Researchers have introduced Any2Building, a novel framework designed to reconstruct detailed 3D building models (Level of Detail 2) from incomplete point cloud data, overhead images, or a combination of both. The system utilizes shared spatial queries grounded in occupancy to integrate geometric and visual information within a common 3D grid. This approach allows the model to learn structural priors across different data modalities and provides spatial guidance for reconstructing missing elements. To support this research, the team also released BuildingWorld-Image, a dataset containing approximately one million paired samples of partial point clouds, overhead images, and LoD2 building models across 18 cities. AI
IMPACT This research advances 3D urban modeling by enabling more accurate and robust building reconstruction from diverse data sources.
RANK_REASON Research paper detailing a new model and dataset for 3D reconstruction. [lever_c_demoted from research: ic=1 ai=1.0]
AI-generated summary · Google Gemini · from 1 sources. How we write summaries →