pytorch-layoutnet icon indicating copy to clipboard operation
pytorch-layoutnet copied to clipboard
verified publisher icon sunset1995

Why Eval with flip and rotation

Open robotshanxi opened this issue 6 years ago • 2 comments

It looks to me the network training has no flip or rotation, why they are added as evaluation? Thanks

robotshanxi avatar Jan 23 '19 08:01 robotshanxi

Hi, The network was actually trained with flip and rotation augmentation (see dataset.py).

sunset1995 avatar Jan 23 '19 11:01 sunset1995

Thanks for the reply, I was thinking args.no_flip is true. My mistake.

robotshanxi avatar Jan 24 '19 00:01 robotshanxi