The network architecture is ResNet18.
Dataset from COVID-19 Radiography Dataset on Kaggle
The model is trained for 1 epoch. After even 1 epochs, calculated accuracy is about 0.989.
Loss function for the training is basically just a binary crossentropy. The optimizer is Adam.
This tutorial depends on the following libraries:
- Pytorch
- torchvision
- numpy
- PIL
- matplotlib
Follow the notebook to run the code.
Use the trained model to do classification on test images.
