This is the code for article MMAE: A universal image fusion method via mask attention mechanism. The detailed code has been uploaded.

这是文章 MMAE: A universal image fusion method via mask attention mechanism 的代码。 详细代码已经上传。

link 链接：https://pan.baidu.com/s/1up-FGeMLt0_LQACDWjDYoQ?pwd=6356

extraction code 提取码：6356

link 链接：https://pan.baidu.com/s/1cP7DKddQHjxO6W6ABZc7cg?pwd=hwso

extraction code 提取码：hwso

Training method: you need to modify the dataset address in the TRAIN file and the save address of the training parameters according to the actual situation. The dataset reading can be modified by yourself according to the actual situation. A reference is provided here.

多聚焦图像融合任务的训练，运行train.py

红外和可见光图像融合任务的训练，运行train-5.py

医学图像融合任务的训练，运行train-4.py

Training for the multifocus image fusion task, running train.py

Training for infrared and visible image fusion task, running train-5.py

Training for medical image fusion task, running train-4.py

Test method: you need to modify the reading address of the training parameters and the fusion image saving address in the TEST file according to the actual situation.

多聚焦图像融合任务测试，运行test.py

红外和可见光图像融合任务测试，运行test-5.py

医学图像融合任务测试，运行test-6.py

Multi-focus image fusion task test, running test.py

Infrared and visible image fusion task test, running test-5.py

Medical image fusion task test, running test-6.py

Download address for saved training files

link 链接：https://pan.baidu.com/s/1pOE-7MzQ8KWa5NGgwz9AfA?pwd=yjig

extraction code 提取码：yjig

或者直接使用pth文件夹中的epoch_82_loss_6.166936.pth文件

Or just use the epoch_82_loss_6.166936.pth file in the pth folder

文章已上线，欢迎引用。 The article is now online. Feel free to cite it.

Article Link 文章链接: https://www.sciencedirect.com/science/article/abs/pii/S0031320324007921

@article{WANG2025111041, title = {MMAE: A universal image fusion method via mask attention mechanism}, journal = {Pattern Recognition}, volume = {158}, pages = {111041}, year = {2025}, issn = {0031-3203}, doi = {https://doi.org/10.1016/j.patcog.2024.111041}, url = {https://www.sciencedirect.com/science/article/pii/S0031320324007921}, author = {Xiangxiang Wang and Lixing Fang and Junli Zhao and Zhenkuan Pan and Hui Li and Yi Li} }