Question about the Table 1. CAM evaluation.

[Shield9999]

First of all, thanks for the nice work! I have a question about your code. I hope the Table 1, seed and dCRF means respectively the initial+CAA and initial+CAA+dCRF CAM. So when I try to run the code eval_cam.py and eval_cam_with_crf.py, the performance results were respectively 65.96%, 68.76%. It has lot of gap between the paper. However, the final performance with that CAM was similar to the paper's result.

Thank you for reading!

[linyq2117]

Thanks for your interest in our work.

It seems that you evaluate the performance of CAM on the train_aug set (10582 images in total). However, the results in Table 1 are based on the train set (1464 images in total). You can change the --split_file to reproduce our results in Table 1. We have also changed it in README.md to avoid confusion.

[Shield9999]

Oh that's the reason! Sorry for my mistake, it was train_aug file. Thanks for the comment.

[henry189]

How to visualize refine cam?

[linyq2117]

How to visualize refine cam?

To visualize cam, you can use from pytorch_grad_cam.utils.image import show_cam_on_image following instructions in pytorch_grad_cam.