Research map / Image generation / Image restoration
Learning to Enhance Low-Light Image via Zero-Reference Deep Curve Estimation
Chongyi Li, Chunle Guo, Change Loy Chen
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 |
|---|---|---|---|
| increase curve iterations from 1 to 8 (ablation) | Zero-DCE (l7-f32) · Part2 test set | PSNR: 12.21 -> 16.57 | Improved |
| increase curve iterations from 8 to 16 (ablation) | Zero-DCE (l7-f32) · Part2 test set | PSNR: 16.57 -> 16.79 | No clear effect |
| increase feature maps per layer from 16 to 32 (ablation) | Zero-DCE (l7) · Part2 test set | PSNR: 15.67 -> 16.57 | Improved |
| increase depth from 3 to 7 conv layers (ablation) | Zero-DCE (f32,n8) · Part2 test set | PSNR: 14.50 -> 16.57 | Improved |
| set well-exposedness E to 0.6 instead of 0.4 (ablation) | Zero-DCE · Part2 test set | PSNR: 11.82 -> 16.57 | Improved |
| set well-exposedness E to 0.6 instead of 0.7 (ablation) | Zero-DCE · Part2 test set | PSNR: 13.40 -> 16.57 | Improved |
| set well-exposedness E to 0.6 instead of 0.5 (ablation) | Zero-DCE · Part2 test set | PSNR: 15.19 -> 16.57 | Improved |
| add depthwise separable convolutions (ablation) | Zero-DCE · Part2 test set | PSNR: 16.57 -> 16.51 | No clear effect |
| add shared curve parameter maps across iterations (ablation) | Zero-DCE · Part2 test set | PSNR: 16.57 -> 16.24 | Hurt |
| combine DSconv and Pshared (Zero-DCE++) (ablation) | Zero-DCE · Part2 test set | PSNR: 16.57 -> 16.42 | No clear effect |
| downsample input 12x (ablation) | Zero-DCE++ · Part2 test set | PSNR: 16.09 -> 16.42 | Improved |
| downsample input 300x instead of 12x (ablation) | Zero-DCE++ · Part2 test set | PSNR: 16.42 -> 14.53 | Hurt |
Closest papers by meaning
- Zero-Reference Low-Light Enhancement via Physical Quadruple Priors (CVPR 2024 · similarity 0.86)
- Interpretable Unsupervised Joint Denoising and Enhancement for Real-World low-light Scenarios (ICLR 2025 · similarity 0.85)
- Zero-Shot Low-Light Image Enhancement via Latent Diffusion Models (AAAI 2025 · similarity 0.85)
- ZERO-IG: Zero-Shot Illumination-Guided Joint Denoising and Adaptive Enhancement for Low-Light Images (CVPR 2024 · similarity 0.85)
- Boosting Object Detection with Zero-Shot Day-Night Domain Adaptation (CVPR 2024 · similarity 0.85)
- Enhancing Low-Light Images: A Synthetic Data Perspective on Practical and Generalizable Solutions (AAAI 2025 · similarity 0.85)
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.
