BUILDING-GUIDED PSEUDO-LABEL LEARNING FOR CROSS-MODAL BUILDING DAMAGE MAPPING

Jiepan Li, He Huang, Yu Sheng, Yujun Guo, Wei He ∗ · 2025

Read BUILDING-GUIDED PSEUDO-LABEL LEARNING FOR CROSS-MODAL BUILDING DAMAGE MAPPING by Jiepan Li, He Huang, Yu Sheng, Yujun Guo, Wei He ∗ (2025) in the Tower of Babel, a free online library of public-domain works preserved in clean, searchable text. No sign-up. The full text is served as a plain-text file at a predictable URL, where text is present: https://antarctic-labs.com/tower-of-babel/arxiv-2505-04941.txt.

Description

Jiepan Li and co-authors win in 2025 a data contest with a building-guided method that maps damage from mixed image types. Damage maps from before-and-after images guide rescue and recovery plans. Mixing light photos from before with radar images from after makes the task hard.

Models first pull building outlines from pre-disaster photos, boosted by fusion and test tricks. A change model then sorts damage using true labels plus fake labels cleaned by low doubt. Building outlines from step one guide the fake labels in step two, which cuts noise. Top overlap scores place the entry first in the contest field.

Catalog record

CREATORJiepan Li, He Huang, Yu Sheng, Yujun Guo, Wei He ∗
YEAR2025
COLLECTIONPAPERS
FORMATTXT
FILE SIZE17 KB
VERSION1.0
LICENSEhttp://creativecommons.org/licenses/by/4.0/
RIGHTSCC BY
SOURCEarXiv

Source: https://arxiv.org/abs/2505-04941

Browse the full library · About the Tower of Babel