Wildfire risk prediction improves using street views and satellite images
Wildfire-SSR: Geolocation-Aware Multimodal Fusion for Structure-Level Wildfire Risk Assessment
Computer Vision and Pattern Recognition
Summary
Wildfires threaten homes, especially where neighborhoods meet wild areas, but it's hard to tell which houses are most vulnerable using just one type of data. The authors created Wildfire-SSR, a system that combines satellite pictures with street-level photos to better identify risks for each building. This tool uses AI to understand details about buildings from street views and matches them with satellite images, improving the accuracy of wildfire damage predictions. It processes large areas quickly and helps focus efforts on the most vulnerable structures.
What this means in practice
- •For urban planners: Assess neighborhood wildfire risk using combined street and satellite images to prioritize mitigation and resilience planning.
- •For insurance underwriters: Estimate property wildfire vulnerability more accurately by integrating multiple imagery sources for better risk-based insurance pricing.
Authors
Yu-Hsuan Ho, Yiming Xiao, Hao-En Lai, Ali Mostafavi, Samuel D. Brody
Abstract
Structure-level wildfire vulnerability assessment is critical for reducing losses in wildland-urban interface communities, yet existing methods often fail to combine parcel-scale ignition pathways with community-scale geographic coverage. This study introduces Wildfire-SSR (Satellite and Street-view for wildfire Risk), a geolocation-aware multimodal framework that integrates macro-scale satellite context with micro-scale street-view evidence. A Vision-Language Model extracts wildfire-relevant building and site attributes from pre-event street-view panoramas and converts them into structured semantic text. These features are spatially aligned with satellite imagery through a decoupled cross-attention fusion decoder, while a frozen satellite-pretrained DINOv3 encoder captures overhead visual context. Alternative geolocation embedding strategies are evaluated to anchor text-derived features within each satellite patch. The framework is tested on approximately 37,000 structures affected by the 2025 Eaton and Palisades fires in Los Angeles, using wildfire risk labels derived from relative post-event damage patterns to better represent inherent structural vulnerability. Rotary Positional Embedding provides the strongest spatial grounding, achieving a building localization F1 score of 0.6735 and a harmonic mean F1 score of 0.2476 for five-class pixel-level risk classification, outperforming unimodal baselines and an unanchored multimodal configuration. The model also shows a conservative bias in higher-risk classes, reducing severe-risk underestimation, and completes end-to-end assessment of the study area in 37 minutes on a single GPU. Wildfire-SSR reframes cross-view fusion as localized semantic alignment and provides a scalable tool for mitigation prioritization, resilience planning, and insurance-oriented wildfire risk assessment.