Whenever I walk into grocery stores during COVID, I noticed that there is always someone standing there checking if people are wearing masks. It would be great if that could be detected by an AI instead, allowing the person to do other work.
What it does
Detects if there is a person with no mask walking into a store. If there is, then send an email notification.
How I built it
I trained a model using a mask/no mask dataset found on Kaggle. This model can identify if a person is wearing a mask or not from an image. Then I used opencv to read in webcam data into a PyQt GUI.
Challenges I ran into
The biggest hurdle was training a model that would be sufficient, but also quick. I initially created another model, but it was taking far too long to produce results, which wouldn't be useful for our use case since this is real time.
Accomplishments that I'm proud of
Been to a few hackathons before, I'm just happy to have completed something so quick!
What I learned
I learned a lot more about using PyTorch, and learned a few techniques for PyQt to act like there's a video playing, even though it's just a few images showing up.
What's next for Face Mask Notifier
There's a lot I can do with this project. Currently it can only identify one person, it would be beneficial if it could do more than one person. Also I can have this count number of people walking in.