Skip to content

Face mask detection project developed using CNN to ensure people wear masks.

Notifications You must be signed in to change notification settings

devanshukd04/Face-mask-Detection

Repository files navigation

Face-mask-Detection

Inspiration

  • COVID-19 is an infectious disease caused by the SARS-CoV2 virus, but anyone can become seriously ill from COVID-19, regardless of their age.
  • When an infected person coughs, sneezes, speaks, sings, or breathes, small liquid particles spread the virus.
  • Having a mask on is an effective way to reduce the spread of the COVID-19 virus. This project will assist local authorities in monitoring public areas for public safety.

What it does

  • The task of ensuring everyone wears a facemask is not an easy one.
  • The application recognizes human faces and determines whether they are wearing face masks or not.
  • Face recognition and mask detection is performed in real-time in this application.

How we built it

  • It contains 10000 images for training and 400 images for validation taken from the Kaggel website
  • As the dataset used consists of images with different colors, sizes, and orientations, the images are converted to grayscale and normalized.
  • Data is augmented to provide robustness for the model and to avoid overfitting.
  • A CNN model is then built over the processed data. It is found that the model is 97.5% accurate
  • Using openCV, the model was tested in real-time and deployed over the web.

About

Face mask detection project developed using CNN to ensure people wear masks.

Resources

Stars

Watchers

Forks

Releases

No releases published

Packages

No packages published