search

zengxianyu / crfill

CR-Fill: Generative Image Inpainting with Auxiliary Contextual Reconstruction. ICCV 2021
Other
226 stars 43 forks source link

How to train model on high-resolution images #18

Open herbiezhao opened 2 years ago

herbiezhao commented 2 years ago

If I want to inference on high-resolution images(2048x2048), how do I need to train the model? The current effect is that the higher the resolution, the worse the effect. In addition to modifying the crop size, do I need to modify the discriminator and others?

zengxianyu commented 2 years ago

You can try adding more layers to the discriminator

roimulia2 commented 2 years ago

Hey @herbiezhao, did you manage to solve this challenge?