Can be applied to the field of image repair/denoising？

[Breeze-Zero]

Thank you very much for your work. This lightweight network structure is really great, but when I actually use my own data, I find that its training loss does not decrease much, and the output visual image shows that the model has been repaired to some extent, but there are still some deficiencies in texture. In addition, I hope it can be applied to the field of repair or noise reduction. I wonder if you have done relevant experiments, and I hope you can tell me whether it is feasible

[SkyeLu]

Hi, if you are training on your own dataset, I think it is better that the reference image and the gt-HR image look similar, in this way, the training difficulty can be reduced. For example, the traing set we use is CUFED5, whose reference image and HR image are from the same video clip, they look very similar.
As for the denoising task, we haven't done experiments on it. I think it is possible to transfer structures or texture details from reference images to oversmooth denoised results via such reference-based method.