Skip to content

Building a Face Mask Classifier with CNN and Transfer Learning

Notifications You must be signed in to change notification settings

hayriyeanill/FaceMaskClassifier

Repository files navigation

Artificial Neural Networks Class

FaceMaskClassifier

Covid - 19 is a contagious virus that causes a newly discovered respiratory tract infection and can be passed from person to person. On January 30, 2020, Covid - 19 was declared a global health emergency by the World Health Organization. On March 11, 2020, the virus was declared a pandemic, that is, a global epidemic. It is vital to monitor the prevention of stroke and taking the necessary measures to reduce the contagious impact of the Covid - 19 disease. In this project, a face mask classifier was created using a transfer learning model by fine-tuning CNN and the pre-trained state-of-the-art deep learning model InceptionV3. In order to better train and test the model, the image augmentation technique was adopted to address the limited data availability. According to the results, the transfer learning model developed with CNN with 20 epoch number gave the best result with 0.9990.

Dataset

The dataset contains 2 folders labeled as to which class they belong to the 2 classes are "with_mask", "without_mask". Each folder holds 2994 images of people that belong to such a labeled class

Dataset available on: https://www.kaggle.com/vijaykumar1799/face-mask-detection

Methods

  1. Convolutional Neural Networks
  2. Image Data Generator
  3. Transfer Learning

About

Building a Face Mask Classifier with CNN and Transfer Learning

Topics

Resources

Stars

Watchers

Forks

Releases

No releases published

Packages

No packages published