Skip to content
/ DPSR Public
forked from Cagatayisil/DPSR

Deep Plug-and-Play Super-Resolution for Arbitrary Blur Kernels (CVPR, 2019) (PyTorch)

Notifications You must be signed in to change notification settings

gdma96/DPSR

 
 

Repository files navigation

DPSR

Deep Plug-and-Play Super-Resolution for Arbitrary Blur Kernels (CVPR, 2019)

Training and testing codes for the super-resolver prior (PyTorch)


The left is the blurry LR image. The right is the super-resolved image by DPSRGAN with scale factor 4.

Run demo_test_dpsr.py to produce the following results.


Super-resolved images of LR image chip.png by DPSR with scale factors 2, 3 and 4.

Run demo_test_dpsr_real.py to produce the following results.

LR

x2

x3

x4

Requirements and Dependencies

  • Spyder (Python 3.6)
  • PyTorch 0.4.1
  • Windows 10

Citation

@inproceedings{zhang2019deep,
  title={Deep Plug-and-Play Super-Resolution for Arbitrary Blur Kernels},
  author={Zhang, Kai and Zuo, Wangmeng and Zhang, Lei},
  booktitle={IEEE Conference on Computer Vision and Pattern Recognition},
  year={2019}
}

About

Deep Plug-and-Play Super-Resolution for Arbitrary Blur Kernels (CVPR, 2019) (PyTorch)

Resources

Stars

Watchers

Forks

Releases

No releases published

Packages

No packages published

Languages

  • Python 96.4%
  • MATLAB 3.5%
  • M 0.1%