**Update 08/11/2020: Released new dataset with over 14000 CXR images containing 473 COVID-19 train samples. Train and Test data are kept the same as provided bt this script https://github.com/lindawangg/COVID-Net/blob/master/docs/COVIDx.md
If you are using this dataset, please cite this paper; https://arxiv.org/abs/2010.06418
- Please download the original COVIDx dataset by following https://github.com/ieee8023/covid-chestxray-dataset.
- Save the Train and Test data under folders train and test
- Run the test.py file which uses our pre-trained weighits to generate masks for COVIDx dataset using a pre-trained unet model
- Make sure to change the file paths in test.py in accordance to your directory structure
The current COVIDx dataset is constructed by the following open source chest radiography datasets:
- https://github.com/ieee8023/covid-chestxray-dataset
- https://github.com/agchung/Figure1-COVID-chestxray-dataset
- https://github.com/agchung/Actualmed-COVID-chestxray-dataset
- https://www.kaggle.com/tawsifurrahman/covid19-radiography-database
- https://www.kaggle.com/c/rsna-pneumonia-detection-challenge (which came from: https://nihcc.app.box.com/v/ChestXray-NIHCC )