face-mask detection and Warning System

Facemask detection program and pre-trained weights!

This is my first project on the live object detection, so please feel free to suggest changes where necessary

The code was made for the Colab environment so when running offline make appropriate changes.

• Project inspired from : perceptron
• Dataset from : Ohyicong
• For full augmented dataset visit my drive

Note : The model uses a CNN architecture so errors due to outside noise may cause variation from expected output.

Edit : As part of a hackathon we decided to modify the program a little further to include face recogninition and messaging functions. This project is now a joined effort of Vyshak T ,Azeem M Basheer , Kallu Sudarshan and Martin Joe C.

Program Structure

1. Load the pretrained model
2. Find the ROI with face using haarcascade
3. Detect if mask present or not
4. If mask present display "Mask is present".
5. If mask not present , check for the persons identity using facial recognition library
6. Using messaging script, send a whatsapp warning
7. display the video footage with bounting box and labels around detected faces