Is this method applicable to the self-supervised denoising task of real-world noise images?

[wam730]

Dear authors, I sincerely appreciate your outstanding work! Upon reading your paper, I noticed that your method has thus far been exclusively applied to the self-supervised denoising tasks using synthetic noise datasets. My questions are as follows:
1.Is your method applicable to real-world noisy images?
2.Additionally, I would like to inquire whether your method is restricted to U-shaped networks only?
Looking forward to your reply very much!

[liuhanze623]

Thank you very much for your interest in our work!
To address your questions:

1. Yes, our method is indeed applicable to real-world noisy images. We have included relevant experimental results demonstrating its effectiveness in real-world scenarios in the supplementary material of our paper.
2. Our method is not limited to U-shaped networks. The rotation-equivariant convolutions we employ are designed to be plug-and-play, meaning they can replace vanilla convolutions in a variety of architectures to enhance performance.
   Please feel free to reach out if you have any further questions or thoughts!