gzhu-image-lab / clriqa Goto Github PK

Hi!
Thanks for your great job!
I am trying to use your method in my project. But I find out that may be there is a little wrong in the underexposed luminance component function.

I draw a curve for the function, and the result is as follow:

Look forward your reply!
Beas wishes!
Aaron

Dear author:

Thanks for such a good idea!

I have a few questions for you.

1. I encountered a problem when reproducing your code (as shown in the figure below). Can you give me some suggestions?
2. In "test_file = '../../data/csiq/train_data/' + 'ft_test.txt'", ft_test.txt is a label for images? (such as "Original""Level1""Level2""Leve3")
3. When can your model be uploaded？

We look forward to your reply.

您好，有幸拜读您的论文，请问本项目什么时候开源？

Dear author:

Thanks for such a good idea!

I have a little question for the expansion of synthetically distorted image database. For a synthetically distorted image with a quality score, how many images are generated from it in total? One synthetically distorted image are deformed with different types of distortion and different distortion levels, or just deformed with one distortion type which consistent with the distortion type in the original synthetically distorted image?

looking forward to your reply. Thank you very much.