😷 Face Mask Detector – Deep Learning Project

This is a deep learning-based Face Mask Detector built using Convolutional Neural Networks (CNN) in TensorFlow/Keras.
It classifies whether a person is wearing a mask or not using image data.

Trained on over 7,500 images with GPU acceleration using Google Colab, this model achieved 93.9% validation accuracy.

---

📂 Dataset

• ✅ with_mask/ – 3,725 images
• ✅ without_mask/ – 3,828 images
• Total: 7,553 images
• Format: .jpg images with two class folders

---

🧠 Model Overview

• Custom CNN with:
  • 3 Convolutional Layers
  • Batch Normalization
  • Max Pooling
  • Dropout Regularization
• Optimizer: Adam
• Loss Function: Categorical Crossentropy
• Accuracy Achieved: 93.9% on validation set
• Model File: high_accuracy_mask_detector.h5

---

📈 Features

• Data Augmentation with ImageDataGenerator
• Train/Test Split using Scikit-Learn
• EarlyStopping to prevent overfitting
• Normalized input data
• One-hot label encoding
• Visualized training/validation accuracy
• Exported trained model for deployment

---

🔍 Output Screenshot

Final Training Result (93.9% Validation Accuracy)

---

⚙️ Libraries Used

• TensorFlow / Keras
• OpenCV
• NumPy
• Matplotlib
• Scikit-learn

---

🚀 How to Use

1. Clone the repository
2. Load the dataset with with_mask/ and without_mask/ folders
3. Run the notebook: face_mask_detector.ipynb in Google Colab
4. Trained model will be saved as high_accuracy_mask_detector.h5
5. You can deploy it using:
   • Streamlit (web app)
   • OpenCV (real-time webcam detection)
   • Flask/Gradio (REST API or interface)

---

🌍 Future Work

• Real-time detection with webcam
• Deploy on Streamlit or Hugging Face Spaces
• Convert model to TFLite for mobile apps
• Integrate into security systems or thermal cameras

---

✨ Author

Muhammad Rayan Shahid
Founder of ByteBrilliance AI
💼 Future AI Engineer | 🇵🇰 Pakistan | 🧠 AI Enthusiast

---

🏷️ Tags

#DeepLearning #CNN #FaceMaskDetection #AI #TensorFlow #Keras #ComputerVision #ByteBrillianceAI

---