Skip to content

An attempt at a PyTorch Implementation of "Zero-Shot" Super-Resolution using Deep Internal Learning by Shocher et al. CVPR 2018

License

Notifications You must be signed in to change notification settings

MayankSingal/PyTorch-Zero-Shot-Super-Resolution

Folders and files

NameName
Last commit message
Last commit date

Latest commit

 

History

9 Commits
 
 
 
 
 
 
 
 
 
 
 
 
 
 

Repository files navigation

PyTorch-Zero-Shot-Super-Resolution

An attempt at a PyTorch Implementation of "Zero-Shot" Super-Resolution using Deep Internal Learning by Shocher et al. CVPR 2018

I have tried to implement everything that has been given in the paper. Results are still lacking, but look promising. A lot of the preprocessing code has been taken directly from the original author's amazing repo, which is in tensorflow.

Things left to do:

  1. Get results close to the paper
  2. Study effects of non-local networks, other attention based methods on this task.

To run the model, change the path of the image to be resized in enhancer.py. Then run enhancer.py. Dependencies: PyTorch 0.4 and Python 3.5.

Results on some images which have been resized to 2x size are given below. The blurry image is the one obtained after bicubic interpolation while the sharper image is the one obtained by this implementation.

Alt text

Alt text

About

An attempt at a PyTorch Implementation of "Zero-Shot" Super-Resolution using Deep Internal Learning by Shocher et al. CVPR 2018

Resources

License

Stars

Watchers

Forks

Releases

No releases published

Packages

No packages published

Languages