RS-MUCA

KO | EN

gitlite — search

#javascript #python #hacktoberfest #react #ai #typescript #llm #go #golang #android #machine-learning #rust #deep-learning #linux

★ 13 Open GitHub ↗

[TGRS 2025] Semi-supervised semantic segmentation for remote sensing images via multi-scale uncertainty consistency and cross-teacher-student attention

// repository documentation

Was this content helpful?

★ 0 (0 ratings)

# RS-MUCA

The code implementation of the MUCA model.
![muca](https://raw.githubusercontent.com/wangshanwen001/RS-MUCA/master/docs/muca.png)

### Environment

---

```
python == 3.8.10
torch == 1.13.0+cu117
torchvision == 0.14.0+cu117
mmcv == 2.2.0
tensorboard == 2.13.0
pandas == 2.0.2
numpy == 1.24.3
pillow == 9.5.0
tqdm == 4.65.0
einops == 0.8.0
opencv-python == 4.10.0.84
scipy == 1.10.1
```

### Dataset

Your file structure will be like:

---

```
├── [Your Dataset Path]
    ├── JPEGImages
        ├── img_001.jpg
        ├── img_002.jpg
        └── ...
    ├── SegmentationClass
        ├── label_001.png
        ├── label_002.png
        └── ...
    └── ImageSets
    	└──Segmentation
    		└──split.txt
```
Modify your dataset path in train.py
``
Dataset_path  = 'YourDataset'
`` 
.

### Pretrained Backbone
Download the pretrained [Segformer](https://pan.baidu.com/s/1tH4wdGnACtIuGOoXb0_rAw). Code = ``tyjr``.
```
├── model_data
    └── segformer_backbone.pth
```

### Training

---

```
python3 train.py
```

### Evaluating

---
```
python3 evaluate.py
```

### Citation
If you find it useful, please consider citing:
```
@ARTICLE{11062866,
  author={Wang, Shanwen and Sun, Xin and Chen, Changrui and Hong, Danfeng and Han, Jungong},
  journal={IEEE Transactions on Geoscience and Remote Sensing}, 
  title={Semi-Supervised Semantic Segmentation for Remote Sensing Images via Multiscale Uncertainty Consistency and Cross-Teacher–Student Attention}, 
  year={2025},
  volume={63},
  number={},
  pages={1-15},
  keywords={Semantic segmentation;Perturbation methods;Uncertainty;Training;Feature extraction;Computational modeling;Supervised learning;Buildings;Transformers;Semantics;Consistency regularization;deep learning;remote sensing (RS) images;semi-supervised semantic segmentation},
  doi={10.1109/TGRS.2025.3585489}}
```

RS-MUCA

Today's picks

Was this content helpful?

Recent Feedback

RS-MUCA

Today's picks

Was this content helpful?

Recent Feedback

Download README