search

shepnerd / inpainting_gmcnn

Image Inpainting via Generative Multi-column Convolutional Neural Networks, NeurIPS2018
MIT License
427 stars 98 forks source link

help me #39

Closed 12-wu closed 4 years ago

12-wu commented 5 years ago

Excuse me,I want to run prtrained models of CelebA-HQ-256,but have a error as follows :‘ return files[sorted(range(len(file_times)), key=lambda x: file_times[x])[-1]] IndexError: list index out of range’ I can't solve it.

shepnerd commented 5 years ago

It seems the path of the pretrained model is invalid. You can check the path first.

12-wu commented 5 years ago

thank you very much.I have already run the pretrained model of the celeba-hq-256.but the results are not very well,such as : image image image image image image image image I think may be the mask size is too large to inpainting well,can you help me solve it?and make the visual effect of the results better.Thanks again

shepnerd commented 5 years ago

Your used testing images are from celeba and the pretrained model is for celeba-hq (with 256 x 256 resolution). It seems the performance of celebahq model would degrade in this case as the differences between these two datasets are not to be neglected. Our pretrained model for celeba will be released at the end of this month.

12-wu commented 5 years ago

Excuse me,I have a problem when I try to train as follows: 1 2 I help you can help me.thanks a lot

shepnerd commented 5 years ago

Refer to #11 .

12-wu commented 5 years ago

I'm so sorry to bother again. I have a problem when I want train such as: result but I have tensoflow.layers.and I have install tensorflow again. this is my command： order. The whole process is as follows all all2

shepnerd commented 4 years ago

It might be caused the tensorflow version. Please refer to here.