Dataset Filename Mismatch with data_reorganize.csv

[Nech-C]

Hi Dr. Ji,

Thank you for this valuable project. I am trying to build the ColonINST dataset, but I encountered an issue where the filenames in the dataset do not align with the filenames referenced in the data_reorganize.csv for the polyp section of the Kvasir(Kvasir-V2).

Specifically, the CSV expects files with cju-prefixed names (e.g., cju0qkwl35piu0993l0dewei2.png) for images under kvasir-dataset-v2/polyps, but the dataset contains filenames in a UUID format (e.g., fd49618e-96fc-45ec-a917-9d04421faa3f.jpg). All other categories from Kvasir-V2 have valid paths in the data_reorganize.csv file. As a result, about 1000 polyp images are missing.

I downloaded the dataset through the link provided by this repo and also tried other sources, but none of them have polyp images with the expected filenames.

Could you clarify how the cju-prefixed filenames are derived? I appreciate any help you can give!