资源介绍

We introduce COVIDx, an open access benchmark dataset that we generated comprising of 13,975 CXR images across 13,870 patient patient cases, with the largest number of publicly available COVID-19 positive cases to the best of the authors' knowledge.

Instruction

• Requirements to install on your system
• How to generate COVIDx dataset
• Steps for training, evaluation and inferenceof COVIDNet
• Steps for inferenceof COVIDNet lung severity scoring
• Results
• links to pretrained models

Citation

Please use the following citation when referencing the dataset:

@misc{wang2020covidnet,
    title={COVID-Net: A Tailored Deep Convolutional Neural Network Design for Detection of
COVID-19 Cases from Chest Radiography Images},
    author={Linda Wang, Zhong Qiu Lin and Alexander Wong},
    year={2020},
    eprint={2003.09871},
    archivePrefix={arXiv},
    primaryClass={cs.CV}
}