发表机构
Chongqing University（重庆大学）
机构由 AI 辅助整理，请以论文原文为准。AI 中文总结
SG-Layout通过两阶段训练将结构化空间知识融入LLMs，在三类任务中提升了空间推理与几何一致性，尤其适配关系密集的复杂场景。
AI 中文摘要
从自然语言理解和生成空间连贯的布局，对大语言模型（LLMs）而言是一项基础却极具挑战性的任务。现有LLMs常难以捕捉物体间明确的几何关系与结构依赖。为解决该问题，我们提出SG-Layout，一种图引导的布局生成框架，将结构化空间知识明确融入LLMs。SG-Layout遵循两阶段训练范式：（1）图-语言特征对齐阶段，训练关系图编码器与投影器，将场景图嵌入映射至LLMs的语言空间；（2）指令微调阶段，基于LoRA的适配器实现指令驱动布局生成的高效微调，同时保留主干模型参数固定。我们在图像布局生成、室内场景合成及机器人物体重排任务中评估SG-Layout，实验结果显示，相较于紧凑的开源主干模型，SG-Layout提升了空间推理准确率与几何一致性，在关系密集、结构复杂的场景中优势尤为显著，这些结果凸显了图结构化特征对齐对增强可控布局生成的有效性。
英文摘要
Understanding and generating spatially coherent layouts from natural language remains a fundamental yet challenging task for large language models (LLMs). Existing LLMs often struggle to capture explicit geometric relationships and structural dependencies between objects. To address this issue, we propose SG-Layout, a graph-guided layout generation framework that explicitly incorporates structured spatial knowledge into LLMs. SG-Layout follows a two-stage training paradigm: (1) a graph-language feature alignment stage, where a relational graph encoder and a projector are trained to map scene-graph embeddings into the LLM's linguistic space; and (2) an instruction tuning stage, where LoRA-based adapters enable efficient fine-tuning for instruction-driven layout generation while keeping the backbone frozen. We evaluate SG-Layout on image layout generation, indoor scene synthesis and robotic object rearrangement tasks. Experimental results show that SG-Layout improves spatial reasoning accuracy and geometric consistency over the compact open-source backbone, with particularly clear advantages in relation-dense and compositionally complex scenes. These results highlight the effectiveness of graph-structured feature alignment for enhancing controllable layout generation.
Comments16 pages, 5 figures. Accepted at WAICA 2026
