Controllable Generation of Large-Scale 3D Urban Layouts with Semantic and Structural Guidance
This work provides an effective tool for city planning, scene synthesis, and gaming by enabling controllable generation of 3D urban layouts, though it appears incremental as it builds on existing graph-based and image-based methods.
The paper tackled the problem of generating large-scale 3D urban layouts by addressing limitations in existing methods, such as lack of geometric continuity and scalability, and it resulted in a controllable framework that produces valid, large-scale urban models for data-driven planning and design.
Urban modeling is essential for city planning, scene synthesis, and gaming. Existing image-based methods generate diverse layouts but often lack geometric continuity and scalability, while graph-based methods capture structural relations yet overlook parcel semantics. We present a controllable framework for large-scale 3D vector urban layout generation, conditioned on both geometry and semantics. By fusing geometric and semantic attributes, introducing edge weights, and embedding building height in the graph, our method extends 2D layouts to realistic 3D structures. It also enables users to directly control the output by modifying semantic attributes. Experiments show that it produces valid, large-scale urban models, offering an effective tool for data-driven planning and design.