HIPT

HIPT copied to clipboard

AymenSe

HI, thanks for your great work. I'm currently using your pretrained dino network backbone (ViT-S/16) to extract patch (256x256) features. What mean and std should I use? ImageNet mean and...

weiaicunzai

Hi People, First, thanks for your astonishing work! I am currently trying out how HIPT performs in a Grading task and recognised, performance is underwhelming using the features I extracted...

mbu93

Gradient accumulation not properly implemented

2

Hi, based on the following lines, it seems gradient accumulation is not properly implemented: https://github.com/mahmoodlab/HIPT/blob/a9b5bb8d159684fc4c2c497d68950ab915caeb7e/2-Weakly-Supervised-Subtyping/utils/core_utils.py#L285-L290 A proper implementation should look like the following: ```python loss = loss / gc loss.backward()...

clemsgrs

Hello I followed the command in the attention visualization .ipynb for 4096 individual (not concatenated). I am noticing that on the raw image there are some discolorations but it is...

matthew-lee1

Hi there, I just want to know how can I get the pertaining dataset used for this article? Cheers

aCoalBall

Thanks for releasing the codes and the data. I noticed that there are several WSIs in each patient in TCGA. So I need to extract patches from all WSIs and...

WinterPan2017

Thank you for this great work! Would it be possible to make the coordinates of the pre-extracted "region-level" feature embeddings (3-Self-Supervised-Eval/embeddings_slide_lib/embeddings_slide_lib/vit256mean_tcga_slide_embedding) available?

florianbuettner

If I wanted to extract a WSI image with a higher resolution, for example: [150000 x 150000] patch tokens encoded using [4096 x 4096] patch tokens, what should i do?

yuzhuogu

hi, thanks for the interesting paper! I am trying to replicate the results for survival prediction but cannot find the .csv file (not splits, the dataset csv itself) for colon...

umarikkar