CoviScan- CNN based COVID-19 Variant Prediction

This native Android app uses a RESNET-18 model, trained using PyTorch, to detect COVID-19 with chest X-ray radiographs. The accuracy of the model was 98.9% on the COVID-QU-Ex Dataset and 73.8% across real world samples collected at the Powai Polyclinic and Hospital, Mumbai. The APK for this app can be downloaded using the link: https://drive.google.com/file/d/1VrsETSk48RTLPMqV0ly9HU3K8hj3MRuw/view?usp=sharing

File Structure

• load_train_model.ipynb: Jupyter Notebook to load the dataset and train and test the 18-layer mobile CNN.
• CovidPredictor: Source code for the deployed Android application, with which the trained model is locally bundled.