Training CinCGAN with our own dataset

Junshk / CinCGAN-pytorch

Unofficial Implementation of "Unsupervised Image Super-Resolution using Cycle-in-Cycle Generative Adversarial Networks" in CVPR 2018.

142 stars 27 forks source link

Training CinCGAN with our own dataset #3

Open MFarooqAit opened 5 years ago

MFarooqAit commented 5 years ago

How can we train CinCGAN with our own dataset? Kindly help in this matter.

Junshk commented 5 years ago

Let me know about your dataset structure. Generally, you can train with our own dataset by modifying code/data/srdata.py and code/data/div2k.py

MFarooqAit commented 5 years ago

Dear Junshk, Thank you so much for your reply!

I want to use CelebA dataset with same images size as mentioned in paper (Unsupervised Image Super-resolution using CinCGAN). I want to used 128 x 128 HR and 32 x 32 LR images for training.