FloorplanVLM：一种用于平面图向量化的视觉-语言模型
FloorplanVLM: A Vision-Language Model for Floorplan Vectorization
AI总结：
FloorplanVLM通过像素到序列的范式，实现复杂平面图向量化的高精度与高鲁棒性，其统一框架和渐进训练策略在工程级向量化任务中表现出色。
摘要：
将位图平面图转换为工程级向量图形具有挑战性，因为其拓扑复杂且存在严格的几何约束。为此，我们提出了FloorplanVLM，一个统一的框架，将平面图向量化重新表述为图像条件序列建模任务。与基于像素的方法依赖脆弱的启发式方法或基于查询的Transformer生成碎片化的房间不同，我们的模型直接输出代表全局拓扑的结构化JSON序列。这种'像素到序列'的范式使复杂几何体（如倾斜墙壁和曲线弧）的精确和整体约束满足成为可能。为了支持这种数据需求大的方法，我们引入了一个可扩展的数据引擎：我们构建了一个大规模数据集（Floorplan-2M）和一个高保真子集（Floorplan-HQ-300K）以平衡几何多样性和像素级精度。然后，我们采用渐进训练策略，使用监督微调（SFT）进行结构定位和质量退火，随后使用组相对策略优化（GRPO）进行严格几何对齐。为了标准化复杂布局的评估，我们建立了并开源了FPBench-2K。在该严格基准上的评估表明，FloorplanVLM在结构有效性方面表现出色，实现了92.52%的外部墙IoU，并在非曼哈顿建筑架构上实现了稳健的泛化能力。
英文摘要：
Converting raster floorplans into engineering-grade vector graphics is challenging due to complex topology and strict geometric constraints. To address this, we present FloorplanVLM, a unified framework that reformulates floorplan vectorization as an image-conditioned sequence modeling task. Unlike pixel-based methods that rely on fragile heuristics or query-based transformers that generate fragmented rooms, our model directly outputs structured JSON sequences representing the global topology. This 'pixels-to-sequence' paradigm enables the precise and holistic constraint satisfaction of complex geometries, such as slanted walls and curved arcs. To support this data-hungry approach, we introduce a scalable data engine: we construct a large-scale dataset (Floorplan-2M) and a high-fidelity subset (Floorplan-HQ-300K) to balance geometric diversity and pixel-level precision. We then employ a progressive training strategy, using Supervised Fine-Tuning (SFT) for structural grounding and quality annealing, followed by Group Relative Policy Optimization (GRPO) for strict geometric alignment. To standardize evaluation on complex layouts, we establish and open-source FPBench-2K. Evaluated on this rigorous benchmark, FloorplanVLM demonstrates exceptional structural validity, achieving $\textbf{92.52%}$ external-wall IoU and robust generalization across non-Manhattan architectures.
机构、总结、中文摘要均由AI分析而来，可能存在错误，请以论文正文为准。
