KO | EN

GitLite

Minimalist Open Source Repository Search Engine

gitlite — search

#dsh #dsh-plugin #developer-tools #ai-agent #claude-code #codex #typescript #ai-agents #deepseek-harness #mcp #cli #rust

OpenEarthMap-SAR

★ 21 Open GitHub ↗

OpenEarthMap-SAR: A benchmark dataset for land cover mapping under all-weather conditions

// repository documentation

Was this content helpful?

★ 0 (0 ratings)

<div align="justify">
	
## OpenEarthMap-SAR
<p align="justify">
The <a href="https://arxiv.org/abs/2501.10891v2">OpenEarthMap Synthetic Aperture Radar</a> (OpenEarthMap-SAR) is a high-resolution SAR benchmark dataset for land cover mapping under all-weather conditions. The motivation of this benchmark dataset is to facilitate advancements in SAR-based geospatial analysis for global high-resolution land cover mapping. 
</p> 
<p>	
The OpenEarthMap-SAR dataset was served as the official dataset for the <a href="https://www.grss-ieee.org/technical-committees/image-analysis-and-data-fusion/?tab=data-fusion-contest">2025 IEEE GRSS Data Fusion Contest Track 1</a> organized by the IEEE GRSS Image Analysis and Data Fusion Technical Committee, the University of Tokyo, RIKEN, and ETH Zurich. The contest aims to foster the development of innovative solutions for all-weather land cover and building damage mapping using multimodal SAR and optical EO data at submeter resolution. Check out the winners of the contest at <a href="https://www.grss-ieee.org/community/technical-committees/winners-of-the-2025-ieee-grss-data-fusion-contest-all-weather-land-cover-and-building-damage-mapping/">here</a>.
</p>
</div>

<div align="center">
	
[![GitHub license](https://badgen.net/github/license/Naereen/Strapdown.js)](https://github.com/Naereen/StrapDown.js/blob/master/LICENSE)
<a href="https://pytorch.org/get-started/locally/"><img alt="PyTorch" src="https://img.shields.io/badge/PyTorch-ee4c2c?logo=pytorch&logoColor=white"></a>
![Python](https://img.shields.io/badge/python-3.9+-blue.svg)
</div>

## Dataset
<div align="justify">
<p>	
The OpenEarthMap-SAR dataset consists of 1.5 million segments of 5033 aerial and satellite images covering 35 regions from Japan, France and the USA; and with partially manually annotated labels and fully pseudo labels of 8 land cover classes. Each image has a size of 1024x1024 pixels at a ground sampling distance of 0.15m--0.5m. The dataset has been made publicly available at <a href="https://zenodo.org/records/14622048">Zenodo</a>, where you can download it. Below are examples of the dataset.
</p>
<p><img src="assets/data_examples.png"></p>
</div>

## Baseline Models
<div align="justify">
<p>The baseline models for three different tasks (semantic segmentation, UDA, and image translation) on the OpenEarthMap-SAR benchmark dataset are provided as follows.</p>

### Semantic Segmentation
<p>Yon can download all the pre-trained basleine models for the semantic segmentation task from <a href="https://drive.google.com/drive/folders/138jNYpgIey8uNOVLoW6p9dA1A5dhr8_K?usp=drive_link">here</a>.</p>
<table align="center">
    
    <tr align="center">
        <th>Method</th>
        <th>Modality</th> 
	    <th>Labelling Scenario</th> 
	    <th>mIoU</th>
        <th>Pretrained</th> 
    </tr>
    <tr align="center">
        <td rowspan="9">U-Net with EfficientNet-B4</td>
        <td rowspan="3"> Optical </td> 
	    <td align="left"> Pseudo labels</td> 
	    <td> 56.56 </td> 
	    <td> <a href="https://drive.google.com/file/d/1tsbFhc0ukHggocr7dAzbRs6FMPFnQiMG/view?usp=drive_link">Download</a> </td> 
    </tr>
     <tr align="center">
	    <td align="left"> Pseudo labels + 5 real labels per region</td> 
	    <td> 58.19 </td> 
	    <td> <a href="https://drive.google.com/file/d/1hyHlC6stjt4OQ9BWXIv4OP87YkOLpQHr/view?usp=drive_link">Download</a> </td> 
    </tr>
     <tr align="center">
	    <td align="left"> Only 5 real labels per region</td> 
	    <td> 65.10 </td> 
	    <td> <a href="https://drive.google.com/file/d/16REK9OcnhzrUte68hnBj7SJCqLpt1c_U/view?usp=drive_link">Download</a> </td> 
    </tr>
    
    <tr align="center">
        <td rowspan="3"> SAR </td> 
	    <td align="left"> Pseudo labels</td> 
	    <td> 35.13 </td> 
	    <td> <a href="https://drive.google.com/file/d/1hvgJIWttGg-NloVfeQiXIoFi26RvAUVg/view?usp=drive_link">Download</a> </td> 
    </tr>
     <tr align="center">
	    <td align="left"> Pseudo labels + 5 real labels per region</td> 
	    <td> 36.84 </td> 
	    <td> <a href="https://drive.google.com/file/d/1BSe2ywsFbAcf0CS_7j9_fvWTe5Z89TlC/view?usp=drive_link">Download</a> </td> 
    </tr>
     <tr align="center">
	    <td align="left"> Only 5 real labels per region</td> 
	    <td> 33.86 </td> 
	    <td> <a href="https://drive.google.com/file/d/1Iy_VE6LpoUfOuLFM8V54WTmSSKBOKNwm/view?usp=drive_link">Download</a> </td> 
    </tr>
    
    <tr align="center">
        <td rowspan="3">SAR + Optical</td> 
	    <td align="left">Pseudo labels</td> 
	    <td> 56.32 </td> 
	    <td> <a href="https://drive.google.com/file/d/1AFRU0sreULHzb2fCCTag8Jjx6XxJUUiw/view?usp=drive_link">Download</a> </td> 
    </tr>
    <tr align="center">
	    <td align="left">Pseudo labels + 5 real labels per region</td> 
	    <td> 49.79 </td> 
	    <td> <a href="https://drive.google.com/file/d/18Y_NHJ78zT5lOL5h_muCNtrivu6aLl4V/view?usp=drive_link">Download</a> </td> 
    </tr>
    <tr align="center">
	    <td align="left"> Only 5 real labels per region</td> 
	    <td> 60.16 </td> 
	    <td> <a href="https://drive.google.com/file/d/1raL8qVwSSraqYBcvHO06yHHpw-vCkD_a/view?usp=drive_link">Download</a> </td> 
    </tr>
</table>
The code for the <a href="https://arxiv.org/abs/1505.04597">U-Net</a> with <a href="https://arxiv.org/abs/1905.11946">EfficientNet-B4</a> is borrowed from the <a href="https://github.com/qubvel-org/segmentation_models.pytorch">Segmentation Models Pytorch</a>. Thanks to the authors for making their code publicly available.
<p>&nbsp;</p>

### UDA Semantic Segmentation
<p>Country-wise UDA task with France as the source domain and Japan and USA as target domains. 
Yon can download all the pre-trained basleine models for the UDA semantic segmentation task from <a href="https://drive.google.com/drive/folders/10LB26t1yw6T1uFPMwQBsRc5oUjWbAYJ4?usp=drive_link">here</a>.</p>
<table align="center">
    <tr align="center">
        <th>Method</th>
        <th>Modality</th> 
        <th>UDA Setup</th> 
	    <th>Labelling Scenario</th> 
	    <th>mIoU</th>
        <th>Pretrained</th> 
    </tr>
    
    <tr align="center">
        <td rowspan="12"><a href="https://openaccess.thecvf.com/content/CVPR2022/papers/Hoyer_DAFormer_Improving_Network_Architectures_and_Training_Strategies_for_Domain-Adaptive_Semantic_CVPR_2022_paper.pdf">DAFormer</a></td>
        <td rowspan="4"> Optical </td> 
	    <td align="left" rowspan="2"> France-to-USA </td> 
	    <td align="left"> Pseudo labels</td> 
	    <td> 39.38 </td> 
	    <td> <a href="https://drive.google.com/file/d/1YLgYgTPcsnrsfh7OF0z1jjcETsfve-S5/view?usp=drive_link">Download</a> </td> 
    </tr>
    <tr align="center">
	    <td align="left"> Pseudo labels + 5 real labels per region</td> 
	    <td> 39.60 </td> 
	    <td> <a href="https://drive.google.com/file/d/1Ya8xiaFNbw7zh2SgHMHRxy4CKGf4vSCI/view?usp=drive_link">Download</a> </td> 
    </tr>
    <tr align="center">
	    <td align="left" rowspan="2"> France-to-Japan </td> 
	    <td align="left"> Pseudo labels</td> 
	    <td> 44.35 </td> 
	    <td> <a href="https://drive.google.com/file/d/1aiGa2yQ4EmalVFRpvaPEXdq4DqkjjIOP/view?usp=drive_link">Download</a> </td> 
    </tr>
    <tr align="center">
	    <td align="left"> Pseudo labels + 5 real labels per region</td> 
	    <td> 43.74 </td> 
	    <td> <a href="https://drive.google.com/file/d/1uSkQgX_2SHwzESSyxZ1KlA_NAbrqFtsq/view?usp=drive_link">Download</a> </td> 
    </tr>
    
   <tr align="center">
        <td rowspan="4"> SAR </td> 
	    <td align="left" rowspan="2"> France-to-USA </td> 
	    <td align="left"> Pseudo labels</td> 
	    <td> 19.35 </td> 
	    <td> <a href="https://drive.google.com/file/d/1757Jxecp5SVaXuYNMK0zYqfmjQE73Inr/view?usp=drive_link">Download</a> </td> 
    </tr>
    <tr align="center">
	    <td align="left"> Pseudo labels + 5 real labels per region</td> 
	    <td> 21.26 </td> 
	    <td> <a href="https://drive.google.com/file/d/1OGX4xhr32gzuAYmZt8rlMipmnbk6LXVS/view?usp=drive_link">Download</a> </td> 
    </tr>
    <tr align="center">
	    <td align="left" rowspan="2"> France-to-Japan </td> 
	    <td align="left"> Pseudo labels</td> 
	    <td> 13.49 </td> 
	    <td> <a href="https://drive.google.com/file/d/1GkYX7zVCcmfmF7liWTCFIuJ50XoHO_a6/view?usp=drive_link">Download</a> </td> 
    </tr>
    <tr align="center">
	    <td align="left"> Pseudo labels + 5 real labels per region</td> 
	    <td> 12.50 </td> 
	    <td> <a href="https://drive.google.com/file/d/1Sc5HWtYDCZpWvcre8W0Tm-IufZLaAri4/view?usp=drive_link">Download</a> </td> 
    </tr>
    
    <tr align="center">
        <td rowspan="4">SAR + Optical</td> 
	    <td align="left" rowspan="2"> France-to-USA </td> 
	    <td align="left"> Pseudo labels</td> 
	    <td> 40.45 </td> 
	    <td> <a href="https://drive.google.com/file/d/1xluCoIT-pMGR4AgmK0tMZIjGdj4YfWXi/view?usp=drive_link">Download</a> </td> 
    </tr>
    <tr align="center">
	    <td align="left"> Pseudo labels + 5 real labels per region</td> 
	    <td> 37.21 </td> 
	    <td> <a href="https://drive.google.com/file/d/1LRUQrMuzXbksluw1DFGVJy2kXSPYcwbB/view?usp=drive_link">Download</a> </td> 
    </tr>
    <tr align="center">
	    <td align="left" rowspan="2"> France-to-Japan </td> 
	    <td align="left"> Pseudo labels</td> 
	    <td> 40.03 </td> 
	    <td> <a href="https://drive.google.com/file/d/1MyZ7tMjy4NaDm3IAfKoJuOD8Ch4VitU9/view?usp=drive_link">Download</a> </td> 
    </tr>
    <tr align="center">
	    <td align="left"> Pseudo labels + 5 real labels per region</td> 
	    <td> 39.97 </td> 
	    <td> <a href="https://drive.google.com/file/d/1LGqMu7NRTiSojKvWfiFLKJfQXEfirZf5/view?usp=drive_link">Download</a> </td> 
    </tr>
</table>
The code for the UDA semantic segmentation task is borrowed from the <a href="https://github.com/lhoyer/DAFormer">DAFormer</a>. Thanks to the authors for making their code publicly available.
<p>&nbsp;</p>

### Image Translation
<p>Yon can download all the pre-trained basleine models for the image translation task from <a href="https://drive.google.com/drive/folders/1TQIglCWyOOEMLf9pwYHVoetffaCvnwmw?usp=drive_link">here</a>.</p>
<table align="center">
    <tr align="center">
        <th>Method</th>
        <th>Task</th> 
        <th>Sub-Task</th> 
	    <th>Modality</th> 
	    <th>PSNR</th>
	    <th>SSIM</th>
        <th>Pretrained</th> 
    </tr>
    
    <tr align="center">
        <td rowspan="6"><a href="https://arxiv.org/abs/2007.15651">CUT</a></td>
	    <td rowspan="3" align="left"> RGB Generation </td> 
        <td align="left"> Pseudo Semantic2RGB </td> 
	    <td align="left"> Pseudo Semantic Map</td> 
	    <td> 4.62 </td> 
	    <td> 0.2799 </td> 
	    <td> <a href="https://drive.google.com/drive/folders/1H-yENaO-LrpwXuN3sG0ksAKuBfn5Ca-K?usp=drive_link">Download</a> </td> 
    </tr>
    <tr align="center">
        <td align="left"> Real Semantic2RGB </td> 
	    <td align="left"> Real Semantic Map </td> 
	    <td> 14.57 </td> 
	    <td> 0.2141 </td> 
	    <td> <a href="https://drive.google.com/drive/folders/10xPg1L2WLHnw_ji-INrsLH-HwwOUHG0D?usp=drive_link">Download</a> </td> 
    </tr>
    <tr align="center">
        <td align="left"> SAR2RGB </td> 
	    <td align="left"> SAR </td> 
	    <td> 12.47 </td> 
	    <td> 0.1075 </td> 
	    <td> <a href="https://drive.google.com/drive/folders/1fsABom2c4lwflLdrVXLp6cUcIV9vZvW7?usp=drive_link">Download</a> </td> 
    </tr>
    
       <tr align="center">
	    <td rowspan="3" align="left"> SAR Generation </td> 
        <td align="left"> Pseudo Semantic2SAR </td> 
	    <td align="left"> Pseudo Semantic Map</td> 
	    <td> 9.48 </td> 
	    <td> 0.0213 </td> 
	    <td> <a href="https://drive.google.com/drive/folders/1at81X_upRHHMZZgIwknF96XfwBZF3ooo?usp=drive_link">Download</a> </td> 
    </tr>
    <tr align="center">
        <td align="left"> Real Semantic2SAR </td> 
	    <td align="left"> Real Semantic Map </td> 
	    <td> 9.27 </td> 
	    <td> 0.0239 </td> 
	    <td> <a href="https://drive.google.com/drive/folders/1jhITJcppFZdU85Wf6gP3xQ1XBYcwt8Ys?usp=drive_link">Download</a> </td> 
    </tr>
    <tr align="center">
        <td align="left"> RGB2SAR </td> 
	    <td align="left"> RGB </td> 
	    <td> 9.18 </td> 
	    <td> 0.0292 </td> 
	    <td> <a href="https://drive.google.com/drive/folders/1pbUq3zFyCXDKsvW4rWYsLi4ctLZ6pn9I?usp=drive_link">Download</a> </td> 
    </tr>
</table>
The code for the image translation task is borrowed from the <a href="https://github.com/taesungp/contrastive-unpaired-translation">Contrastive Unpaired Translation (CUT)</a>. Thanks to the authors for making their code publicly available.
<p>&nbsp;</p>

## Citation
<div align="justify">
For any scientific publication using this data, the following paper should be cited:
<pre style="white-space: pre-wrap; white-space: -moz-pre-wrap; white-space: -pre-wrap; white-space: -o-pre-wrap; word-wrap: break-word;">
@article{xia2025openearthmap,
  title   = {OpenEarthMap-SAR: A Benchmark Synthetic Aperture Radar Dataset for Global High-Resolution Land Cover Mapping},
  author  = {Xia, Junshi and Chen, Hongruixuan and Broni-Bediako, Clifford and Wei, Yimin and Song, Jian and Yokoya, Naoto},
  journal = {arXiv preprint arXiv:2501.10891},
  year    = {2025}
}
</pre>
</div>

## Acknowledgements
<div align="justify">
We are most grateful to the authors of the following works for making their source codes publicly available:
<ul>
    <li><a href="https://github.com/qubvel/segmentation_models.pytorch?tab=readme-ov-file">Semantic Segmentation PyTorch</a></li>
    <li><a href="https://github.com/lhoyer/DAFormer">DAFormer</a></li>
    <li><a href="https://github.com/wangkaihong/CDAC">CDAC</a></li>
    <li><a href="https://github.com/junyanz/BicycleGAN">BicycleGAN</a></li>
    <li><a href="https://github.com/taesungp/contrastive-unpaired-translation">CUT</a></li>
    <li><a href="https://github.com/junyanz/pytorch-CycleGAN-and-pix2pix">CycleGAN-Pix2Pix</a></li>
    <li><a href="https://github.com/Mid-Push/Decent">Decent</a></li>
    <li><a href="https://github.com/alpc91/NICE-GAN-pytorch">NICE-GAN</a></li>
    <li><a href="https://github.com/NVlabs/SPADE">SPADE</a></li>
    <li><a href="https://github.com/cyclomon/UNSB">UNSB</a></li>
</ul>
</div>

OpenEarthMap-SAR

Explore Similar Repositories

Was this content helpful?

Recent Feedback

Download README