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 2602.03916cs.CVcs.CEcs.CLcs.LG
收藏

SpatiaLab: 视觉-语言模型能否在真实环境中进行空间推理?

SpatiaLab: Can Vision-Language Models Perform Spatial Reasoning in the Wild?

Computational Intelligence and Operations Laboratory(计算智能与运筹实验室);Shahjalal University of Science and Technology(沙赫jalal科技大学);BRAC University(BRAC大学);North South University(北南大学);Monash University(墨尔本大学);Qatar Computing Research Institute(卡塔尔计算研究院)
Azmine Toushik Wasi, Wahid Faisal, Abdur Rahman, Mahfuz Ahmed Anik, Munem Shahriar, Mohsin Mahmud Topu, Sadia Tasnim Meem, Rahatun Nesa Priti, Sabrina Afroz Mit… 展开作者

Azmine Toushik Wasi, Wahid Faisal, Abdur Rahman, Mahfuz Ahmed Anik, Munem Shahriar, Mohsin Mahmud Topu, Sadia Tasnim Meem, Rahatun Nesa Priti, Sabrina Afroz Mitu, Md. Iqramul Hoque, Shahriyar Zaman Ridoy, Mohammed Eunus Ali, Majd Hawasly, Mohammad Raza, Md Rizwan Parvez

AI总结:

SpatiaLab通过真实场景下的空间推理任务评估视觉-语言模型的能力,揭示其在复杂空间关系、深度感知和3D几何方面的不足。

摘要:

空间推理是人类认知的基本方面,但仍然是当代视觉-语言模型(VLMs)的主要挑战。以往的工作大多依赖于合成或LLM生成的环境,任务设计和谜题样设置有限,无法捕捉现实世界中的复杂性、视觉噪声和多样的空间关系。为此,我们引入SpatiaLab,一个全面的基准,用于评估VLMs在真实、无约束环境中的空间推理能力。SpatiaLab包含1,400个视觉问题-答案对,涵盖六个主要类别:相对位置、深度与遮挡、方向、大小与比例、空间导航和3D几何,每个类别有五个子类别,产生30种不同的任务类型。每个子类别至少包含25个问题,每个主要类别至少包含200个问题,支持多选和开放性评估。在不同最先进的VLMs上进行的实验揭示了空间推理能力与人类之间存在显著差距。在多选设置中,InternVL3.5-72B的准确率为54.93%,而人类为87.57%。在开放性设置中,所有模型的性能下降约10-25%,GPT-5-mini的最高得分为40.93%,而人类为64.93%。这些结果突显了处理复杂空间关系、深度感知、导航和3D几何的关键限制。通过提供多样化的现实世界评估框架,SpatiaLab揭示了VLMs空间推理的关键挑战和机遇,为未来研究提供了一个基准,以引导研究朝着稳健、与人类对齐的空间理解发展。SpatiaLab可在:https://spatialab-reasoning.github.io/访问。

英文摘要:

Spatial reasoning is a fundamental aspect of human cognition, yet it remains a major challenge for contemporary vision-language models (VLMs). Prior work largely relied on synthetic or LLM-generated environments with limited task designs and puzzle-like setups, failing to capture the real-world complexity, visual noise, and diverse spatial relationships that VLMs encounter. To address this, we introduce SpatiaLab, a comprehensive benchmark for evaluating VLMs' spatial reasoning in realistic, unconstrained contexts. SpatiaLab comprises 1,400 visual question-answer pairs across six major categories: Relative Positioning, Depth & Occlusion, Orientation, Size & Scale, Spatial Navigation, and 3D Geometry, each with five subcategories, yielding 30 distinct task types. Each subcategory contains at least 25 questions, and each main category includes at least 200 questions, supporting both multiple-choice and open-ended evaluation. Experiments across diverse state-of-the-art VLMs, including open- and closed-source models, reasoning-focused, and specialized spatial reasoning models, reveal a substantial gap in spatial reasoning capabilities compared with humans. In the multiple-choice setup, InternVL3.5-72B achieves 54.93% accuracy versus 87.57% for humans. In the open-ended setting, all models show a performance drop of around 10-25%, with GPT-5-mini scoring highest at 40.93% versus 64.93% for humans. These results highlight key limitations in handling complex spatial relationships, depth perception, navigation, and 3D geometry. By providing a diverse, real-world evaluation framework, SpatiaLab exposes critical challenges and opportunities for advancing VLMs' spatial reasoning, offering a benchmark to guide future research toward robust, human-aligned spatial understanding. SpatiaLab is available at: https://spatialab-reasoning.github.io/.

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。

↑