LoveDA Semantic Segmentation

Organized by kingdrone - Current server time: April 25, 2025, 5:50 p.m. UTC

Current

First phase
Oct. 1, 2021, midnight UTC

End

Competition Ends
Never

LoveDA Semantic Segmentation Challenge

LoveDA Semantic Segmentation Challenge is organized by Junjue Wang, RSIDEA group, Wuhan University.

This contest aims to promote research of large-scale land-cover mapping and deep learning applied to high spatial resolution satellite images.

Instructions:

  1. Develop your semantic segmentation models on the available Train and Validation sets.
  2. Use your optimized model to obtain results based on the available Test images.
  3. Submit your results on this platform and you will get your Test score.

The official benchmark code has been released at Github. Please feel free to develop your model and I'm look forward to your exciting results. If you have any questions, please contact me: kingdrone@whu.edu.cn

Evaluation

Performance is assessed using the pixel-wise mean Intersection over Union (mIoU). The evaluation code is available at Github.

Submission format

The format of submitted zip is as follows:

Result.zip
- 4191.png
- 4192.png
- 4193.png
...
- 5986.png

Classes indexes

Background - 0
Building - 1
Road - 2
Water - 3
Barren - 4
Forest - 5
Agriculture - 6

Notice: The Result.zip must be exactly the same as the file name in Test (1796 images). Same as the test image, each PNG has 1024 x 1024 pixels. The evaluation script will automatically ignore invalid pixels in the test images.

Terms and Conditions

Participants of this challenge acknowledge that they have read and agree to the following Contest Terms and Conditions:

    • The owners of the data and of the copyright on the data are RSIDEA, Wuhan University.
    • Use of the Google Earth images must respect the "Google Earth" terms of use.
    • All images and their associated annotations in LoveDA can be used for academic purposes only,

      but any commercial use is prohibited.

    • Any scientific publication using the data shall refer to the following papers:
      @inproceedings{wang2021loveda,
              title={Love{DA}: A Remote Sensing Land-Cover Dataset for Domain Adaptive Semantic Segmentation},
              author={Junjue Wang and Zhuo Zheng and Ailong Ma and Xiaoyan Lu and Yanfei Zhong},
              booktitle={Proceedings of the Neural Information Processing Systems Track on Datasets and Benchmarks},
              editor = {J. Vanschoren and S. Yeung},
              year={2021},
              volume = {1},
              pages = {},
              url={https://datasets-benchmarks-proceedings.neurips.cc/paper/2021/file/4e732ced3463d06de0ca9a15b6153677-Paper-round2.pdf}
          }
      @article{wang2022lovecs,
          title = {Cross-sensor domain adaptation for high spatial resolution urban land-cover mapping: From airborne to spaceborne imagery},
          journal = {Remote Sensing of Environment},
          volume = {277},
          pages = {113058},
          year = {2022},
          issn = {0034-4257},
          doi = {https://doi.org/10.1016/j.rse.2022.113058},
          url = {https://www.sciencedirect.com/science/article/pii/S0034425722001729},
          author = {Junjue Wang and Ailong Ma and Yanfei Zhong and Zhuo Zheng and Liangpei Zhang},
          }
      

First phase

Start: Oct. 1, 2021, midnight

Competition Ends

Never

You must be logged in to participate in competitions.

Sign In
# Username Score
1 Derivative3 0.57661
2 Ccnu_Has_pfz 0.57391
3 ivica.dimitrovski 0.57363