Research map / Detection and segmentation / CNN features
Multi3Net: Segmenting Flooded Buildings via Fusion of Multiresolution, Multisensor, and Multitemporal Satellite Imagery
Tim G. J. Rudner, Marc Rußwurm, Jakub Fil, Ramona Pelich, Benjamin Bischke, Veronika Kopačková, Piotr Biliński
Read the paperCheck your idea against 137,217 papers
AI-extracted. The measured effects below were read by AI from the paper's full text (no summary yet). Numbers and claims can be wrong or missing context, so check the paper before you cite or build on it.
Measured effects reported in the paper
| What was changed | Where | Measured | Effect |
|---|---|---|---|
| Fuse Sentinel-1, Sentinel-2, and VHR rather than using VHR alone | Multi3Net · Flooded buildings | mIoU: 74.2% -> 75.3% | Improved |
| Fuse Sentinel-1 and Sentinel-2 rather than using Sentinel-1 alone | Multi3Net · Flooded buildings | mIoU: 50.2% -> 59.7% | Improved |
| Use Sentinel-1, Sentinel-2, and VHR instead of VHR alone | Multi3Net · Building footprints | mIoU: 78.9% -> 79.9% | Improved |
| Use Sentinel-1 and Sentinel-2 instead of Sentinel-1 alone | Multi3Net · Building footprints | mIoU: 69.3% -> 76.1% | Improved |
| Replace the U-Net fusion architecture with Multi3Net | Flooded-building segmentation using Sentinel-1, Sentinel-2, and VHR · Flooded buildings | bIoU: 44.2% -> 57.5% | Improved |
| Concatenate pre- and post-disaster images in early fusion rather than use a shared-weight two-stream network or subtract feature activations (ablation) | Multi3Net temporal fusion · Flooded buildings | model performance | No clear effect |
| Use late fusion with added convolutional layers rather than early fusion, encoder feature fusion, sum fusion, or element-wise multiplication (ablation) | Multi3Net multisensor fusion · Building and flooded-building segmentation | model performance | Improved |
Closest papers by meaning
- DisasterM3: A Remote Sensing Vision-Language Dataset for Disaster Damage Assessment and Response (NeurIPS 2025 · similarity 0.85)
- Generalizable Disaster Damage Assessment via Change Detection with Vision Foundation Model (AAAI 2025 · similarity 0.84)
- Spatial-Logic-Aware Weakly Supervised Learning for Flood Mapping on Earth Imagery (AAAI 2024 · similarity 0.83)
- Flood Insights: Integrating Remote and Social Sensing Data for Flood Exposure, Damage, and Urgent Needs Mapping (AAAI 2024 · similarity 0.83)
- MONITRS: Multimodal Observations of Natural Incidents Through Remote Sensing (NeurIPS 2025 · similarity 0.82)
- Kuro Siwo: 33 billion $m^2$ under the water. A global multi-temporal satellite dataset for rapid flood mapping (NeurIPS 2024 · similarity 0.82)
Writing a paper in this area? scime draft is a free LaTeX editor with AI included, real-time co-authors and a check of your references and venue rules before you submit. It also imports OpenAI Prism projects.
