lusinlu/gradient-variance-loss
Code of the ICASSP 2022 paper "Gradient Variance Loss for Structure Enhanced Super-Resolution"
This project helps image processing specialists and researchers enhance the visual quality of low-resolution images. It takes a blurry or pixelated image and produces a sharper, more detailed version, focusing on restoring fine textures and structural details that often get lost. Anyone working with digital images who needs to improve their clarity and detail, such as in scientific imaging or media production, would find this useful.
No commits in the last 6 months.
Use this if you need to super-resolve images and prioritize the natural, crisp appearance of textures over purely numerical accuracy metrics like PSNR.
Not ideal if your primary concern is to reduce overall pixel error without specific emphasis on gradient details or if you're working with non-image data.
Stars
33
Forks
—
Language
Python
License
—
Category
Last pushed
Feb 27, 2022
Commits (30d)
0
Get this data via API
curl "https://pt-edge.onrender.com/api/v1/quality/diffusion/lusinlu/gradient-variance-loss"
Open to everyone — 100 requests/day, no key needed. Get a free key for 1,000/day.
Higher-rated alternatives
Fanghua-Yu/SUPIR
SUPIR aims at developing Practical Algorithms for Photo-Realistic Image Restoration In the Wild....
Maclory/SPSR
Pytorch implementation of Structure-Preserving Super Resolution with Gradient Guidance (CVPR...
The-Harsh-Vardhan/HNDSR
A hybrid Neural Operator-Diffusion model for continuous-scale satellite image super-resolution....