This is official Pytorch implementation of "A Full-Scale Hierarchical Encoder-Decoder Network with Cascading Edge-prior for Infrared and Visible Image Fusion"

@article{luo2023full,
  title={A Full-Scale Hierarchical Encoder-Decoder Network with Cascading Edge-prior for Infrared and Visible Image Fusion},
  author={Luo, Xiaoqing and Wang, Juan and Zhang, Zhancheng and Wu, Xiao-jun},
  journal={Pattern Recognition},
  pages={110192},
  year={2023},
  publisher={Elsevier}
}

The overall framework of the proposed FSFusion.

1. Downloading the pre-trained checkpoint from hed_pretrained_bsds.caffemodel and putting it in ./HED.
2. The pre-trained checkpoint is put in ./models/Final.model.
3. The test datasets are put in ./images/.
4. The result data_root are put in ./outputs/dataset_name.

Then running test.py