/BDLFusion

Codes about Bi-level Dynamic Learning for Jointly Multi-modality Image Fusion and Beyond

Primary LanguagePythonMIT LicenseMIT

BDLFusion

Codes of Bi-level Dynamic Learning for Jointly Multi-modality Image Fusion and Beyond

Zhu Liu, Jinyuan Liu, Guanyao Wu, Long Ma, Xin Fan, Risheng Liu. In IJCAI 2023. Paper

Requirements

  • Python 3.7
  • PyTorch 1.10.1
  • Checkpoint of detection Checkpoint

Usage

Testing

Run "python test.py" to test the model.

Training

Run "python train.py" to train the model.

The warmstart checkpoint and saliency weight maps are provided at url (https://drive.google.com/drive/folders/1WxxuXFDX4-18DfAJjDWoemREBDRcEaDH?usp=drive_link)

Workflow

Results of detection

Results of segmentation

Citation

If you use this code for your research, please cite our paper.

@article{liu2023bilevel,
  title={Bi-level Dynamic Learning  for Jointly Multi-modality Image Fusion and Beyond},
  author={Zhu Liu and Jinyuan Liu and Guanyao Wu and Long Ma and Xin Fan and Risheng Liu},
  journal={IJCAI},
  year={2023},
}