fengwang
Request for instructions
Hello,
I would very much like to test this denoiser for my own datasets.
Unfortunately, I find it difficult to run Noise2Atom. What exactly is one required to do in order to get it to run on their own data.
If understand it correctly, one hast to first simulate the gaussian-like noise data. To what extent is that specific to ones own data?
Next, it says to the configuration, but what exactly is required to run it.
I have several movies (tiff files with stacks of frames) and would like to subject them to Noise2Atom.
Could you provide a little bit more of a required workflow for Noise2Atom and maybe an example that demonstrates running Noise2Atom covering the different required steps. So far I am unsure how to apply Noise2Atom.
Thank you very much for your help.
Best
Dennis
Sorry for my late reply. But could you please take a look at this paper: https://arxiv.org/abs/2110.11911 If you have massive images, you can try this self-supervised denoising without simulating the ground truth images.
Okay, I will have a look.
But still, could you provide a rough instruction/doc/readme on how to actually use noise2atom. I mean it appears to perform quite well, so I maybe one can make it a bit more accessible by providing some instructions (creating the required environment, input data required for training, training, and actually denoising with a trained model, or of the like). This I believe would be very helpful and much appreciated by users.
Having read the suggested paper of you, I am eager to test the framework on my own datasets. Unfortunately, I cannot find any reference to the implementation of the denoising software introduced in said paper. Am I missing something or where can one find the source code?
Okay, I will have a look.
But still, could you provide a rough instruction/doc/readme on how to actually use noise2atom. I mean it appears to perform quite well, so I maybe one can make it a bit more accessible by providing some instructions (creating the required environment, input data required for training, training, and actually denoising with a trained model, or of the like). This I believe would be very helpful and much appreciated by users.
Having read the suggested paper of you, I am eager to test the framework on my own datasets. Unfortunately, I cannot find any reference to the implementation of the denoising software introduced in said paper. Am I missing something or where can one find the source code?
1.) I will prepare a quick readme file for noise2atom. 2.) Please find a reference implementation from this file: https://github.com/fengwang/Noisy_Predestination/blob/main/train_7.py Let me know if you need help.
1.) I will prepare a quick readme file for noise2atom.
That would be great!
2.) Please find a reference implementation from this file: https://github.com/fengwang/Noisy_Predestination/blob/main/train_7.py Let me know if you need help.
Unfortunately, the link is not functional (Page not found - 404). If you could also provide a short instruction readme for the implementation that accompanies the work published in this paper: "Self-supervised denoising for massive noisy images"
1.) I will prepare a quick readme file for noise2atom.
That would be great!
2.) Please find a reference implementation from this file: https://github.com/fengwang/Noisy_Predestination/blob/main/train_7.py Let me know if you need help.
Unfortunately, the link is not functional (Page not found - 404). If you could also provide a short instruction readme for the implementation that accompanies the work published in this paper: "Self-supervised denoising for massive noisy images"
Would you please try the link again? It was a private project, and I have made it public.