Datasets used in ArtiFusion

[dmhdmhdmh]

Hello！I'm sorry to bother you. As you mentioned in the article: "To evaluate the performance of artifact restoration, a training set is curated from a subset of Camelyon17. It comprises a total number of 2445 artifact-free images and another 2547 images with artifacts, where all histological images are scaled to the resolution of 256 × 256 pixels at the magnitude of 20×." Could you tell me which subset of Camelyon17 was used? I was wondering if you could kindly share the dataset with me (a total number of 2445 artifact-free images and another 2547 images with artifacts, where all histological images are scaled to the resolution of 256 × 256 pixels). I promise that the use of this dataset is only for academic research. And I will be very glad to cite your work in my paper in the future.

[yiqings]

Hi, the dataset was a manually curated subset in previous work [1], and please refer to the downloading link (OneDrive / Baidu Cloud) in the associated repo [2].

[1] Ke, Jing, et al. "Artifact Detection and Restoration in Histology Images with Stain-Style and Structural Preservation." IEEE Transactions on Medical Imaging (2023).
[2] https://github.com/yunboer/AR-classifier-and-AR-CycleGAN#datasets

[kpbhat25]

Can you please upload the link to the masks you have created ?
gt_masks during the inference time