Flood Segmentation on Sentinel-1 SAR Imagery with Semi-Supervised Learning
Reviewed by Pithpith:A4PQN63Ropen to challenge →
read the original abstract
Floods wreak havoc throughout the world, causing billions of dollars in damages, and uprooting communities, ecosystems and economies. The NASA Impact Flood Detection competition tasked participants with predicting flooded pixels after training with synthetic aperture radar (SAR) images in a supervised setting. We propose a semi-supervised learning pseudo-labeling scheme that derives confidence estimates from U-Net ensembles, progressively improving accuracy. Concretely, we use a cyclical approach involving multiple stages (1) training an ensemble model of multiple U-Net architectures with the provided high confidence hand-labeled data and, generated pseudo labels or low confidence labels on the entire unlabeled test dataset, and then, (2) filter out quality generated labels and, (3) combine the generated labels with the previously available high confidence hand-labeled dataset. This assimilated dataset is used for the next round of training ensemble models and the cyclical process is repeated until the performance improvement plateaus. We post process our results with Conditional Random Fields. Our approach sets a new state-of-the-art on the Sentinel-1 dataset with 0.7654 IoU, an impressive improvement over the 0.60 IoU baseline. Our method, which we release with all the code and models, can also be used as an open science benchmark for the Sentinel-1 dataset.
This paper has not been read by Pith yet.
Forward citations
Cited by 3 Pith papers
-
Urban Flood Observations: A hand-labeled training and validation dataset of post-flood inundation
UFO is a new publicly available hand-labeled dataset of 215 PlanetScope image chips from 14 urban flood events annotated for inundated and non-inundated areas, validated via segmentation model with 77.3 mean IoU and c...
-
Urban Flood Observations: A hand-labeled training and validation dataset of post-flood inundation
RL-finetuning Qwen3-8B with LLM-as-judge rewards produces a red teaming model that generates effective attacks for novel adversarial goals not seen during training.
-
Beyond Backscatter: AlphaEarth Land-Cover Priors for Rapid SAR Flood Segmentation Across Foundation Backbones
AlphaEarth land-cover priors improve SAR flood segmentation IoU over SAR-only and DEM baselines across CNN and ViT backbones on held-out events like Hurricane Florence.
discussion (0)
Sign in with ORCID, Apple, or X to comment. Anyone can read and Pith papers without signing in.