# caiyuanhao1998/Retinexformer

"Retinexformer: One-stage Retinex-based Transformer for Low-light Image Enhancement" (ICCV 2023 Top-10 Cited 🏆) & (NTIRE 2024 Runner-Up 🏆) & (NTIRE 2025 Winner 🏆) & (NTIRE 2026 Winner 🏆)

Repository: https://github.com/caiyuanhao1998/Retinexformer
Canonical: https://ross.abutalabs.com/products/retinexformer
Homepage: https://arxiv.org/abs/2303.06705
Language: Python
License: MIT
License Family: permissive
Topics: low-light-image-enhancement, low-light-vision, object-detection, nighttime-enhancement, image-restoration, transformer, detection, ntire, basicsr, iccv2023, low-light-enhance, low-light-enhancement, low-light-enhancer
Last push: 2026-05-23T16:09:23+00:00

## Health v2 (maintenance only)
Score: 66/100 (v2, computed 2026-09-02T17:46:02.011165+00:00)
- activity 83, release rhythm 35, longevity 81
- inputs: {"age_days": 1145, "days_push": 102, "days_rel": null, "gap_med": null, "n_releases_24m": 0}
- flags: no_releases
- formula: round(0.45*activity + 0.35*rhythm + 0.20*longevity); archived -> min(score, 10)

## Adoption (not part of the score)
Stars 1518, forks 122 (observed 2026-08-28T04:04:57.437621+00:00)

## What it is
Retinexformer is a one-stage Retinex-based Transformer model and toolbox for low-light image enhancement, published at ICCV 2023. It supports over 15 benchmarks, extremely high-resolution enhancement, and low-light object detection applications, and has won or inspired winning solutions in NTIRE challenges from 2024 to 2026.

## Use cases
- enhance dark low-light photos
- brighten nighttime images with a transformer model
- preprocess low-light images for object detection
- compare low-light enhancement methods across benchmarks
- enhance very high-resolution images up to 4000x6000
- reproduce ICCV 2023 low-light enhancement research
- build a low-light enhancement solution for NTIRE challenges

## When to choose
- you need state-of-the-art low-light image enhancement with pretrained models
- you want a baseline and toolbox covering 15+ benchmarks
- you need illumination-guided transformer modeling for dark images
- you want to enhance images before low-light object detection
- you need a proven competition-winning enhancement method

## When to avoid
- you need general-purpose photo editing rather than low-light enhancement
- you want a lightweight CPU-only real-time enhancer
- you need a polished end-user application with a GUI
- your task is unrelated to image restoration or enhancement

## Facets
- artifact type: library
- maturity: active
- function: image-processing, computer-vision, deep-learning, machine-learning
- domain: computer-vision, image-processing, deep-learning, machine-learning
- platform: python, cross-platform
- tags: low-light-image-enhancement, retinex, transformer, image-restoration, iccv2023, ntire, nighttime-enhancement, object-detection, basicsr, research-code, gpu, linux

## Member repositories
- caiyuanhao1998/Retinexformer (main) score 66

## Provenance
- Observed fields: from GitHub, fetched 2026-08-28T04:04:57.437621+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 2026-08-30T04:32:00.737885+00:00, confidence not recorded.
  - readme: https://github.com/caiyuanhao1998/Retinexformer (fetched 2026-08-28T04:04:57.437621+00:00, sha 9013d158c201)
  - homepage: https://arxiv.org/abs/2303.06705 (fetched 2026-08-29T11:35:24.641322+00:00, sha 9c0d18f9a0f3)
  - site_page: https://info.arxiv.org/about/donate.html (fetched 2026-08-29T11:35:24.650629+00:00, sha cca9c3a11c56)
  - site_page: https://info.arxiv.org/about/ourmembers.html (fetched 2026-08-29T11:35:24.653911+00:00, sha 47cbc55ff1de)
  - site_page: https://info.arxiv.org/about (fetched 2026-08-29T11:35:24.655529+00:00, sha a1f16f915a9a)
  - site_page: https://info.arxiv.org/labs/index.html (fetched 2026-08-29T11:35:24.652396+00:00, sha b14a8d05a0ec)
- Data as of 2026-08-30T08:39:29.467469+00:00.