How to crop CelebA dataset for training

andreas128 / SRFlow

Official SRFlow training code: Super-Resolution using Normalizing Flow in PyTorch

Other

824 stars 112 forks source link

How to crop CelebA dataset for training #30

Open dongheehand opened 3 years ago

dongheehand commented 3 years ago

Thanks for your great work! I have a question about training using CelebA dataset In SRFlow paper, we pre-process the dataset by cropping aligned patches, which are resized to the HR resolution of 160 × 160.

I am wondering how to crop celeba aligned dataset! Thank you!