Real-time Web-based Face Mask Detection
Contents
This project is built with OpenCV, YOLOv3 and some Computer Vision concepts in order to detect face masks in real-time video streams.
1 - Key Features
- Train YOLO with custom dataset.
- Enabling face mask detection in real-time video streams using Streamlit.
2 - Motivation
In the midst of the persistent COVID-19 pandemic, the need for effective face mask detection applications has surged. However, the scarcity of substantial datasets containing images depicting individuals wearing masks has presented a significant obstacle, intensifying the complexity and difficulty of this undertaking.
3 - Dataset
Kaggle link: link
This dataset contains 853 images belonging to the 3 classes, as well as their bounding boxes in the PASCAL VOC format.
The classes include:
- With mask
- Without mask
- Mask worn incorrectly