search

yuanzhi-zhu / DiffPIR

"Denoising Diffusion Models for Plug-and-Play Image Restoration", Yuanzhi Zhu, Kai Zhang, Jingyun Liang, Jiezhang Cao, Bihan Wen, Radu Timofte, Luc Van Gool.
https://yuanzhi-zhu.github.io/DiffPIR/
MIT License
388 stars 31 forks source link

FFHQ 256×256 and ImageNet 256×256 validation dataset #27

Closed NebiyouTen closed 10 months ago

NebiyouTen commented 10 months ago

Dear Authors,

Thanks for your great work and for sharing your code!

I am currently working on reproducing the results, and I wanted to ask how you sampled the 100 images for evaluation (if possible, which specific samples were used) and how they were pre-processed to obtain the final 256x256 HR images for each dataset.

Thank you.

yuanzhi-zhu commented 10 months ago

Hi @NebiyouTen,

I have just uploaded the names of the validation data at https://github.com/yuanzhi-zhu/DiffPIR/blob/main/testsets/ffhq_val_100/ffhq_val_100.txt and https://github.com/yuanzhi-zhu/DiffPIR/blob/main/testsets/imagenet_val_100/imagenet_val_100.txt.

You can download them from the corresponding 256*256 datasets.

NebiyouTen commented 10 months ago

Thank you.