flyywh/image-denoising-state-of-the-art

image-denoising-state-of-the-art's Issues

Redirecting to https://github.com/wenbihan/reproducible-image-denoising-state-of-the-art

Since https://github.com/wenbihan/reproducible-image-denoising-state-of-the-art has a better record on all the new denoising techniques, why not?

wrong PDF link

wrong PDF link for
Non-Local Recurrent Network

Information of ReNOIR dataset can be updated

ReNOIR: Journal of Visual Communication and Image Representation 51, No. 2, 144-154, 2018

Official Pytorch code for CycleISP (CVPR 2020--Oral)

Hi @flyywh,

Could you please add the following denoising method in your list: https://github.com/swz30/CycleISP

Thank you.

How about "Deep Image Prior"?

Deep Image Prior

Abstract
Deep convolutional networks have become a popular tool for image generation and restoration. Generally, their excellent performance is imputed to their ability to learn realistic image priors from a large number of example images. In this paper, we show that, on the contrary, the structure of a generator network is sufficient to capture a great deal of low-level image statistics prior to any learning. In order to do so, we show that a randomly-initialized neural network can be used as a handcrafted prior with excellent results in standard inverse problems such as denoising, super-resolution, and inpainting. Furthermore, the same prior can be used to invert deep neural representations to diagnose them, and to restore images based on flash-no flash input pairs.

PolyU data set should be listed under benchmarks

Link for the arxiv:
https://arxiv.org/pdf/1804.02603.pdf

Link to the data set:
https://github.com/csjunxu/PolyU-Real-World-Noisy-Images-Dataset

Also another paper targeting real-world image denoising:

https://arxiv.org/pdf/1807.04686.pdf
https://github.com/GuoShi28/CBDNet

Which papers have code or trained models, does the author know?