Automatic verification of protective masks on the face.
The project was designed to automatically detect faces in pictures/videos and verify that they have a properly applied mask.
I trained the model on publicly available datasets
- Python 3.72
To run this project
$ git clone https://github.com/jedrzejd/Mask-Detector.git
$ cd Mask-Detector
$ python3 -m venv venv
$ . venv/bin/activate
$ pip install -r requirements.txt
$ python app.py
-
Now
- The app recognizes a face and draws a rectangle with the probability that the person has a correctly fitted mask.
-
In Future
- Added the ability to use multiple video sources.
- Improving the accuracy of the model (currently 95% accuracy on the training set)
- Greater generalization of the model, for example: for people with glasses, different skin color