AIFloodSense [1] (a Global Aerial Imagery Dataset for Semantic Segmentation and Understanding of Flooded Environments) provides worldwide coverage and more recent aerial imagery than prior benchmarks. AIFloodSense is an aerial imagery dataset comprising 470 high-resolution images from 230 distinct flood events across 64 countries and six continents.
Annotation supports multiple computer vision tasks:
(i) image classification by environment, camera angle, and continent;
(ii) semantic segmentation of three key visual categories—flood, sky, and buildings; and
(iii) visual question answering (VQA) with task-specific queries.
AIFloodSense dataset establishes a new benchmark for advancing flood-related computer vision research and contributes to the broader effort of building AI tools for climate resilience.
Figure 1: Geographic distribution of flood events in the proposed dataset. Training samples are indicated in blue and test samples in red, plotted on a global map to illustrate spatial coverage.
Figure 2: Samples from the AIFloodSense Dataset.
Table 1:Summary of the dataset distribution. Rows correspond to each continent, and columns report the total number of images, the number of images collected per year (2022–2024), the number of images per environment type (Urban/Peri-urban, Rural), and the number of images per camera angle (Sky presence, Sky absence).
Dataset
AIFloodSense Dataset (470 images) [1] : to be announced very soon
Other Datasets
Flood Area Segmentation Dataset (290 images) [2-3]: https://www.kaggle.com/datasets/faizalkarim/flood-area-segmentation
Flood Semantic Segmentation Dataset (663 images) [2-3]: https://www.kaggle.com/datasets/lihuayang111265/flood-semantic-segmentation-dataset
Real, Synthetic (SD_s) and semi-synthetic (SD_ip) Datasets [4]: https://drive.google.com/file/d/1XLU5tpOONZ1zjdEgWUdpuVVFN2DRZz7c/view?usp=sharing
Related Publications
[1] Georgios Simantiris, Konstantinos Bacharidisa, Apostolos Papanikolaou, Petros Giannakakis, Costas Panagiotakis, AIFloodSense: A Global Aerial Imagery Dataset for Semantic Segmentation and Understanding of Flooded Environments, arXiv preprint arXiv:
[2] Georgios Simantiris and Costas Panagiotakis, Unsupervised Color Based Flood Segmentation in UAV imagery, vol. 16, no 12, Remote Sensing, 2024.
[3] G. Simantiris and C. Panagiotakis, Unsupervised Deep Learning for Flood Segmentation in UAV imagery, 13th IAPR Workshop on Pattern Recognition in Remote Sensing, 2024.
[4] G. Simantiris, K. Bacharidis and C. Panagiotakis, Closing the Domain Gap: Can Pseudo-Labels from Synthetic UAV Data Enable Real-World Flood Segmentation?, 25(12), Sensors, 2025.
[5] https://sites.google.com/site/costaspanagiotakis/research/flood-detection