# GTMFuse
**Repository Path**: qazYP/GTMFuse
## Basic Information
- **Project Name**: GTMFuse
- **Description**: No description available
- **Primary Language**: Unknown
- **License**: Not specified
- **Default Branch**: main
- **Homepage**: None
- **GVP Project**: No
## Statistics
- **Stars**: 0
- **Forks**: 0
- **Created**: 2025-03-12
- **Last Updated**: 2025-03-12
## Categories & Tags
**Categories**: Uncategorized
**Tags**: None
## README
# GTMFuse: Group-attention transformer-driven multiscale dense feature-enhanced network for infrared and visible image fusion
⭐ This code has been completely released ⭐
⭐ our [article](https://doi.org/10.1016/j.knosys.2024.111658) ⭐
If our code is helpful to you, please cite:
```
@article{mei2024gtmfuse,
title={GTMFuse: Group-attention transformer-driven multiscale dense feature-enhanced network for infrared and visible image fusion},
author={Mei, Liye and Hu, Xinglong and Ye, Zhaoyi and Tang, Linfeng and Wang, Ying and Li, Di and Liu, Yan and Hao, Xin and Lei, Cheng and Xu, Chuan and others},
journal={Knowledge-Based Systems},
volume={293},
pages={111658},
year={2024},
publisher={Elsevier}
}
```
## To Train
```
python train.py
```
## To Test
1. Downloading the pre-trained checkpoint from [best_model.pth](https://pan.baidu.com/s/1KRxEpTCM0t4fPgvPz9iRaQ?pwd=of0u) and putting it in **./checkpoints**.
2. python test.py
## HBUT dataset
Downloading the HBUT dataset from [HBUT](https://pan.baidu.com/s/1AcPukklhBTSL3SOJZC2D2Q?pwd=31ys)
## overall network
## Results
### TNO Datasset
#### Qualitative result
- Four representative images of the TNO test set.In alphabetical order they are infrared image, visible image, GTF, FusionGAN, SDNet, RFN–Nest, U2Fusion, LRRNet, SwinFusion, CDDFuse, DATFuse, and GTMFuse.
#### Quantitative Results
| **Methods** | **EN** | **SD** | **SF** | **VIF** | **AG** | **Qabf** |
|:----------------:|:---------:|:---------:|:---------:|:------------:|:-----------------------:|:---------:|
| **GTF** | 6.60008 | 8.69847 | 0.04613 | 0.49451 | 4.36880 | 0.43436 |
| **FusionGAN** | 6.50420 | 8.31568 | 0.03139 | 0.61350 | 3.20322 | 0.25814 |
| **SDNet** | 6.58356 | 8.58165 | 0.05916 | 0.63884 | 6.03142 | 0.44332 |
| **RFN–Nest** | 6.67323 | 8.80744 | 0.03991 | 0.67467 | 4.11122 | 0.43302 |
| **U2Fusion** | 6.90395 | 8.98294 | 0.05854 | 0.68953 | 6.27575 | 0.45158 |
| **LRRNet** |6.70679 | 9.14174 | 0.05434 |0.74519 |4.82270 | 0.38131 |
| **SwinFusion** | 6.69018 | 8.74623 | 0.04868 | 0.81244 | 4.74248 | 0.52084 |
| **CDDFuse** | 7.02021 | 8.92531 | 0.06416| 0.82307 | 6.07101 | 0.51846 |
| **DATFuse** | 6.77604 | 8.82027 | 0.04612 | 0.82066 | 4.51106 | 0.51235 |
| **GTMFuse** | **7.03991** |**9.22010** | **0.06607** | **0.84018** | **6.65676** | **0.60472**|
### RoadScene Datasset
#### Qualitative result
- Four representative images of the RoadScene test set.In alphabetical order they are infrared image, visible image, GTF, FusionGAN, SDNet, RFN–Nest, U2Fusion, LRRNet, SwinFusion, CDDFuse, DATFuse, and GTMFuse.
#### Quantitative Results
| **Methods** | **EN** | **SD** | **SF** | **VIF** | **AG** | **Qabf** |
|:----------------:|:---------:|:---------:|:---------:|:------------:|:-----------------------:|:---------:|
| **GTF** | 7.45805 | 10.4952 | 0.04605 | 0.57953 | 4.04458 | 0.37079 |
| **FusionGAN** | 7.09511 | 10.0518 | 0.04323 | 0.56307 | 4.11028 | 0.28132 |
| **SDNet** | 7.3388 | 10.1153 | 0.07541 | 0.74513 | 7.55126 | 0.51691 |
| **RFN–Nest** | 7.34281 | 10.2000 | 0.05192 | 0.75382 | 5.13552 | 0.45230 |
| **U2Fusion** | 7.21249 | 10.1205 | 0.07469 | 0.67670 | 7.42630 | 0.51831 |
| **LRRNet** |7.09023 | 10.1468 | 0.06907 | 0.64912 | 6.19723 | 0.41013 |
| **SwinFusion** | 7.18569 | 10.3193 | 0.06757 | 0.80244 | 6.52487 | 0.57124 |
| **CDDFuse** | **7.48812** | **10.6921** | **0.09099**| 0.78466 | **8.33022** | 0.49671 |
| **DATFuse** | 6.89646 | 10.4078 | 0.05495 | 0.79045 | 5.06397 | 0.50003 |
| **GTMFuse** | 7.35795 |10.5113| 0.08181 | **0.87918** | 7.92432 | **0.60665** |
### MSRS Dataset
#### Qualitative result
- Four representative images of the MSRS test set. In alphabetical order they are infrared image, visible image, GTF, FusionGAN, SDNet, RFN–Nest, U2Fusion, LRRNet, SwinFusion, CDDFuse, DATFuse, and GTMFuse.
#### Quantitative Results
| **Methods** | **EN** | **SD** | **SF** | **VIF** | **AG** | **Qabf** |
|:----------------:|:---------:|:---------:|:---------:|:------------:|:-----------------------:|:-------------------------:|
| **GTF** | 4.44195 | 6.11111 | 0.05620 | 0.48176 | 3.53966 | 0.39194 |
| **FusionGAN** | 5.86785 | 6.79263 | 0.03654 | 0.61998 | 3.00051 | 0.24709 |
| **SDNet** | 5.54468 | 6.13925 | 0.05910 | 0.48644 | 4.40836 | 0.41903 |
| **RFN–Nest** | 5.81096 | 7.91701 | 0.04982 | 0.74520 | 4.12198 | 0.50474 |
| **U2Fusion** | 5.03625 | 6.78870 | 0.06157 | 0.57216 | 4.48894 | 0.42512 |
| **LRRNet** |5.89799 | 7.30930 | 0.04548 | 0.38422 | 3.64204 | 0.19980 |
| **SwinFusion** | 6.61543 | 8.46817 | 0.06756 | 0.99403 | 5.26562 | 0.66481 |
| **CDDFuse** | 6.32740 | 8.53021 | **0.08130** | 0.97155 | 6.12164 | 0.66558 |
| **DATFuse** | 6.29844 | 7.71886 | 0.07247 | 0.71196 | 5.96856 | 0.54618 |
| **GTMFuse** | **6.78256** | **8.60603** | 0.08105 | **1.00857** | **6.39748** | **0.69590** |
If you have any questions, please contact me by email (hux18943@gmail.com).