PyTorch implementation of SMCF-Net for bi-temporal SAR flood segmentation, with training code, configurations, and dataset splits.
-
Updated
Oct 8, 2026 - Python
PyTorch implementation of SMCF-Net for bi-temporal SAR flood segmentation, with training code, configurations, and dataset splits.
A Streamlit application for flood segmentation using Sentinel-1 SAR imagery, featuring SAR preprocessing and U-Net based inference for flood extent mapping.
Terrain-aware multimodal GeoAI framework for flood segmentation using Sentinel-1 SAR imagery and deep learning.
Source code for "Cross-Resolution Domain Shift: From U-Net Encoder–Decoder to the Prithvi Hybrid Model" | Gunawan, A. A. S., Kam, R. M.
Open-source PyTorch tools for radar, optical, and terrain-based flood-water segmentation, uncertainty analysis, and event-level robustness evaluation.
A research paper that answers the question "How much does the performance of a SAR-based flood-segmentation model decrease when it is trained on some Sen1Floods11 regions or events and tested on a geographically unseen region or event, compared with a conventional random train–test split?"
To associate your repository with the flood-segmentation topic, visit your repo's landing page and select "manage topics."