HBANet

Official code of HBANet: A hybrid boundary-aware attention network for infrared and visible image fusion.

Introduction

This repository contains the official code of the paper "HBANet: A hybrid boundary-aware attention network for infrared and visible image fusion". The code is implemented in PyTorch.

Installation

1. Clone this repository.

git clone https://github.com/LuoXubo/HBANet.git
cd HBANet

2. Install the required packages.

pip install -r requirements.txt

Usage

1. Download the dataset and put it in the data folder.
2. Train the model (Optional).

python train.py

3. Test the model.

python test.py

Acknowledgement

The code is based on the following repositories:

• BA-Transformer
• SwinFusion

Citation

If you find this work helpful, please consider citing:

@article{LUO2024104161,
title = {HBANet: A hybrid boundary-aware attention network for infrared and visible image fusion},
journal = {Computer Vision and Image Understanding},
volume = {249},
pages = {104161},
year = {2024},
issn = {1077-3142},
doi = {https://doi.org/10.1016/j.cviu.2024.104161},
url = {https://www.sciencedirect.com/science/article/pii/S107731422400242X},
author = {Xubo Luo and Jinshuo Zhang and Liping Wang and Dongmei Niu}
}