TIRAuxCloud: A Thermal Infrared Dataset for Day and Night Cloud Detection

TIRAuxCloud is a multi-modal dataset designed to support all day cloud segmentation, with a particular focus on thermal infrared (TIR) imagery. It combines multispectral observations from Landsat and VIIRS with auxiliary information such as elevation, land cover, meteorological variables, and cloud-free reference imagery. The dataset includes both automatically generated cloud masks and manually annotated samples, enabling the development and evaluation of advanced cloud detection methods.

The full paper describing the dataset, methodology, and benchmarks is available on: https://arxiv.org/abs/2602.21905

---

GitHub Repository

The code associated with this dataset is available on: https://github.com/Orion-AI-Lab/TIRAuxCloud

The repository can be used to reproduce the experiments presented in the paper.

---

Citation

if you use this work please cite:

@misc{apostolakis2026tirauxcloudthermalinfrareddataset,
      title={TIRAuxCloud: A Thermal Infrared Dataset for Day and Night Cloud Detection}, 
      author={Alexis Apostolakis and Vasileios Botsos and Niklas Wölki and Andrea Spichtinger and Nikolaos Ioannis Bountos and Ioannis Papoutsis and Panayiotis Tsanakas},
      year={2026},
      eprint={2602.21905},
      archivePrefix={arXiv},
      primaryClass={cs.CV},
      url={https://arxiv.org/abs/2602.21905}, 
}