We found train_connect_component_mask.m would only process the last image in the list.
In matlab_code/train_connect_component_mask.m, line 67
The code is like "for i = size(list,2):size(list,2)".
Should it be changed to "for i = 1:size(list,2)"?
I have tried several times to download the training results, but still cannot get them. I'm wondering if the file in Google Drive is broken or just because the file(4.3G) is too large for browser. Could you please upload results with smaller sizes (like to divide them into several parts)? Thank you for your help : )