SpA-Former-shadow-removal icon indicating copy to clipboard operation
SpA-Former-shadow-removal copied to clipboard

Published 20 hours ago • verified publisher icon zhangbaijin

Mis-Match Results

Open Hamzat0014 opened this issue 1 year ago • 1 comments

Hello Brother, Thanks for the amazing work.

I have done inference of ISTD test data set images with the the checkpoint provided by you. I am unable to get the good results. The model is performing extremally well when there is a "grey" color image at its input. 95-4

But when there is a combination of colors i.e. even 2 colors in an image, there is hardly any difference between the input image and the output shadow image produced. Few of the results are attached. 90-1 91-1 96-6

Hamzat0014 avatar Sep 30 '22 10:09 Hamzat0014

yeah,there is a problem haven't been solved, I think the ISTD dataset has a large variety of scenes, and there are not so many scene samples. I suggest that you can train on ISTD+ or improve the number of samples through data enhancement

zhangbaijin avatar Sep 30 '22 11:09 zhangbaijin