zhenqifu / pairlie Goto Github PK

为什么LOL-test中reference的结果与 eval 输出的结果差距这么大呢？按照文章中的方式reference出来的结果好像不太对

Hello, can you please share a code for calculating Delta E to my email?

Email: [email protected]

Hi, noticed you've hardcoded the dataset size to 324 sets of paired images:

class DatasetFromFolder(data.Dataset):

   ...
   ...
   ...

    def __len__(self):
        return 324

(dataset.py line 53). This means that users training on custom datasets may either:

1. Only use the first few images of a large training set
2. Run into an indexing error similar to this one with a smaller dataset

Traceback (most recent call last):
  File "<console>", line 1, in <module>
  File "/opt/conda/lib/python3.7/site-packages/torch/utils/data/dataloader.py", line 628, in __next__
    data = self._next_data()
  File "/opt/conda/lib/python3.7/site-packages/torch/utils/data/dataloader.py", line 671, in _next_data
    data = self._dataset_fetcher.fetch(index)  # may raise StopIteration
  File "/opt/conda/lib/python3.7/site-packages/torch/utils/data/_utils/fetch.py", line 58, in fetch
    data = [self.dataset[idx] for idx in possibly_batched_index]
  File "/opt/conda/lib/python3.7/site-packages/torch/utils/data/_utils/fetch.py", line 58, in <listcomp>
    data = [self.dataset[idx] for idx in possibly_batched_index]
  File "/home/.../PairLIE/dataset.py", line 28, in __getitem__
    data_filenames = [join(join(self.data_dir, str(index+1)), x) for x in listdir(join(self.data_dir, str(index+1))) if is_image_file(x)]
FileNotFoundError: [Errno 2] No such file or directory: 'PairLIE/319'

Would it be possible to flag this in the .README or to add a command line argument for the custom dataset size?

Thanks

Hi, I have noticed your work recently and interested with the idea of PairLIE.
When I read your paper and code, I found that the training progress seems to have no relation to the enhace phase.
So I wondering how do you determine the form of g(L)? As a hyperparameter or something else?
Thank you.

Thank you for sharing your great code. 😺

What is the license for this model? I'd like to cite it to the repository I'm working on if possible, but I want to post the license correctly.

• Converted ONNX + TFLite
  https://github.com/PINTO0309/PINTO_model_zoo/tree/main/418_Diffusion-Low-Light

Thank you.

想请教一下作者，这个网络最终输出的结果如何才能转化成综合优化后的图片呢，而不是这种分解量形式的，谢谢

Based on the paper

L: Illuminaton
R: Reflectance
I: Image

And I guess:

D: Residual
X: Noise

Am I correct?

However, I feel like I (left) looks worse than R (right).

Could you please clarify it? Thanks!