InfiniBench：面向视觉空间推理的可定制场景复杂度无限基准生成
InfiniBench: Infinite Benchmarking for Visual Spatial Reasoning with Customizable Scene Complexity
InfiniBench 提供了一种可参数化控制场景复杂度的自动化基准生成方法，视觉研究者可用它系统性地评测 VLM 在测量、视角转换和时空追踪等空间推理任务上的失效模式，并借助其基于聚类的布局优化与任务感知相机轨迹优化生成高复杂度视频输入。
原文依据 InfiniBench, a fully automated, customizable and user-friendly benchmark generator that can synthesize a theoretically infinite variety of 3D scenes with parameterized control on scene complexity.
摘要译文
由模型翻译 · 术语统一 · 请对照原文核对关键结论现代视觉语言模型（VLM）被期望具备处理多样场景复杂度的空间推理能力，但由于缺乏既多样、可扩展又完全可定制的基准，评估此类能力十分困难。现有基准对场景复杂度的可定制性有限，且无法隔离和分析 VLM 在特定空间条件下的具体失效模式。为填补这一空白，本文提出 InfiniBench，一个全自动、可定制且用户友好的基准生成器，可合成理论上无限多种具有参数化场景复杂度控制的 3D 场景，而非针对不同场景复杂度分别呈现基准。InfiniBench 独特地将自然语言场景描述转化为具有复杂且物理合理 3D 布局的照片级真实视频。这通过三项关键创新实现：1）基于 LLM 的智能体框架，从场景描述中迭代细化程序化场景约束；2）灵活的基于聚类的布局优化器，可生成程序化方法此前难以处理的密集杂乱场景；3）任务感知的相机轨迹优化方法，将场景渲染为具有完整物体覆盖的视频作为 VLM 输入。实验表明，InfiniBench 在提示词保真度和物理合理性方面优于最优的程序化及基于 LLM 的 3D 生成方法，尤其在高复杂度场景中。我们进一步通过为测量、视角转换和时空追踪等代表性空间推理任务生成基准，展示了 InfiniBench 的实用性。InfiniBench 的源代码见 https://github.com/pittisl/infinibench，部分由 InfiniBench 生成的示例基准见 https://huggingface.co/datasets/Haoming645/infinibench。
英文原文摘要 1956 字符
Modern vision-language models (VLMs) are expected to have abilities of spatial reasoning with diverse scene complexities, but evaluating such abilities is difficult due to the lack of benchmarks that are not only diverse and scalable but also fully customizable. Existing benchmarks offer limited customizability over the scene complexity and are incapable of isolating and analyzing specific VLM failure modes under distinct spatial conditions. To address this gap, instead of individually presenting benchmarks for different scene complexities, in this paper we present InfiniBench, a fully automated, customizable and user-friendly benchmark generator that can synthesize a theoretically infinite variety of 3D scenes with parameterized control on scene complexity. InfiniBench uniquely translates scene descriptions in natural language into photo-realistic videos with complex and physically plausible 3D layouts. This is achieved through three key innovations: 1) a LLM-based agentic framework that iteratively refines procedural scene constraints from scene descriptions; 2) a flexible cluster-based layout optimizer that generates dense and cluttered scenes previously intractable for procedural methods; and 3) a task-aware camera trajectory optimization method that renders scenes into videos with full object coverage as VLM input. Experiments demonstrate that InfiniBench outperforms state-of-the-art procedural and LLM-based 3D generation methods in prompt fidelity and physical plausibility, especially in high-complexity scenarios. We further showcased the usefulness of InfiniBench, by generating benchmarks for representative spatial reasoning tasks including measurement, perspective-taking and spatiotemporal tracking. The source codes of InfiniBench are available at https://github.com/pittisl/infinibench, and some sample benchmarks being generated by InfiniBench are available at https://huggingface.co/datasets/Haoming645/infinibench.