GreenScapes: Optimized Vegetation Configurations in Cities
Abstract
Vegetation plays a central role in cities, serving functional roles such as balancing shade and sky exposure, shaping the microclimate, carbon sequestration, and providing acoustic screening. Furthermore, aesthetic choices in landscape styles (formal, modernist, picturesque) are important for citizens' well-being and should be considered in the design of urban landscapes. In this paper, we present a novel method for configuring vegetation — ranging from trees to flower beds — in 3D urban scenes. Our method uses individual lots as input and employs a conditioned Vision-Language Model (VLM) to generate a top-down rendering of vegetation, including lawns, flowers, shrubs, hedges, and trees. This rendering is then converted into a 3D scene by detecting individual plants and selecting appropriate 3D assets. To obtain optimal placements for assets, we define a refinement loop that uses the VLM-generated propositions as an initial condition to perform a low-dimensional set of world-space transformations, along with discrete vegetation asset selections via CMA-ES. To define the objective function for the optimization, we evaluate the generated 3D scenes to derive the physical properties needed to simulate shade and path-based acoustic exposure indices. Furthermore, we render in-scene views in each optimization step to compute novel aesthetic metric scores, enabling the generation of aesthetic as well as functional vegetation layouts for urban scenes.
Method
Overview of our pipeline for prompt-driven park vegetation design. Inputs are a user text prompt (a) and OSM-derived site context (b). An image-to-image constrained editor generates multiple top-down layout proposals (c), which are converted to a metric top-down representation (d) and processed by instance extraction (e) to obtain candidate pedestrian viewpoints (f) and vegetation instance positions and coarse types (g). These instances are projected into world coordinates and assembled into a 3D park scene (h). We evaluate candidate scenes by rendering multiple in-scene views to compute perceptual descriptors and biomass-related quantities (i) and by running physics-inspired field models for functional measures such as shade and sound (j). CMA-ES then refines continuous instance parameters and triggers discrete species substitutions under a Gaussian-mixture asset prior using the combined perceptual and functional objectives (k), yielding the final optimized park layout (l).
Optimization
CMA-ES refinement under three optimization objectives. From left to right: Sky View Factor (SVF), aesthetic fitness, and path-based acoustic exposure. Each panel compares the initial VLM-proposed layout (before) with the result after optimization (after). The three runs are time-aligned so all panels advance through their respective optimizations together.
BibTeX
@article{GreenScapes2026,
title={GreenScapes: Optimized Vegetation Configurations in Cities},
author={Bartkowiak, Patryk and Marlewski, Marcin and W{\'o}jcik, Pawe{\l} and Lee, Jae Joong and Michels, Dominik L. and Benes, Bedrich and Pirk, S{\"o}ren and Palubicki, Wojciech},
note={To appear at SIGGRAPH Asia 2026},
year={2026}
}