Covid times have been tough. I wanted to ideally make a checking system for door way entries in which people would be allowed only when they have their masks on. For prototyping purposes I have a "Face Mask Detection" for identifying if the Mask is put on a face or not so as to contain virus spread in COVID times.
What it does
Currently it detects if a person is wearing a mask or not.
How I built it
I've used p5.js for making it on the web and use TensorFlow js (ml5.js) to classify. I've used my own images for model training and using transfer learning I was able to classify the things.
Challenges I ran into
Apart from hardware and network issues, I was stuck at hosting the model for real time webcam interface on a site.
What's next for Mask detector
A door lock system, which opens only when the camera detects that user is wearing a mask.