Conflict-aware models improve disaster damage assessments from photos
Trust the View That Sees the Target: Mining Cross-View Conflicts for Reliability-Gated Disaster Damage Assessment
Computer Vision and Pattern Recognition
Summary
After disasters, buildings are checked for damage using both aerial and street-level photos, but these two views don’t always agree. The authors found that in cases where the views conflict, trusting the more reliable view for each building works better than just combining both equally. They created a simple method to decide which photo type to trust based on visibility and confidence. This approach improves accuracy in wildfire and hurricane damage assessments and helps predict damage levels even without labels.
What this means in practice
- •For disaster response teams: Use conflict-based gating to improve accuracy when assessing building damage from aerial and street photos after disasters.
- •For urban planners: Predict areas with higher building damage from conflict patterns in image data without needing damage labels.
Authors
Yifan Yang
Abstract
After a disaster, building damage is assessed from overhead tiles and ground-level photographs, and most methods fuse the two views symmetrically, trusting both equally for every building. This paper focuses on the samples where that assumption fails: the conflict cases, on which two independently trained single-view models disagree. We mine such cases from three paired collections (inspection photographs from the 2025 Eaton wildfire and street-view panoramas from Hurricanes Ian and Milton, each matched to very-high-resolution overhead tiles), where they make up 10-33% of the data. On these samples an oracle that simply trusts the correct view beats every fusion method we tested by 0.37-0.41 accuracy, and the gap survives longer training, calibration, and backbone changes. We recover part of it with a visibility-conditioned reliability gate: a linear model that decides which view to trust from building-visibility features, calibrated per-view confidences, and the disagreement itself. On the wildfire data the gate is the only method that significantly beats calibrated probability averaging (+0.051 on conflicts, p=0.0001) and end-to-end fusion (+0.072, p<10^-4); on the panoramic datasets it matches them. A controlled field-of-view experiment explains why: cropping panoramas toward the building doubles the benefit of fusion, whereas random crops of the same size do not. Finally, the spatial density of conflicts predicts tile-level damage without labels (Spearman r=0.615, p=0.001). Mining conflicts turns "does fusion help?" into "which view should be trusted, where, and why?".