Saved in:
| Main Authors: | , , , , |
|---|---|
| Format: | Preprint |
| Published: |
2025
|
| Subjects: | |
| Online Access: | https://arxiv.org/abs/2510.13067 |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| _version_ | 1866915555588440064 |
|---|---|
| author | Yang, Kaixuan Xiang, Wei Chen, Zhenshuai Jin, Tong Liu, Yunpeng |
| author_facet | Yang, Kaixuan Xiang, Wei Chen, Zhenshuai Jin, Tong Liu, Yunpeng |
| contents | Infrared and visible image fusion aims to integrate complementary information from co-registered source images to produce a single, informative result. Most learning-based approaches train with a combination of structural similarity loss, intensity reconstruction loss, and a gradient-magnitude term. However, collapsing gradients to their magnitude removes directional information, yielding ambiguous supervision and suboptimal edge fidelity. We introduce a direction-aware, multi-scale gradient loss that supervises horizontal and vertical components separately and preserves their sign across scales. This axis-wise, sign-preserving objective provides clear directional guidance at both fine and coarse resolutions, promoting sharper, better-aligned edges and richer texture preservation without changing model architectures or training protocols. Experiments on open-source model and multiple public benchmarks demonstrate effectiveness of our approach. |
| format | Preprint |
| id |
arxiv_https___arxiv_org_abs_2510_13067 |
| institution | arXiv |
| publishDate | 2025 |
| record_format | arxiv |
| spellingShingle | Direction-aware multi-scale gradient loss for infrared and visible image fusion Yang, Kaixuan Xiang, Wei Chen, Zhenshuai Jin, Tong Liu, Yunpeng Computer Vision and Pattern Recognition Infrared and visible image fusion aims to integrate complementary information from co-registered source images to produce a single, informative result. Most learning-based approaches train with a combination of structural similarity loss, intensity reconstruction loss, and a gradient-magnitude term. However, collapsing gradients to their magnitude removes directional information, yielding ambiguous supervision and suboptimal edge fidelity. We introduce a direction-aware, multi-scale gradient loss that supervises horizontal and vertical components separately and preserves their sign across scales. This axis-wise, sign-preserving objective provides clear directional guidance at both fine and coarse resolutions, promoting sharper, better-aligned edges and richer texture preservation without changing model architectures or training protocols. Experiments on open-source model and multiple public benchmarks demonstrate effectiveness of our approach. |
| title | Direction-aware multi-scale gradient loss for infrared and visible image fusion |
| topic | Computer Vision and Pattern Recognition |
| url | https://arxiv.org/abs/2510.13067 |