detection: Upload images and get instant predictions
• Interactive UI: Beautiful, modern interface with animations and visual feedback
• Confidence Scoring: See how confident the model is in its predictions
• Multiple Pages: Home, Detection, Model Info, and About sections
• Responsive Design: Works on desktop and mobile devices
• Visual Analytics: Interactive charts and gauges for prediction confidence

🛠️ Technology Stack

• Frontend: Streamlit with custom CSS styling
• Backend: TensorFlow/Keras
• Model: VGG19 with Transfer Learning
• Image Processing: OpenCV, PIL
• Visualization: Plotly
• UI Components: Streamlit Option Menu

📋 Prerequisites

• Python 3.8 or higher
• Virtual environment (recommended)
• At least 4GB RAM (for TensorFlow)
• vgg19_face_mask_model.keras model file ⚠️ See MODEL_DOWNLOAD.md

🔧 Installation & Setup

Step 1: Clone or Download

Make sure you have all the project files in your directory:

• app.py (main Streamlit application)
• requirements.txt (dependencies)
• vgg19_face_mask_model.keras (trained model) ⚠️ Download Required
• README.md (this file)

Step 2: Create Virtual Environment

# Navigate to project directory
cd "c:\Users\Ziad mahmoud\Downloads\MaskDetection project"

# Create virtual environment
py -m venv mask_detection_env

# Activate virtual environment
# On Windows:
mask_detection_env\Scripts\activate

# On macOS/Linux:
# source mask_detection_env/bin/activate

Step 3: Install Dependencies

# Make sure your virtual environment is activated
pip install -r requirements.txt

Step 4: Run the Application

streamlit run app.py

The app will automatically open in your default web browser at http://localhost:8501

📱 How to Use

1. Launch the App: Run the command above and navigate to the provided URL
2. Navigate: Use the horizontal menu to switch between different sections
3. Upload Image: Go to "Detection" tab and upload an image with a face
4. Get Prediction: Click "Detect Mask" to analyze the image
5. View Results: See the prediction with confidence score and visual feedback

🎯 Usage Tips

• Image Quality: Use clear, well-lit images for best results
• Face Visibility: Ensure the face is clearly visible and not too small
• Supported Formats: JPG, JPEG, PNG, BMP, TIFF
• Single Face: Works best with one face per image

📊 Model Information

• Architecture: VGG19 with custom classifier layers
• Training: Transfer learning from ImageNet
• Input Size: 224 × 224 × 3 pixels
• Output: Binary classification (Mask/No Mask)
• Accuracy: ~93-95% on validation data

🐛 Troubleshooting

Common Issues:

1. Model Not Found

   Error: Model could not be loaded
   

   • Ensure vgg19_face_mask_model.keras is in the same directory as app.py

2. Import Errors

   ModuleNotFoundError: No module named 'streamlit'
   

   • Make sure virtual environment is activated
   • Run pip install -r requirements.txt

3. Memory Issues

   ResourceExhaustedError: OOM when allocating tensor
   

   • Close other applications to free up RAM
   • Use smaller images or restart the app

4. Port Already in Use

   Port 8501 is already in use
   

   • Use a different port: streamlit run app.py --server.port 8502
   • Or kill the existing process

🔧 Configuration

You can customize the app by modifying these parameters in app.py:

• Image Size: Change (224, 224) in the preprocess_image function
• Confidence Threshold: Modify 0.5 in the predict_mask function
• Styling: Update the CSS in the st.markdown section
• Colors: Change color schemes in the CSS classes

📂 Project Structure

MaskDetection project/
├── app.py                          # Main Streamlit application
├── requirements.txt                # Python dependencies
├── vgg19_face_mask_model.keras    # Trained model file
├── README.md                       # Project documentation
├── mask_detection_env/             # Virtual environment (created after setup)
└── VGG19_MaskDetection_ZiadMahmoudElShazly.ipynb  # Training notebook

🤝 Contributing

Feel free to contribute to this project by:

• Reporting bugs
• Suggesting new features
• Improving the UI/UX
• Optimizing model performance

📄 License

This project is for educational and research purposes.

👨‍💻 Developer

Ziad Mahmoud ElShazly Email: ziad.m.elshazly@gmail.com

Built with ❤️ using Streamlit, TensorFlow, and Python

🔄 Version History

• v1.0: Initial release with basic mask detection
• v1.1: Added interactive UI and confidence scoring
• v1.2: Enhanced styling and multiple page navigation