It was a guided project in coursera where I tried to detect COVID-19 with Chest X-Ray pictures using PyTorch where a convolutional neural network (CNN) model named ResNet-18 has been used to train the model to create an image classifier to classify in between three classes which are "Normal", "Viral Pneumonia Affected" and "COVID-19 Affected" with testing accuracy more than 98%.
Dataset: https://www.kaggle.com/tawsifurrahman/covid19-radiography-database
Though the dataset was pretty small as COVID-19 came recently but it can be improvised more if the dataset gets enriched. And as the dataset is small, it can be used for learning purpose for now.