arXivDaily arXiv每日学术速递 周一至周五更新
arXiv周末暂无论文更新,休息一下吧,周末愉快~~
arXiv 2603.26512cs.AI
收藏

CADSmith: 基于程序化几何验证的多智能体CAD生成

CADSmith: Multi-Agent CAD Generation with Programmatic Geometric Validation

Carnegie Mellon University(卡内基梅隆大学)

Jesse Barkley, Rumi Loghmani, Amir Barati Farimani

AI总结:

CADSmith通过多智能体管道生成CadQuery代码,并通过双重嵌套修正循环进行迭代优化,结合精确测量和视觉评估提升CAD生成质量。

摘要:

现有文本到CAD生成方法要么单次通过无几何验证,要么依赖损失性视觉反馈无法解决尺寸误差。我们提出CADSmith,一种多智能体管道,从自然语言生成CadQuery代码,然后通过两个嵌套修正循环进行迭代优化:内循环解决执行错误,外循环基于程序化几何验证。外循环结合OpenCASCADE内核的精确测量(边界框尺寸、体积、实体有效性)与独立视觉语言模型Judge的总体视觉评估。系统使用API文档增强生成而非微调,维护当前数据库作为CAD库的基础。我们在包含100个提示的定制基准上评估,三个难度层级(T1至T3)和三种消融配置。与零样本基线相比,CADSmith实现100%执行率(从95%提升),中位F1分数从0.9707提升至0.9846,中位IoU从0.8085提升至0.9629,均值Chamfer距离从28.37降至0.74,证明闭环优化与程序化几何反馈显著提升LLM生成CAD模型的质量和可靠性。

英文摘要:

Existing methods for text-to-CAD generation either operate in a single pass with no geometric verification or rely on lossy visual feedback that cannot resolve dimensional errors. We present CADSmith, a multi-agent pipeline that generates CadQuery code from natural language. It then undergoes an iterative refinement process through two nested correction loops: an inner loop that resolves execution errors and an outer loop grounded in programmatic geometric validation. The outer loop combines exact measurements from the OpenCASCADE kernel (bounding box dimensions, volume, solid validity) with holistic visual assessment from an independent vision-language model Judge. This provides both the numerical precision and the high-level shape awareness needed to converge on the correct geometry. The system uses retrieval-augmented generation over API documentation rather than fine-tuning, maintaining a current database as the underlying CAD library evolves. We evaluate on a custom benchmark of 100 prompts in three difficulty tiers (T1 through T3) with three ablation configurations. Against a zero-shot baseline, CADSmith achieves a 100% execution rate (up from 95%), improves the median F1 score from 0.9707 to 0.9846, the median IoU from 0.8085 to 0.9629, and reduces the mean Chamfer Distance from 28.37 to 0.74, demonstrating that closed-loop refinement with programmatic geometric feedback substantially improves the quality and reliability of LLM-generated CAD models.

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。

↑