Research Article

LightFloodNet: A lightweight CBAM-guided deep learning architecture with depthwise separable convolutions for SAR-based flood extent mapping

Volume: 8 July 3, 2026

LightFloodNet: A lightweight CBAM-guided deep learning architecture with depthwise separable convolutions for SAR-based flood extent mapping

Abstract

Rapid and accurate flood extent mapping is critical for disaster response and mitigation. Synthetic Aperture Radar (SAR) imagery from Sentinel-1 enables all-weather, all-day flood monitoring; however, existing deep learning approaches often rely on architectures with millions of parameters, which may constrain their use in time-sensitive or resource-limited flood-mapping workflows. We propose LightFloodNet, a lightweight encoder-decoder network that integrates established CBAM attention mechanisms and depthwise separable convolutions within a U-Net-inspired framework as a computationally efficient candidate for SAR flood segmentation. On the geographically unseen test regions of the Sen1Floods11 benchmark, the proposed model achieves an IoU of 0.5399 and precision of 0.7554 using only 1.57 million parameters, approximately five times fewer than a standard U-Net baseline. It maintains comparable segmentation performance while attaining higher precision (+0.104) and a 29-fold reduction in computational cost relative to the same baseline. A systematic ablation study shows that Tversky Loss mainly improves recall, CBAM primarily shifts predictions toward higher precision, and Test-Time Augmentation provides small but consistent inference gains. The results indicate that the proposed architecture is a promising candidate for near-real-time SAR flood mapping in resource-constrained settings, pending hardware-specific validation.

Keywords

Thanks

I am grateful for the opportunity to submit my manuscript to your esteemed journal. On behalf of this submission, I would like to extend my sincere thanks to you, the editorial office, and the reviewers in advance for the time, effort, and careful consideration devoted to the review process.

References

  1. Bonafilia, D., Tellman, B., Anderson, T., & Issenberg, E. (2020). Sen1Floods11: A georeferenced dataset to train and test deep learning flood algorithms for Sentinel-1. Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition Workshops (CVPRW), 835–845. https://doi.org/10.1109/CVPRW50498.2020.00113
  2. Buslaev, A., Iglovikov, V. I., Khvedchenya, E., Parinov, A., Druzhinin, M., & Kalinin, A. A. (2020). Albumentations: Fast and flexible image augmentations. Information, 11(2), 125. https://doi.org/10.3390/info11020125
  3. Chen, L.-C., Zhu, Y., Papandreou, G., Schroff, F., & Adam, H. (2018). Encoder-decoder with atrous separable convolution for semantic image segmentation. Proceedings of the European Conference on Computer Vision (ECCV), 833–851. https://doi.org/10.1007/978-3-030-01234-2_49
  4. Diakogiannis, F. I., Waldner, F., Caccetta, P., & Wu, C. (2020). ResUNet-a: A deep learning framework for semantic segmentation of remotely sensed data. ISPRS Journal of Photogrammetry and Remote Sensing, 162, 94–114. https://doi.org/10.1016/j.isprsjprs.2020.01.013
  5. Howard, A. G., Zhu, M., Chen, B., Kalenichenko, D., Wang, W., Weyand, T., Andreetto, M., & Adam, H. (2017). MobileNets: Efficient convolutional neural networks for mobile vision applications. arXiv preprint arXiv:1704.04861. https://arxiv.org/abs/1704.04861
  6. Katiyar, V., Tamkuan, N., & Nagai, M. (2021). Near-real-time flood mapping using off-the-shelf models with SAR imagery and deep learning. Remote Sensing, 13(12), 2334. https://doi.org/10.3390/rs13122334
  7. Li, Y., Martinis, S., Wieland, M., Schlaffer, S., & Natsuaki, R. (2019). Urban flood mapping using SAR intensity and interferometric coherence via Bayesian network fusion. Remote Sensing, 11(19), 2231. https://doi.org/10.3390/rs11192231
  8. Lin, T.-Y., Goyal, P., Girshick, R., He, K., & Dollár, P. (2017). Focal loss for dense object detection. Proceedings of the IEEE International Conference on Computer Vision (ICCV), 2980–2988. https://doi.org/10.1109/ICCV.2017.324

Details

Primary Language

English

Subjects

Image Processing, Artificial Intelligence (Other)

Journal Section

Research Article

Publication Date

July 3, 2026

Submission Date

May 8, 2026

Acceptance Date

June 17, 2026

Published in Issue

Year 2026 Volume: 8

APA
Kınalıoğlu, İ. H. (2026). LightFloodNet: A lightweight CBAM-guided deep learning architecture with depthwise separable convolutions for SAR-based flood extent mapping. Turkish Journal of Remote Sensing, 8. https://doi.org/10.51489/tuzal.1947416
AMA
1.Kınalıoğlu İH. LightFloodNet: A lightweight CBAM-guided deep learning architecture with depthwise separable convolutions for SAR-based flood extent mapping. TJRS. 2026;8. doi:10.51489/tuzal.1947416
Chicago
Kınalıoğlu, İsmail Hakkı. 2026. “LightFloodNet: A Lightweight CBAM-Guided Deep Learning Architecture With Depthwise Separable Convolutions for SAR-Based Flood Extent Mapping”. Turkish Journal of Remote Sensing 8 (July). https://doi.org/10.51489/tuzal.1947416.
EndNote
Kınalıoğlu İH (July 1, 2026) LightFloodNet: A lightweight CBAM-guided deep learning architecture with depthwise separable convolutions for SAR-based flood extent mapping. Turkish Journal of Remote Sensing 8
IEEE
[1]İ. H. Kınalıoğlu, “LightFloodNet: A lightweight CBAM-guided deep learning architecture with depthwise separable convolutions for SAR-based flood extent mapping”, TJRS, vol. 8, July 2026, doi: 10.51489/tuzal.1947416.
ISNAD
Kınalıoğlu, İsmail Hakkı. “LightFloodNet: A Lightweight CBAM-Guided Deep Learning Architecture With Depthwise Separable Convolutions for SAR-Based Flood Extent Mapping”. Turkish Journal of Remote Sensing 8 (July 1, 2026). https://doi.org/10.51489/tuzal.1947416.
JAMA
1.Kınalıoğlu İH. LightFloodNet: A lightweight CBAM-guided deep learning architecture with depthwise separable convolutions for SAR-based flood extent mapping. TJRS. 2026;8. doi:10.51489/tuzal.1947416.
MLA
Kınalıoğlu, İsmail Hakkı. “LightFloodNet: A Lightweight CBAM-Guided Deep Learning Architecture With Depthwise Separable Convolutions for SAR-Based Flood Extent Mapping”. Turkish Journal of Remote Sensing, vol. 8, July 2026, doi:10.51489/tuzal.1947416.
Vancouver
1.İsmail Hakkı Kınalıoğlu. LightFloodNet: A lightweight CBAM-guided deep learning architecture with depthwise separable convolutions for SAR-based flood extent mapping. TJRS. 2026 Jul. 1;8. doi:10.51489/tuzal.1947416

 SCImago Journal & Country Rank             Flag Counter