Research map / Image generation / Image restoration
Toward Fast, Flexible, and Robust Low-Light Image Enhancement
Check your idea against 108,967 papers
AI-extracted summary. Written by AI from the paper's title only (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 |
|---|---|---|---|
| Use residual illumination learning instead of directly learning illumination (ablation) | SCI illumination estimator · Low-light image enhancement | visual exposure quality | Helped |
| Add the self-calibrated module during training so stage outputs converge (ablation) | SCI · Low-light image enhancement | inference computation | Helped |
| Use one 3-3-channel block instead of three 3-3-3-3-channel blocks (ablation) | SCI · MIT testing dataset | PSNR: 20.4459 -> 20.6074 | Helped |
| Increase the illumination estimator from 1 to 2 blocks (ablation) | SCI · MIT testing dataset | PSNR: 20.6074 -> 20.5809 | No clear effect |
| Increase the illumination estimator from 2 to 3 blocks (ablation) | SCI · MIT testing dataset | FLOPs (G): 0.0410 -> 0.0619 | Hurt |
| Apply SCI training to RUAS with one iterative block instead of using RUAS with three (ablation) | RUAS · LSRW | PSNR: 14.4372 -> 14.7352 | Helped |
| Apply SCI training to RUAS with one iterative block instead of using RUAS with three (ablation) | RUAS · LSRW | FLOPs (G): 0.2813 -> 0.0936 | Helped |
| Use SCI+ joint training of SCI and the face detector rather than fine-tuning the detector on enhanced images (ablation) | S3FD with SCI · DARK FACE | detection precision and recall | Helped |
| Use SCI enhancement before PSPNet rather than the compared enhancement methods | PSPNet · ACDC | mIoU: 45.6 -> 46.3 | Helped |
| Use SCI instead of RUAS | Low-light enhancement · MIT | PSNR: 18.5372 -> 20.4459 | Helped |
Closest papers by meaning
- Zero-Reference Low-Light Enhancement via Physical Quadruple Priors (CVPR 2024 · similarity 0.89)
- Luminance-Aware Statistical Quantization: Unsupervised Hierarchical Learning for Illumination Enhancement (NeurIPS 2025 · similarity 0.88)
- Improving Visual and Downstream Performance of Low-Light Enhancer with Vision Foundation Models Collaboration (CVPR 2025 · similarity 0.87)
- Zero-Shot Low-Light Image Enhancement via Latent Diffusion Models (AAAI 2025 · similarity 0.87)
- Boosting Object Detection with Zero-Shot Day-Night Domain Adaptation (CVPR 2024 · similarity 0.87)
- Interpretable Unsupervised Joint Denoising and Enhancement for Real-World low-light Scenarios (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.