Questions pursued through code, experiments, and answers that required better questions.
Per-building damage assessment from pre- and post-disaster satellite images on xBD, with Mask R-CNN building detection and a fine-tuned, satellite-pretrained DINOv3 ViT-L.
Recall on the xBD test split with known building footprints: 97% no damage, 61% minor, 73% major, 84% destroyed.
