Hello!
The GT images seem to have a low illumination compared to the input images. I am a little bit confused about the objective. For instance, if we consider this an image-to-image translation task, the generated images should have a low brightness similar to the GT images. However, if the objective is to remove only the shadows by keeping brightness, contrast, exposure, etc., then the GT images are a bit confusing. In my understanding, exposure-related information has a huge impact on quantitative evaluation. Please consider clarifying the evaluation criteria and the objective of the competition.
Thanks, much.
Posted by: CVRG @ Feb. 21, 2023, 2:07 a.m.Also, the input and GT images are not aligned. The disparity between both images is significantly noticeable.
Posted by: CVRG @ Feb. 21, 2023, 2:20 a.m.We have the same questions about the brightness and alignment!
Posted by: BowenZhao @ Feb. 25, 2023, 4:03 a.m.The same questions are bothering us too.
Posted by: jiangchengzhi @ Feb. 25, 2023, 8:42 a.m.I too would like to ask the target of this challenge.
Posted by: Yuki-11 @ Feb. 25, 2023, 8:48 a.m.