search

ZhengJun-AI / MoirePhotoRestoration-MCNN

This is an unofficial reproduction of paper Moir´e Photo Restoration Using Multiresolution Convolutional Neural Networks.(PyTorch)
MIT License
91 stars 12 forks source link

Questions about the dataset #8

Closed shallowdream204 closed 1 year ago

shallowdream204 commented 1 year ago

Thanks a lot for sharing the dataset and code. I've noticed that in most paper[1,2,3] introductions, it is mentioned that the TIP2018 dataset contains 135,000 images with a resolution of approximately 400x400. However, the dataset downloaded from the Baidu Cloud link has a smaller number of images than 135,000 and a resolution of 800x800. Is this normal?

  1. Real-Time Image Demoireing on Mobile Devices. ICLR2023.
  2. Towards Efficient and Scale-Robust Ultra-High-Definition Image Demoireing. ECCV2022.
  3. FHDe2Net: Full High Definition Demoireing Network. ECCV2020.
ZhengJun-AI commented 1 year ago

Yes, the TIP2018 dataset doesn't match the description in the paper. You can also see the table shown in https://github.com/ZhengJun-AI/MoirePhotoRestoration-MCNN#dataset

I suppose papers you mentioned directly copy the description from the original paper.

shallowdream204 commented 1 year ago

Thank you for your prompt response. Do we need to crop the central area of the image during testing, or just test with the entire image?

Yes, the TIP2018 dataset doesn't match the description in the paper. You can also see the table shown in https://github.com/ZhengJun-AI/MoirePhotoRestoration-MCNN#dataset

I suppose papers you mentioned directly copy the description from the original paper.

ZhengJun-AI commented 1 year ago

Thank you for your prompt response. Do we need to crop the central area of the image during testing, or just test with the entire image?

Yes, the TIP2018 dataset doesn't match the description in the paper. You can also see the table shown in ZhengJun-AI/MoirePhotoRestoration-MCNN#dataset I suppose papers you mentioned directly copy the description from the original paper.

The authors use centercrop if my memory serves me right... Maybe you should read the original paper carefully, because I do not keey track of demoiring fo a long time.

shallowdream204 commented 1 year ago

Thank you very much for your timely response before the CVPR 2024 deadline. I found the code for cropping at the ECCV 2022 paper. Thanks again.

Thank you for your prompt response. Do we need to crop the central area of the image during testing, or just test with the entire image?

Yes, the TIP2018 dataset doesn't match the description in the paper. You can also see the table shown in ZhengJun-AI/MoirePhotoRestoration-MCNN#dataset I suppose papers you mentioned directly copy the description from the original paper.

The authors use centercrop if my memory serves me right... Maybe you should read the original paper carefully, because I do not keey track of demoiring fo a long time.