AI RESEARCH
Test-Time Adaptation for Height Completion via Self-Supervised ViT Features and Monocular Foundation Models
arXiv CS.CV
•
ArXi:2604.02009v1 Announce Type: new Accurate digital surface models (DSMs) are essential for many geospatial applications, including urban monitoring, environmental analyses, infrastructure management, and change detection. However, large-scale DSMs frequently contain incomplete or outdated regions due to acquisition limitations, reconstruction artifacts, or changes in the built environment. Traditional height completion approaches primarily rely on spatial interpolation or which assume spatial continuity and therefore fail when objects are missing.