Summary
It is hard to tell if computer models that look at satellite images to judge building damage after disasters are really assessing each building or just guessing based on nearby buildings. The authors found that many models often rely too much on the damage status of neighboring buildings, missing isolated damaged ones or incorrectly flagging intact buildings in damaged areas. They created an add-on that helps models look at individual buildings more carefully without retraining the whole system. This add-on improves accuracy, fixing many mistakes caused by relying too much on neighbors and spotting isolated damaged buildings better.
What this means in practice
- •For disaster response teams: Provide more reliable building damage maps by reducing errors from neighbourhood assumptions, enabling better targeted relief efforts after disasters.
- •For satellite image analytics companies: Improve existing damage assessment models’ accuracy without retraining by integrating the building-evidence plug-in to reduce false alerts in complex scenes.$Commercial implications: This plug-in can be sold as a software upgrade improving model reliability for clients in insurance or emergency services.
Authors
JianChong Guo, YuTing Wan, Kefan Li, AiLong Ma, YanFei Zhong
Abstract
Building damage assessment (BDA) from very-high-resolution imagery supports disaster response, and deep models now achieve high overall accuracy on public benchmarks. Because damage is spatially clustered, most buildings share the damage level of their neighbours, so overall scores are dominated by such buildings, whereas damaged buildings among intact ones are rare and easily overlooked. A high overall score therefore cannot tell whether a model judges the building itself or relies on its neighbourhood as a shortcut, and this study tests which of the two current models do. We stratify buildings by their agreement with their neighbours, move buildings into new neighbourhoods with their pixels unchanged, and diagnose nine models covering the main paradigms of BDA. All models miss a third to half of the isolated damaged buildings, and in non-flood events they call 73--87% of intact buildings damaged once these are moved into a damaged neighbourhood. We propose a building-evidence plug-in that adds two experts to an existing model without retraining it. One expert sees only the building, and the other sees its surroundings but is trained with seamlessly swapped neighbourhoods so that it learns not to trust them. With each of the nine models, the plug-in maintains or improves overall accuracy, reduces errors on buildings that disagree with their neighbours by 8--18%, and cuts by 40--69% the intact buildings that a damaged neighbourhood turns into damaged ones. It also raises the detection of synthetic isolated damage from 1--6% to 13--39% and carries over to optical--SAR data and to an unseen earthquake. High overall scores thus conceal a systematic reliance on the neighbourhood, which building-level evidence reduces at low cost.