search

taki0112 / UGATIT

Official Tensorflow implementation of U-GAT-IT: Unsupervised Generative Attentional Networks with Adaptive Layer-Instance Normalization for Image-to-Image Translation (ICLR 2020)
MIT License
6.17k stars 1.04k forks source link

Single channel training set #11

Open TKassis opened 5 years ago

TKassis commented 5 years ago

Trying to train on single-channel grayscale images using '--img_ch 1' argument during training but I get the following error immediately after starting training:

ValueError: Trying to share variable generator_A/conv/conv2d/kernel, but specified shape (7, 7, 3, 64) and found shape (7, 7, 1, 64).

yxt132 commented 5 years ago

Similar problem here. I am using 4-channel image and get an error message: ValueError: Trying to share variable generator_A/conv/conv2d/kernel, but specified shape (7, 7, 3, 64) and found shape (7, 7, 4, 64).

Did you solve the problem? what if the input is 4-channel and output is 1-channel?

TKassis commented 5 years ago

Similar problem here. I am using 4-channel image and get an error message: ValueError: Trying to share variable generator_A/conv/conv2d/kernel, but specified shape (7, 7, 3, 64) and found shape (7, 7, 4, 64).

Did you solve the problem? what if the input is 4-channel and output is 1-channel?

I have not. Just been using 3 channels instead.

taki0112 commented 5 years ago

Is there an error in the train phase?

chokyungjin commented 3 years ago

Same problems, in training Code.