search

dongkyunk / CheXFusion

[ ICCV CVAMD 2023] Official implementation of "CheXFusion: Effective Fusion of Multi-View Features using Transformers for Long-Tailed Chest X-Ray Classification" 🏆 1st place solution 🏆 for the ICCV CVAMD 2023 Shared Task: CXR-LT: Multi-Label Long-Tailed Classification on Chest X-Rays.
MIT License
35 stars 3 forks source link

Folder structure configuration #3

Closed oscarloch closed 5 months ago

oscarloch commented 5 months ago

Hello, I attempted to use the code but encountered difficulty in determining the appropriate folder structure for training the model.

  1. What file (or path) do I need to modify to commence training?
  2. What is the appropriate folder structure for training the model? Could you specify the one you utilized?

I already have the mimic-cxr-jpg and chexpert downloaded but currently I can't configure this model correctly.

I'm sorry if my question is too elementary. Thank you so much for your help.

dongkyunk commented 5 months ago

Hi, please refer to chexpert_dataset and cxr_dataset. This code is designed for use of multiple datasets, so you will also have to also edit the datamodule code. About the folder structure, you can refer to the code.