arXiv cs.AI· Qingqi Song, Yuzhou Chen, Lingqian Hu, Shenhao Wang·· 12 小时前AI 评分31
DualControlNet：用生成式 AI 从文本提示合成土地利用图与卫星影像
Generative AI for urban planning: synthesizing land-use maps with satellite imagery via DualControlNet
AI 导读
DualControlNet 是一个可联合生成高分辨率土地利用图与卫星影像的生成式土地利用模型，基于 Orlando 和 Chicago 超 4 万对瓦片训练。相比单模态基线，其土地利用多样性一致性提升 62.8%、建筑覆盖率一致性提升 16.3%，跨模态空间一致性提升逾一倍、视觉质量提升三倍。模型支持多维控制并响应“高密度开发”等文本提示，降低城市未来情景探索门槛。
正文 · 原文
Abstract:Various land-use models characterize land-use patterns and facilitate scenario planning. However, existing models are largely predictive, forecasting transitions based on numeric data without synthesizing new land-use maps from simple text prompts. Meanwhile, generative AI in urban systems remains unimodal, weakly controllable, and evaluated mostly by generic image quality rather than domain-specific metrics. To address these gaps, we introduce DualControlNet, a generative land-use model that jointly synthesizes high-resolution land-use maps and satellite imagery, addressing multimodality, controllability, and verification challenges. Trained on over 40,000 paired tiles from Orlando and Chicago, DualControlNet is evaluated using land-use, spatial consistency, and image quality metrics. Relative to unimodal baselines, DualControlNet improves land-use diversity consistency by 62.8% and building coverage ratio consistency by 16.3%, more than doubles cross-modal spatial consistency, and triples visual quality. Scenario-based experiments demonstrate that DualControlNet supports sustainable development through multidimensional controls, increasing density and diversity, and responding to text prompts like "high-density development." Overall, this work advances land-use modeling via a multimodal, controllable, and verifiable generative framework, lowering technical barriers and enabling intuitive exploration of urban futures.
| Subjects: | Computer Vision and Pattern Recognition (cs.CV); Artificial Intelligence (cs.AI) |
| Cite as: | arXiv:2611.05859 [cs.CV] |
| (or arXiv:2611.05859v1 [cs.CV] for this version) | |
| https://doi.org/10.48550/arXiv.2611.05859 arXiv-issued DOI via DataCite (pending registration) |
Submission history
From: Qingqi Song [view email]
[v1]
Wed, 4 Nov 2026 02:49:55 UTC (8,809 KB)
来源：arXiv cs.AI · arxiv.org