pneumonia-detection-pytorch

a pytorch implementation of kagge pneumonia competition

DataSet

Data are installed from kaggle competition 'RSNA Pneumonia Detection Challenge'

3-class classification
- lung opacity
- normal
- not normal / no lung opacity
trian with densenet121
- fine-tuning classifier
Grad_CAM unsupervised learning
- improve explainability
- thresholding heatmap to output bbox

Name		Name	Last commit message	Last commit date
Latest commit History 11 Commits
pictures		pictures
README.md		README.md
densenet_cam.py		densenet_cam.py
fix.py		fix.py
main.py		main.py
pneumonia.py		pneumonia.py
pred_class.py		pred_class.py
predict.py		predict.py
test.py		test.py
thresholding.py		thresholding.py
train.py		train.py