Clarification needed on image standardisation before inputing to models

[ff98li]

Dear authors,

Thank you for the great work!

I am currently using path-50x1-remedis-s to extract WSI patch features for weakly supervised classification, following reference code provided in the notebook. However, in the image preprocessing step (e.g. preprocess_for_eval), I did not find code for standardising images based on means and std. Could you please confirm whether the images were standardised during the training/validation phase or were only clipped to the range [0, 1] without standardisation?

Any clarification or insight from you would be greatly appreciated.