Hello, I would like to know if the noise is not zero-mean, if only part of the code needs to be modified to converge or the method can only be run under the assumption of zero-mean noise;

Hello,I would like to ask,what is the matrix c below denote?

It would be appreciate if you could explain the meaning of std below：

HI！If I want to train on our own dataset，and how should I prepare data， such as ‘Scene_Instances.txt’

PS E:\githubProjects\CVF-SID_PyTorch\src> python test.py --device 0 --config ../models/CVF_SID/SIDD_Val/config.json --resume ../models/CVF_SID/SIDD_Val/model_best.pth
Traceback (most recent call last):
File "test.py", line 11, in
from torch.distributed.optim import DistributedOptimizer
File "D:\python3.8\lib\site-packages\torch\distributed\optim_init_.py", line 8, in
from .optimizer import DistributedOptimizer
File "D:\python3.8\lib\site-packages\torch\distributed\optim\optimizer.py", line 8, in
from torch.distributed.rpc import RRef
ImportError: cannot import name 'RRef' from 'torch.distributed.rpc' (D:\python3.8\lib\site-packages\torch\distributed\rpc_init_.py)

Hello, I would like to ask, why it is mentioned in the paper that we do not need a clean image for self-supervised denoising, but the ground truth is required in the code？

Thank you！

Hi, it's me again! I am really interested in your work!
But I have found some questions about the technical detail.

According to figure 3a of the paper, you feed the signal-independent noise map with signal-dependent noise map into the network again and expect to get . I have some questions about this part.

1. Why we need to feed the signal-dependent noise map as well? Why not just feed the signal-independent noise map and expect to get .
2. According to the provided source code, it seems the signal-dependent noise map wasn't fed into the network?
   

   CVF-SID_PyTorch/src/trainer/trainer.py

   Line 87 in c89242a

   noise_w3, noise_b3, clean3 = self.model(padr_crop((noise_b)))

Apart from this, can you share more detail about the setup of single image training? For example, the hyperparameters you have used (epochs, learning rate, etc) for the image shown in the appendix.

Thank you so much!

Hello!
Thank you for your amazing job!
Could you release your detail information about the processing of SIDD-Medium Dataset?
There are only 320 noisy images in this dataset, so did you crop them in many patches before training (e.g. overlap, patch size), and then cropped these patches into 40*40 in training iter/epoch?
Or what's the length of your trainloader? Is ==320 or >>320 ?
Thank you in advance!

Thanks for sharing the source code and the pre-trained model on SIDD validation dataset. I tested the provided pre-trained model on SIDD validation dataset and only got 34.14/0.855 (SPNR/SSIM).

Moreover, I tried to train the denoiser ( CVF-SID (S) ) on SIDD validation dataset directly, and I can only get 34.05/0.847.

I am using the same versions of python and pytorch as yours. I would like to know if I missed something important.

I followed the steps you provided but the test code doesn't work as expected
After I executed the commands:
python test.py --device 0 --config ../models/CVF_SID/SIDD_Val/config.json --resume ../models/CVF_SID/SIDD_Val/model_best.pth
It returns the network structure and these are the last two lines:
Loading checkpoint: ../models/CVF_SID/SIDD_Val/model_best.pth
Checkpoint loaded. Resume training from epoch 501
And the program ends here and output nothing.

I think there's something wrong with the config or the code?

Hello, I want to know how to modify the test code to realize my demand, produce the denoised images. Could you tell me the solution? thx.

self.train_metrics = MetricTracker('Total_loss', writer=self.writer)
I am puzzled by how to maintain the values of self.train_metrics in trainer.py？
I did not find the implementation of the self.train_metrics.update() in your coding

Dear Author:I am not able to get the DND benchmark dataset at the moment. Can you please share the dnd benchmark dataset zip package with me?

FileNotFoundError: [Errno 2] No such file or directory: '../data/SIDD_Small_sRGB_Only/ValidationGtBlocksSrgb.mat'