Detecting face masks using deep learning
This project aims at using transfer learning on pretrained model for detecting face with masks on them. The model used is ResNet50 which is available in PyTorch. The approach used for prediction using the trained model is as follows:
- First we find ROI in the image which is face. This is done by passing our image to MTCNN face detector
- Then we crop the face, which is our ROI and pass it to our model to detect if face is covered with mask or not.
- PyTorch
- MTCNN face detector
The dataset used in this project can be found here - https://github.com/prajnasb/observations with some of my own images and their augmented images.
There's always room for improvement. In this project, we can further detect if the faces are fully covered by mask or not. It is beneficial to do so because improper use of mask may decrease it's efficiency.
I would be so happy if you give this repository a star ⭐ or share it with your friends ❤️