Building Dataset

Building datasets are crucial for advancing urban analysis, environmental modeling, and architectural design, with current research focusing on creating comprehensive and accurate representations of building exteriors and footprints. This involves leveraging diverse data sources like street view imagery, satellite imagery, LiDAR point clouds, and OpenStreetMap data, often processed using deep learning models such as UNet, transformers, and diffusion models to extract building attributes (height, function, age) and generate realistic 3D models. These improved datasets and associated algorithms enable more accurate urban simulations, improved building energy efficiency predictions, and enhanced understanding of urban morphology and dynamics, ultimately supporting better urban planning and policy decisions.