search

DIAGNijmegen / pathology-he-auto-augment

H&E tailored Randaugment: automatic data augmentation policy selection for H&E-stained histopathology.
Apache License 2.0
51 stars 7 forks source link

Great paper and great repo but why tensorflow #3

Closed TumVink closed 1 year ago

TumVink commented 1 year ago

As title ;(

KhrystynaFaryna commented 1 year ago

Those were the days, my friend:) The randaugment-related stuff and well as augmentations are independent of a framework. You can just keep everything augmentation related, use a pytorch dataloader and apply distort_image_with_randaugment(image, m, n, ra_type) function to each image/patch within the dataloader:)