LBAM_Pytorch icon indicating copy to clipboard operation
LBAM_Pytorch copied to clipboard

Severe artifacts for central hole

Open C-H-D opened this issue 6 years ago • 2 comments

@Vious Thanks for your excellent work! I'm training your model on Paris StreetView dataset with central square hole, but the results all suffer from severe artifacts like these: 350_100 340_1 350_16 I'm training with batchsize=16, do you have any idea why?

C-H-D avatar Dec 10 '19 03:12 C-H-D

Thank you for your interest in our work, we design the network and loss (gan loss) mainly for irregular holes. For a certain kind of holes like central holes, gan loss plays a very important role. Although we use a parallel structure of the discriminator, I observe that when the holes are very large, even the gan loss converges well, the central area still has many artifacts. In this central hole inpainting task, I suposse using a local gan which the discriminator is designed for only minizing real/fake distrution of the hole area.

Vious avatar Dec 10 '19 14:12 Vious

Thanks for your timely reply!😀 哈工大nb!

C-H-D avatar Dec 12 '19 02:12 C-H-D