Improved MambdaBDA Framework for Robust Building Damage Assessment Across Disaster Domains
This work addresses domain shift and class imbalance in building damage assessment for disaster response, but it is incremental as it builds on an existing successful model.
The paper tackled the problem of reliable post-disaster building damage assessment from satellite imagery by enhancing the MambaBDA framework with modular components like Focal Loss, Attention Gates, and an Alignment Module, resulting in performance gains of 0.8% to 5% in-domain and up to 27% on unseen disasters.
Reliable post-disaster building damage assessment (BDA) from satellite imagery is hindered by severe class imbalance, background clutter, and domain shift across disaster types and geographies. In this work, we address these problems and explore ways to improve the MambaBDA, the BDA network of ChangeMamba architecture, one of the most successful BDA models. The approach enhances the MambaBDA with three modular components: (i) Focal Loss to mitigate class imbalance damage classification, (ii) lightweight Attention Gates to suppress irrelevant context, and (iii) a compact Alignment Module to spatially warp pre-event features toward post-event content before decoding. We experiment on multiple satellite imagery datasets, including xBD, Pakistan Flooding, Turkey Earthquake, and Ida Hurricane, and conduct in-domain and crossdataset tests. The proposed modular enhancements yield consistent improvements over the baseline model, with 0.8% to 5% performance gains in-domain, and up to 27% on unseen disasters. This indicates that the proposed enhancements are especially beneficial for the generalization capability of the system.