Research map / Image generation / Image restoration
Unsupervised Multi-Exposure Image Fusion Breaking Exposure Limits via Contrastive Learning
Han Xu, Liang Haochen, Jiayi Ma
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 |
|---|---|---|---|
| Use MEF-CL instead of TransMEF | Multi-exposure fusion on SICE · SICE test set | MEF-SSIM: 0.907 ± 0.01 -> 0.911 ± 0.01 | Improved |
| Use MEF-CL instead of Deepfuse | Multi-exposure fusion on SICE · SICE test set | CC: 0.944 ± 0.02 -> 0.946 ± 0.03 | Improved |
| Use MEF-CL instead of Deepfuse | Multi-exposure fusion on SICE · SICE test set | PSNR: 58.687 ± 0.71 -> 58.604 ± 0.73 | Hurt |
| Use MEF-CL instead of IFCNN | Multi-exposure fusion on SICE · SICE test set | EI: 98.784 ± 0.75 -> 80.131 ± 0.68 | Hurt |
| Use MEF-CL instead of Deepfuse | Multi-exposure fusion on SICE · SICE test set | DeltaE: 13.790 ± 0.52 -> 13.572 ± 0.42 | Improved |
| Use MEF-CL instead of MEF-Net | Multi-exposure fusion on SICE · SICE test set | Runtime (second): 0.023 ± 0.01 -> 0.033 ± 0.01 | Hurt |
| Add LMSE (ablation) | MEF-CL · SICE test set | MEF-SSIM: 0.889 -> 0.911 | Improved |
| Add LSSIM (ablation) | MEF-CL · SICE test set | MEF-SSIM: 0.907 -> 0.911 | Improved |
| Add contrastive loss LInfoNCE (ablation) | MEF-CL · SICE test set | MEF-SSIM: 0.875 -> 0.911 | Improved |
| Add attribute block (ablation) | MEF-CL · SICE test set | MEF-SSIM: 0.876 -> 0.911 | Improved |
| Add content block (ablation) | MEF-CL · SICE test set | MEF-SSIM: 0.899 -> 0.911 | Improved |
| Add attribute block (ablation) | MEF-CL · SICE test set | EI: 74.465 -> 80.131 | Improved |
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.
- Deepfuse — DeepFuse: A Deep Unsupervised Approach for Exposure Fusion With Extreme Exposure Image Pairs (ICCV 2017) · better on 2 of 3 shared results
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
- Hybrid-Supervised Dual-Search: Leveraging Automatic Learning for Loss-Free Multi-Exposure Image Fusion (AAAI 2024 · similarity 0.90)
- Conditional Controllable Image Fusion (NeurIPS 2024 · similarity 0.85)
- AGLLDiff: Guiding Diffusion Models Towards Unsupervised Training-free Real-world Low-light Image Enhancement (AAAI 2025 · similarity 0.84)
- One Model for ALL: Low-Level Task Interaction Is a Key to Task-Agnostic Image Fusion (CVPR 2025 · similarity 0.84)
- A Unified Solution to Video Fusion: From Multi-Frame Learning to Benchmarking (NeurIPS 2025 · similarity 0.83)
- Ultra-High-Definition Dynamic Multi-Exposure Image Fusion via Infinite Pixel Learning (AAAI 2025 · similarity 0.83)
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.
