# Fanghua-Yu/SUPIR

SUPIR aims at developing Practical Algorithms for Photo-Realistic Image Restoration In the Wild. Our new online demo is also released at suppixel.ai.

Repository: https://github.com/Fanghua-Yu/SUPIR
Canonical: https://ross.abutalabs.com/products/supir
Homepage: http://supir.xpixel.group/
Language: Python
License: NOASSERTION
License Family: other
Topics: deep-learning, diffusion-models, llava, sdxl, stable-diffusion, super-resolution, restoration, pytorch, pytorch-lightning
Last push: 2025-05-12T09:49:11+00:00

## Health v2 (maintenance only)
Score: 36/100 (v2, computed 2026-09-02T17:46:02.011165+00:00)
- activity 21, release rhythm 35, longevity 70
- inputs: {"age_days": 986, "days_push": 478, "days_rel": null, "gap_med": null, "n_releases_24m": 0}
- flags: no_releases, no_license
- formula: round(0.45*activity + 0.35*rhythm + 0.20*longevity); archived -> min(score, 10)

## Adoption (not part of the score)
Stars 5649, forks 476 (observed 2026-08-28T04:09:27.013917+00:00)

## What it is
SUPIR is a Python-based photo-realistic image restoration system built on SDXL diffusion priors and LLaVA captioning, presented at CVPR 2024. It upscales and restores low-quality photos (portraits, landscapes, vintage images) using text-driven diffusion-based restoration models.

## Use cases
- upscale and restore low-quality photos with AI
- enhance old or vintage photographs
- restore blurry portrait and face images
- super-resolution for landscape and gaming screenshots
- run diffusion-based image restoration locally with PyTorch

## When to choose
- you need state-of-the-art photo-realistic super-resolution and restoration
- you have a GPU and want to run restoration locally with pretrained checkpoints
- you want text-prompt-guided control over image restoration results

## When to avoid
- you need a lightweight CPU-only upscaler
- you want a simple one-click tool without downloading multi-gigabyte checkpoints
- your project requires a permissive open-source license (license is non-standard)

## Facets
- artifact type: library
- maturity: active
- function: image-processing, machine-learning, deep-learning
- domain: image-processing, computer-vision, deep-learning, artificial-intelligence
- platform: python, cross-platform
- tags: super-resolution, image-restoration, diffusion-models, stable-diffusion-xl, llava, pytorch, photo-enhancement, upscaling, gpu, linux

## Member repositories
- Fanghua-Yu/SUPIR (main) score 36

## Provenance
- Observed fields: from GitHub, fetched 2026-08-28T04:09:27.013917+00:00.
- Health v2: computed from the inputs above; adoption is never an input.
- Inferred fields (summary, facets, guidance): AI-extracted, prompt v1, taxonomy v1, on 2026-08-29T17:55:16.769511+00:00, confidence not recorded.
  - readme: https://github.com/Fanghua-Yu/SUPIR (fetched 2026-08-28T04:09:27.013917+00:00, sha 2967c5e248ff)
  - homepage: http://supir.xpixel.group/ (fetched 2026-08-29T08:49:42.850933+00:00, sha 1081549c5b3b)
- Data as of 2026-08-30T08:39:29.467469+00:00.