arXiv cs.CL· Hongxing Li, Jinyue Su, Dingming Li, Wenqi Zhang, Weiming Lu, Jun Xiao, Yueting Zhuang, Yongliang Shen·· 9 天前AI 评分38
SpaceCast-Bench：评估视觉语言模型的预测性空间推理
SpaceCast-Bench: Evaluating Predictive Spatial Reasoning in Vision-Language Models
AI 导读
研究者推出 SpaceCast-Bench，首个直接诊断预测性空间推理的基准，基于 observe-transform-infer 框架，从 182 个真实场景构建 3,862 道题，覆盖静态感知、局部预测、全局预测三个层级共 16 种任务。
正文 · 原文
Abstract:Existing spatial reasoning benchmarks mainly test spatial perception: reading off relations already visible in the input. Yet real-world spatial intelligence demands predictive spatial reasoning: constructing a scene from observations, anticipating how an intervention changes it, and reasoning about the unseen outcome. We introduce SpaceCast-Bench, the first benchmark to directly and diagnostically evaluate this capability. Built around an observe-transform-infer framework, its 3,862 questions from 182 real-world scenes span 16 task types at three levels: static perception, local prediction, and global prediction, progressively requiring scene understanding, spatial state updating, and relational inference over unobserved outcomes. Evaluating 21 models exposes a stark gap: the strongest model reaches only 58.0% against 87.2% human performance, while spatially specialized models remain near random chance. Controlled analyses further reveal that bridge views are critical for integrating distributed observations, and that explicit 3D evidence benefits models more reliably than generated outcome images or videos. Fine-tuning on our programmatically generated data lifts Qwen3-VL-4B from 34.0% to 65.7% with macro-average gains across six out-of-domain benchmarks.
| Comments: | Code: this https URL Dataset: this https URL |
| Subjects: | Computer Vision and Pattern Recognition (cs.CV); Computation and Language (cs.CL) |
| Cite as: | arXiv:2610.12402 [cs.CV] |
| (or arXiv:2610.12402v1 [cs.CV] for this version) | |
| https://doi.org/10.48550/arXiv.2610.12402 arXiv-issued DOI via DataCite (pending registration) |
Submission history
From: Hongxing Li [view email]
[v1]
Thu, 8 Oct 2026 17:44:17 UTC (6,601 KB)
来源：arXiv cs.CL · arxiv.org