Research map / Image generation / Image restoration
Retinexformer: One-stage Retinex-based Transformer for Low-light Image Enhancement
Yuanhao Cai, Hao Bian, Jing Lin, Haoqian Wang, Radu Timofte, Yulun Zhang
Read the paperCheck your idea against 137,217 papers
AI-extracted. The measured effects below were read by AI from the paper's full text (no summary yet). Numbers and claims can be wrong or missing context, so check the paper before you cite or build on it.
Measured effects reported in the paper
| What was changed | Where | Measured | Effect |
|---|---|---|---|
| add one-stage Retinex framework (ORF) to the baseline (ablation) | Retinexformer baseline · SDSD-outdoor | PSNR: 26.47 -> 27.92 | Improved |
| add IG-MSA to the baseline (ablation) | Retinexformer baseline · SDSD-outdoor | PSNR: 26.47 -> 28.86 | Improved |
| combine ORF and IG-MSA (ablation) | Retinexformer baseline · SDSD-outdoor | PSNR: 26.47 -> 29.84 | Improved |
| estimate light-up map L_bar and multiply instead of dividing by illumination map L (ablation) | ORF · SDSD-outdoor | PSNR: 28.97 -> 29.26 | Improved |
| divide by estimated illumination map L (I ./ L) rather than no lighting (ablation) | ORF · SDSD-outdoor | PSNR: 28.86 -> 28.97 | No clear effect |
| feed light-up feature F_lu to the corruption restorer (ablation) | ORF · SDSD-outdoor | PSNR: 29.26 -> 29.84 | Improved |
| use IG-MSA instead of global MSA (ablation) | Retinexformer · SDSD-outdoor | PSNR: 28.43 -> 29.84 | Improved |
| use IG-MSA instead of window MSA (ablation) | Retinexformer · SDSD-outdoor | PSNR: 28.65 -> 29.84 | Improved |
| use Retinexformer instead of SNR-Net | low-light enhancement · LOL-v1 | PSNR: 24.61 -> 25.16 | Improved |
| use Retinexformer as preprocessing for low-light detection instead of SNR-Net | YOLO-v3 on ExDark · ExDark | mean AP: 65.3 -> 66.1 | Improved |
| use Retinexformer instead of SNR-Net on LOL-v2-real SSIM | low-light enhancement · LOL-v2-real | SSIM: 0.849 -> 0.840 | Hurt |
Later papers that compared with this method
36 later accepted papers report Retinexformer as a baseline in their results tables; 28 of them report a better number on most of the results they share with it.
- TPCNet: A Low-Light Image Enhancement Network Inspired by Triple Physical Constraints (ECCV 2026 · better on 5 of 5 shared results)
- Event-Illumination Collaborative Low-light Image Enhancement with a High-resolution Real-world Dataset (CVPR 2026 · better on 4 of 4 shared results)
- Multinex: Lightweight Low-light Image Enhancement via Multi-prior Retinex (CVPR 2026 · better on 1 of 3 shared results)
- Bayesian Neural Networks for One-to-Many Mapping in Image Enhancement (AAAI 2026 · better on 1 of 3 shared results)
- ICLR: Inter-Chrominance and Luminance Interaction for Natural Color Restoration in Low-Light Image Enhancement (AAAI 2026 · better on 2 of 3 shared results)
- Difficulty-Conditioned Attribute-Specific Restoration for Low-Light Image Enhancement (ECCV 2026 · better on 3 of 3 shared results)
- P-PatchDiff: Progressive Patch Diffusion Models for Low-Light Image Enhancement (IJCV 2026 · better on 3 of 3 shared results)
- Resolving Blind Inverse Problems under Dynamic Range Compression via Structured Forward Operator Modeling (ICML 2026 · better on 1 of 2 shared results)
28 more not listed.
Earlier methods this paper compares with
Named baselines from earlier accepted papers in this paper's results tables: better than 1 of 1 on most shared results.
- Restormer — Restormer: Efficient Transformer for High-Resolution Image Restoration (CVPR 2022) · better on 1 of 1 shared result
Matched by method name in results tables read by AI. Authors' own tables tend to favour the newer method, and a comparison can be missed when a method is renamed.
Closest papers by meaning
- IniRetinex: Rethinking Retinex-type Low-Light Image Enhancer via Initialization Perspective (AAAI 2025 · similarity 0.91)
- Spike-RetinexFormer: Rethinking Low-light Image Enhancement with Spiking Neural Networks (NeurIPS 2025 · similarity 0.90)
- You Do Not Need Additional Priors or Regularizers in Retinex-Based Low-Light Image Enhancement (CVPR 2023 · similarity 0.89)
- Boosting Object Detection with Zero-Shot Day-Night Domain Adaptation (CVPR 2024 · similarity 0.87)
- MambaLLIE: Implicit Retinex-Aware Low Light Enhancement with Global-then-Local State Space (NeurIPS 2024 · similarity 0.87)
- Reti-Diff: Illumination Degradation Image Restoration with Retinex-based Latent Diffusion Model (ICLR 2025 · similarity 0.86)
Writing a paper in this area? scime draft is a free LaTeX editor with AI included, real-time co-authors and a check of your references and venue rules before you submit. It also imports OpenAI Prism projects.
