Building Change Detection in Earthquake: A Multi-Scale Interaction Network and A Change Detection Dataset
Researchers have developed a new deep learning model called MSI-Net to improve change detection in remote sensing images, specifically for assessing building damage after earthquakes. This model addresses challenges posed by short imaging intervals and differing angles between temporal images, which can lead to issues like side-looking effects. MSI-Net incorporates joint cross-attention, multi-scale offset calibration, and feature integration modules to enhance feature interaction and alignment, demonstrating superior performance on existing and newly created datasets. AI
IMPACT Enhances the accuracy of post-disaster damage assessment, potentially speeding up emergency response.