Frank031105/AD-SRF26

AD-SRF: Attack-Defense Semantic Reinforcement for Multi-Modal Image Fusion

This repository contains the model and checkpoints introduced in the paper AD-SRF: Attack-Defense Semantic Reinforcement for Multi-Modal Image Fusion.

The model performs multi-modal image fusion (e.g., infrared + visible images) while balancing visual quality and semantic attributes. It achieves state-of-the-art performance on fusion and downstream segmentation tasks.

Code: https://github.com/HaoZhang1018/AD-SRF

Citation:

@article{adsrf2026,
  title={AD-SRF: Attack-Defense Semantic Reinforcement for Multi-Modal Image Fusion},
  author={Zhang, Hao and Zheng, Zhiqian and Tang, Linfeng and Xiang, Xinyu and Ma, Jiayi},
  journal={...},
  year={2026}
}