Mask Detection Using YOLOv8

This project demonstrates real-time face mask detection using the YOLOv8 model. The model is trained to classify images into three categories:

• With Mask
• Without Mask
• Incorrect Mask

Features

• Accurate Mask Classification: Identifies whether individuals are wearing masks correctly.
• Streamlit App: User-friendly interface for uploading and predicting mask usage in images.
• MASK Detection: Fast and efficient detection, making it suitable for health monitoring.

Workflow

1. Dataset: Images labeled into three classes: with mask, without mask, and incorrect mask.
2. Model Training: YOLOv8 was used for training with custom augmentations.
3. Prediction Results: Example predictions can be found below:

Main APP View

APP View when any sample Image Loaded using uploaded image

When Click Predict then Predicted Image

Upload another image and get the prediction

Upload another image and get a prediction

How to Run the App

1. Install dependencies:

   pip install -r requirements.txt
2. Run the Streamlit app:

   streamlit run app.py
3. Upload an image to get predictions.

Conclusion

This YOLOv8-powered mask detection app ensures high accuracy and speed, making it ideal for real-world deployment in areas like public safety and health monitoring.

Very Important Note

I use the custom function to handle the dataset formats for YOLO. However, I did not upload the custom.py code file due to my code privacy