SceneSmith: 面向仿真就绪室内场景的智能体生成
SceneSmith: Agentic Generation of Simulation-Ready Indoor Scenes
AI总结：
提出层次化智能体框架SceneSmith，通过VLM智能体协作从自然语言生成仿真就绪的室内场景，相比先前方法生成3-6倍物体且碰撞率低于2%。
摘要：
仿真已成为大规模训练和评估家用机器人的关键工具，但现有环境未能捕捉真实室内空间的多样性和物理复杂性。当前的场景合成方法生成的房间稀疏布置，缺乏机器人操作所必需的密集杂乱、铰接式家具和物理属性。我们提出了SceneSmith，一个层次化智能体框架，能够从自然语言提示生成仿真就绪的室内环境。SceneSmith通过连续阶段构建场景——从建筑布局到家具放置再到小物体填充——每个阶段都实现为VLM智能体（设计师、评论家和编排者）之间的交互。该框架通过文本到3D合成生成静态物体、数据集检索获取铰接式物体以及物理属性估计，紧密集成了资产生成。SceneSmith生成的物体数量是先前方法的3-6倍，物体间碰撞率低于2%，且96%的物体在物理仿真下保持稳定。在205名参与者参与的用户研究中，与基线相比，平均真实感胜率达到92%，平均提示忠实度胜率达到91%。我们进一步证明了这些环境可用于端到端的自动机器人策略评估流程。
英文摘要：
Simulation has become a key tool for training and evaluating home robots at scale, yet existing environments fail to capture the diversity and physical complexity of real indoor spaces. Current scene synthesis methods produce sparsely furnished rooms that lack the dense clutter, articulated furniture, and physical properties essential for robotic manipulation. We introduce SceneSmith, a hierarchical agentic framework that generates simulation-ready indoor environments from natural language prompts. SceneSmith constructs scenes through successive stages$\unicode{x2013}$from architectural layout to furniture placement to small object population$\unicode{x2013}$each implemented as an interaction among VLM agents: designer, critic, and orchestrator. The framework tightly integrates asset generation through text-to-3D synthesis for static objects, dataset retrieval for articulated objects, and physical property estimation. SceneSmith generates 3-6x more objects than prior methods, with <2% inter-object collisions and 96% of objects remaining stable under physics simulation. In a user study with 205 participants, it achieves 92% average realism and 91% average prompt faithfulness win rates against baselines. We further demonstrate that these environments can be used in an end-to-end pipeline for automatic robot policy evaluation.
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。
