SFPFusion

This is the code of the paper titled as "SFPFusion: An Improved Vision Transformer Combining Super Feature Attention and Wavelet-Guided Pooling for Infrared and Visible Images Fusion".

The paper is accepted by Sensors and can be found here

Framework

Example

Environment

• Python 3.9.13
• torch 1.12.1
• torchvision 0.13.1
• tqdm 4.64.1

To Train

We train our network using MS-COCO 2014(T.-Y. Lin, M. Maire, S. Belongie, J. Hays, P. Perona, D. Ramanan, P. Dollar, and C. L. Zitnick. Microsoft coco: Common objects in context. In ECCV, 2014. 3-5.)

You can run the following prompt:

python train_auto_encoder.py

To Test

Put your image pairs in the "test_images" directory and run the following prompt:

python test_SFPFusion.py

Models

The models for our network can be download from models.

The models should be on:

'./models/model_SFPFusion/1e1/Final_xx_epoch.model'

Acknowledgement

• Our code of training is based on the DenseFuse.
• For calculating the image quality assessments, please refer to this Metric.

Contact Informaiton

If you have any questions, please contact me at yongbiao_xiao_jnu@163.com.

Citation

If this work is helpful to you, please cite it as (BibTeX):

@article{li2023sfpfusion,
  title={SFPFusion: An Improved Vision Transformer Combining Super Feature Attention and Wavelet-Guided Pooling for Infrared and Visible Images Fusion},
  author={Li, Hui and Xiao, Yongbiao and Cheng, Chunyang and Song, Xiaoning},
  journal={Sensors},
  volume={23},
  number={18},
  pages={7870},
  year={2023},
  publisher={MDPI}
}