Research map / Image generation / Image restoration
Attention Guided Low-Light Image Enhancement with a Large Scale Low-Light Simulation Dataset
Feifan Lv, Yu Li, Feng Lu
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 bright loss L_eb to mse (ablation) | Enhancement-Net loss · Synthetic low-light ablation | PSNR: 19.36 -> 20.01 | Improved |
| add structural loss L_es to bright loss (ablation) | Enhancement-Net loss · Synthetic low-light ablation | SSIM: 0.76 -> 0.78; PSNR 20.01 -> 19.92 | Mixed |
| add perceptual loss L_ep (ablation) | Enhancement-Net loss · Synthetic low-light ablation | PSNR: 19.92 -> 20.58 | Improved |
| add regional loss L_er (ablation) | Enhancement-Net loss · Synthetic low-light ablation | PSNR: 20.58 -> 20.84 | Improved |
| add Attention-Net (ablation) | Enhancement-Net · Synthetic low-light ablation | PSNR: 19.12 -> 20.66 | Improved |
| add Noise-Net (ablation) | Enhancement-Net with Attention-Net · Synthetic low-light ablation | PSNR: 20.66 -> 20.84 | Improved |
| increase branch number from 5 to 10 (ablation) | multi-branch Enhancement-Net · Synthetic low-light ablation | PSNR: 20.66 -> 20.84 | Improved |
| increase branch number from 10 to 15 (ablation) | multi-branch Enhancement-Net · Synthetic low-light ablation | PSNR: 20.84 -> 20.83 | No clear effect |
| use attention-guided multi-branch network instead of GLADNet | low-light enhancement · Synthetic low-light (with noise) | PSNR: 19.86 -> 20.84 | Improved |
| use attention-guided network instead of MBLLEN | low-light enhancement · Synthetic low-light (no extra noise) | PSNR: 24.21 -> 25.24 | Improved |
| use Enhancement-Net with knowledge transfer instead of RetinexNet | low-light enhancement · LOL | PSNR: 16.77 -> 20.24 | Improved |
| add Reinforce-Net on LOL | Enhancement-Net · LOL | PSNR: 20.24 -> 19.48; SSIM 0.79 -> 0.81 | Mixed |
Closest papers by meaning
- Low-Light Image Enhancement with Multi-Stage Residue Quantization and Brightness-Aware Attention (ICCV 2023 · similarity 0.90)
- Empowering Low-Light Image Enhancer through Customized Learnable Priors (ICCV 2023 · similarity 0.90)
- Enhancing Low-Light Images: A Synthetic Data Perspective on Practical and Generalizable Solutions (AAAI 2025 · similarity 0.89)
- Semantically Contrastive Learning for Low-Light Image Enhancement (AAAI 2022 · similarity 0.89)
- Zero-Shot Low-Light Image Enhancement via Latent Diffusion Models (AAAI 2025 · similarity 0.88)
- Interpretable Unsupervised Joint Denoising and Enhancement for Real-World low-light Scenarios (ICLR 2025 · similarity 0.87)
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.
