发表机构
The Chinese University of Hong Kong（香港中文大学）
机构由 AI 辅助整理，请以论文原文为准。AI 中文总结
MatLoom提出一种紧凑分层程序语言，利用预训练语言模型生成材质，通过解析器修复和预览批评优化，无需微调即在提示对齐和用户偏好上超越扩散基线。
AI 中文摘要
材质生成不仅应产生外观，还应产生构建外观的规则。我们引入了MatLoom，一种紧凑的、面向层的语言，用于使用预训练语言模型进行文本到材质的生成。每个程序组合了具有alpha遮罩的层，其共享的空间表达式定义了覆盖范围和基于物理的渲染（PBR）通道，使得图案、颜色和凹凸之间的依赖关系变得明确。一个独立的解释器将程序评估为材质贴图，而源代码保留命名字段和层参数以供后续编辑。无需针对特定任务的微调，我们的流程使用解析器引导的修复和基于预览的批评来修改材质设计，然后在保持每个候选的其余源代码固定的同时搜索噪声种子。在一个由141个提示组成的精选基准上，使用六个骨干网络进行评估，我们最佳配置在四个平面布局提示对齐指标上均获得了比三个扩散基线更高的平均分数。其初始程序在批评或种子搜索之前，平均BLIPScore已经超过了所有三个基线。保留的程序在跨骨干网络汇总时，中位长度为21行。在一项涉及30名参与者和20个提示的盲法四路比较中，我们的渲染获得了59.2%的选择，而最受青睐的基线为19.3%。因此，紧凑的可执行程序提供了一种生成与提示对齐的材质的方法，同时将其构建保留为资产的一部分。
英文摘要
Material generation should produce not only an appearance, but also the rules that construct it. We introduce MatLoom, a compact, layer-oriented language for text-to-material generation with pretrained language models. Each program composes alpha-masked layers whose shared spatial expressions define coverage and physically based rendering (PBR) channels, making dependencies between patterns, color, and relief explicit. A standalone interpreter evaluates the program into material maps, while the source retains named fields and layer parameters for subsequent authoring. Without task-specific fine-tuning, our pipeline uses parser-guided repair and preview-based critique to revise material designs, then searches noise seeds while keeping each candidate's remaining source fixed. On a curated benchmark of 141 prompts evaluated with six backbones, our best-performing configuration achieves higher mean scores than three diffusion baselines on all four flat-layout prompt-alignment metrics. Its initial programs already exceed all three baselines on mean BLIPScore, before critique or seed search. Retained programs have a median length of 21 lines when pooled across backbones. In a blind four-way comparison involving 30 participants and 20 prompts, our renders receive 59.2% of choices, compared with 19.3% for the most-preferred baseline. Compact executable programs thus offer a way to generate prompt-aligned materials while retaining their construction as part of the asset.
Comments27 pages, 8 figures
