CD-Mamba: Cloud Detection with Long-Range Spatial Dependency Modeling

• arXiv
• The Biome dataset was split using 4-fold cross-validation, and only the Red, Green, Blue, and Near-Infrared bands were used.
  • Mask Download link
  • Biome Dataset

Remote sensing images are frequently obscured by cloud cover, posing significant challenges to data integrity and reliability. Effective cloud detection requires addressing both short-range spatial redundancies and long-range atmospheric similarities among cloud patches. Convolutional neural networks are effective at capturing local spatial dependencies, while Mamba has strong capabilities in modeling long-range dependencies. To fully leverage both local spatial relations and long-range dependencies, we propose CD-Mamba, a hybrid model that integrates convolution and Mamba’s state-space modeling into a unified cloud detection network. CD-Mamba is designed to comprehensively capture pixel-wise textural details and long-term patch-wise dependencies for cloud detection. This design enables CD-Mamba to manage both pixel-wise interactions and extensive patch-wise dependencies simultaneously, improving detection accuracy across diverse spatial scales. Extensive experiments validate the effectiveness of CD-Mamba and demonstrate its superior performance over existing methods.

Experiment

Train

# Replace the package in your Mamba env lib/python3.10/site-packages/ 
# with the one from folder mamba_ssm.
cd f01
python traincloudmamba.py

Citation

We appreciate it if you cite the following paper:

@Article{xueJARS2025,
  author  = {Tianxiang Xue and Jiayi Zhao and Jingsheng Li and Changlu Chen and Kun Zhan},
  journal = {Journal of Applied Remote Sensing},
  title   = {{CD-Mamba}: Cloud detection with long-range spatial dependency modeling},
  year    = {2025},
  volume  = {19},
}

Contact

https://kunzhan.github.io/

If you have any questions, feel free to contact me. (Email: ice.echo#gmail.com)