search

williamFalcon / pix2pix-keras

Image-to-Image Translation with Conditional Adversarial Networks (Pix2pix) implementation in keras
MIT License
140 stars 50 forks source link

impossible convolution output dim #5

Open pranitapradhan91 opened 6 years ago

pranitapradhan91 commented 6 years ago

Dear all,

I am getting this error when I keep patch size as (64,64) and image size (256, 256).--> ValueError: impossible convolution output dim: expected 1x512x1x1 but received 1x512x2x2. Can someone tell me what is going wrong here? any help is appreciated.

pranitapradhan91 commented 6 years ago

Should image size be equal to patch size?

edmondja commented 6 years ago

Last time i saw his code (it was in december I think), there was a bug making it impossible to have multiple patches (I forgot where it was). I dont think william updated it since then. Therefore the answer to you question is yes.

pranitapradhan91 commented 6 years ago

Okay. Thank you.