Research map / Vision-language and multimodal / Cross-modal learning
LUT-Fuse: Towards Extremely Fast Infrared and Visible Image Fusion via Distillation to Learnable Look-Up Tables
Xunpeng Yi, Yibing Zhang, Xinyu Xiang, Qinglong Yan, Han Xu, 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 |
|---|---|---|---|
| add low-order approximation encoding (ablation) | LUT-Fuse · MFNet | SSIM: 0.533 -> 0.966 | Improved |
| add contextual scene encoding (ablation) | LUT-Fuse · MFNet | SSIM: 0.553 -> 0.966 | Improved |
| add efficient MM-LUT distillation strategy (ablation) | LUT-Fuse · MFNet | SSIM: 0.904 -> 0.966 | Improved |
| add efficient MM-LUT distillation strategy (ablation) | LUT-Fuse · MFNet | Q^AB/F: 0.566 -> 0.628 | Improved |
| add low-order approximation encoding (ablation) | LUT-Fuse · MFNet | entropy: 6.574 -> 6.394 | Hurt |
| use learnable distillation LUT instead of quantization LUT | Tar-DAL · MFNet | SSIM: 0.476 -> 0.745 | Improved |
| use learnable distillation LUT instead of quantization LUT | Diff-IF · MFNet | Q^AB/F: 0.202 -> 0.625 | Improved |
| use learnable distillation LUT instead of quantization LUT | CDDFuse · MFNet | SSIM: 0.553 -> 0.857 | Improved |
| distill MM-Net into MM-LUT (student) instead of using MM-Net | LUT-Fuse · MFNet | MI: 2.764 -> 2.560 | Hurt |
| use LUT-Fuse instead of Tar-DAL for fusion speed | LUT-Fuse · MFNet runtime, RTX 4060 Ti | time (ms): 21.83 -> 4.70 | Improved |
| use LUT-Fuse instead of Tar-DAL on edge device | LUT-Fuse · Jetson Orin NX 720P | time (ms): 389.34 -> 30.54 | Improved |
| use LUT-Fuse fused images for segmentation instead of Text-IF | LUT-Fuse · MFNet segmentation | mIoU: 55.32 -> 54.82 | Mixed |
Later papers that compared with this method
1 later accepted paper reports LutFuse as a baseline in their results tables; 1 of them report a better number on most of the results they share with it.
- P²Fusion: Prompt-based Progressive Infrared-Visible Image Fusion via Dual-Prior Distillation (ECCV 2026 · better on 5 of 5 shared results)
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.
- EMMA — EMMA: Empowering Multi-modal Mamba with Structural and Hierarchical Alignment (ICLR 2025) · better on 2 of 2 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
- EvaNet: Toward More Efficient and Consistent Infrared and Visible Image Fusion Assessment (TPAMI 2026 · similarity 0.87)
- Multi-Modal Gated Mixture of Local-to-Global Experts for Dynamic Image Fusion (ICCV 2023 · similarity 0.86)
- LRRNet: A Novel Representation Learning Guided Fusion Network for Infrared and Visible Images (TPAMI 2023 · similarity 0.86)
- Probing Synergistic High-Order Interaction in Infrared and Visible Image Fusion (CVPR 2024 · similarity 0.84)
- Revisiting Generative Infrared and Visible Image Fusion Based on Human Cognitive Laws (NeurIPS 2025 · similarity 0.84)
- Enhancing Infrared Vision: Progressive Prompt Fusion Network and Benchmark (NeurIPS 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.
