EUVIMEDEUVIMED entdecken
Open Access · CC BY 4.0

Cross-modal distillation for flood extent mapping

Shubhika Garg, Ben Feinstein, Shahar Timnat, Vishal Batchu, Gideon Dror, Adi Gerzi Rosenthal, Varun Gulshan

Environmental Data Science · 2023 · 2

DOI: 10.1017/eds.2023.34 ↗

Kurz gefasst

Fachartikel in der Zeitschrift Environmental Data Science, online erschienen am 07.11.2023. Verfasst von Shubhika Garg und 6 weiteren Autor:innen. Die Arbeit wurde bisher 19-mal zitiert (Crossref) und nennt 55 Literaturangaben. Der Volltext ist frei zugänglich (Open Access, CC BY 4.0).

AbstractBibliografische AngabenZitierenVerwandte Studien

Abstract (Originalsprache)

The increasing intensity and frequency of floods is one of the many consequences of our changing climate. In this work, we explore ML techniques that improve the flood detection module of an operational early flood warning system. Our method exploits an unlabeled dataset of paired multi-spectral and synthetic aperture radar (SAR) imagery to reduce the labeling requirements of a purely supervised learning method. Prior works have used unlabeled data by creating weak labels out of them. However, from our experiments, we noticed that such a model still ends up learning the label mistakes in those weak labels. Motivated by knowledge distillation and semi-supervised learning, we explore the use of a teacher to train a student with the help of a small hand-labeled dataset and a large unlabeled dataset. Unlike the conventional self-distillation setup, we propose a cross-modal distillation framework that transfers supervision from a teacher trained on richer modality (multi-spectral images) to a student model trained on SAR imagery. The trained models are then tested on the Sen1Floods11 dataset. Our model outperforms the Sen1Floods11 baseline model trained on the weak-labeled SAR imagery by an absolute margin of $ 6.53\% $ intersection over union (IoU) on the test split.

Zitieren

APA

Garg, S., Feinstein, B., Timnat, S., Batchu, V., Dror, G., Rosenthal, A. G., & Gulshan, V. (2023). Cross-modal distillation for flood extent mapping. Environmental Data Science, 2. https://doi.org/10.1017/eds.2023.34

Vancouver

Garg S, Feinstein B, Timnat S, Batchu V, Dror G, Rosenthal AG, et al. Cross-modal distillation for flood extent mapping. Environmental Data Science. 2023;2. doi:10.1017/eds.2023.34

BibTeX
@article{Garg2023,
  author = {Garg, Shubhika and Feinstein, Ben and Timnat, Shahar and Batchu, Vishal and Dror, Gideon and Rosenthal, Adi Gerzi and Gulshan, Varun},
  title = {Cross-modal distillation for flood extent mapping},
  journal = {Environmental Data Science},
  year = {2023},
  volume = {2},
  doi = {10.1017/eds.2023.34}
}
Kostenlos starten →