arXivDaily arXiv每日学术速递 周一至周五更新
​通知:联邦节假日调休:哥伦布日 / 原住民日(Columbus Day / Indigenous Peoples' Day),今日(13日)无更新。
arXiv 2502.11859cs.CVcs.CL

定义与评估视觉语言模型的基本空间能力:来自心理测量学的视角

Defining and Evaluating Visual Language Models' Basic Spatial Abilities: A Perspective from Psychometrics

School of Architecture, Tsinghua University(清华大学建筑学院);Department of Electronic Engineering, Tsinghua University(清华大学电子工程系);BNRist, Tsinghua University(清华大学BNRist)

Wenrui Xu, Dalin Lyu, Weihang Wang, Jie Feng, Chen Gao, Yong Li

AI总结:

本研究从心理测量学视角定义并评估视觉语言模型的五种基本空间能力,通过九个实验基准测试13个主流模型,发现其与人类差距显著,并揭示了模型层级结构、规模效应及干预局限,为空间智能评估和具身AI发展提供了诊断工具与路线图。

摘要:

多元智能理论强调了认知能力的层级性。为了推进空间人工智能的发展,我们开创性地提出一个心理测量学框架,定义了视觉语言模型(VLMs)中的五种基本空间能力(BSAs):空间感知、空间关系、空间定向、心理旋转和空间可视化。通过九个经过验证的心理测量实验对13个主流视觉语言模型进行基准测试,揭示了其与人类之间的显著差距(平均得分24.95对比68.38),并得出三个关键发现:1)视觉语言模型反映了人类的层级结构(在二维定向方面最强,在三维旋转方面最弱),且各基本空间能力相互独立(皮尔逊相关系数r<0.4);2)较小的模型如Qwen2-VL-7B超越了较大的模型,其中Qwen领先(30.82),而InternVL2落后(19.6);3)诸如思维链(0.100的准确率提升)和5样本训练(0.259的提升)等干预措施显示出受架构约束的限制。识别出的障碍包括几何编码薄弱和缺乏动态模拟。通过将心理测量学的基本空间能力与视觉语言模型的能力联系起来,我们为空间智能评估提供了一个诊断工具包,为具身人工智能开发提供了方法论基础,并为实现类人空间智能提供了一条基于认知科学的路线图。

英文摘要:

The Theory of Multiple Intelligences underscores the hierarchical nature of cognitive capabilities. To advance Spatial Artificial Intelligence, we pioneer a psychometric framework defining five Basic Spatial Abilities (BSAs) in Visual Language Models (VLMs): Spatial Perception, Spatial Relation, Spatial Orientation, Mental Rotation, and Spatial Visualization. Benchmarking 13 mainstream VLMs through nine validated psychometric experiments reveals significant gaps versus humans (average score 24.95 vs. 68.38), with three key findings: 1) VLMs mirror human hierarchies (strongest in 2D orientation, weakest in 3D rotation) with independent BSAs (Pearson's r<0.4); 2) Smaller models such as Qwen2-VL-7B surpass larger counterparts, with Qwen leading (30.82) and InternVL2 lagging (19.6); 3) Interventions like chain-of-thought (0.100 accuracy gain) and 5-shot training (0.259 improvement) show limits from architectural constraints. Identified barriers include weak geometry encoding and missing dynamic simulation. By linking psychometric BSAs to VLM capabilities, we provide a diagnostic toolkit for spatial intelligence evaluation, methodological foundations for embodied AI development, and a cognitive science-informed roadmap for achieving human-like spatial intelligence.

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。

↑