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 2409.05862cs.CV
收藏

评估人类与图像模型中的多视图对象一致性

Evaluating Multiview Object Consistency in Humans and Image Models

University of California, Berkeley(加利福尼亚大学伯克利分校);Massachusetts Institute of Technology(麻省理工学院)

Tyler Bonnen, Stephanie Fu, Yutong Bai, Thomas O'Connell, Yoni Friedman, Nancy Kanwisher, Joshua B. Tenenbaum, Alexei A. Efros

AI总结:

本文构建多视图三维形状推断基准,采集人类选择、反应时间与注视数据,并评估DINOv2、MAE和CLIP,发现人类显著优于模型且在困难试验中投入更多加工。

摘要:

我们引入了一个基准,用于在三维形状推断任务上直接评估人类观察者与视觉模型之间的对齐程度。我们借鉴认知科学中的一项实验设计,该设计要求对物体形状进行零样本视觉推断:给定一组图像,参与者需要识别哪些图像包含相同或不同的物体,即使视角存在显著变化。我们采用多样化的图像来源,其中既包括常见物体(例如椅子),也包括抽象形状(即通过程序生成的“无意义”物体)。在构建了2000多个独特图像集之后,我们将这些任务布置给人类参与者,从500多名参与者那里收集了3.5万次试验的行为数据。这包括显式选择行为以及反应时间和注视数据等中间测量。随后,我们评估了常见视觉模型(例如DINOv2、MAE和CLIP)的性能。我们发现,人类以显著优势胜过所有模型。通过使用多尺度评估方法,我们识别出模型与人类之间潜在的异同:尽管人类与模型的性能存在相关性,但人类会在具有挑战性的试验上投入更多时间和加工。所有图像、数据和代码均可通过我们的项目页面获取。

英文摘要:

We introduce a benchmark to directly evaluate the alignment between human observers and vision models on a 3D shape inference task. We leverage an experimental design from the cognitive sciences which requires zero-shot visual inferences about object shape: given a set of images, participants identify which contain the same/different objects, despite considerable viewpoint variation. We draw from a diverse range of images that include common objects (e.g., chairs) as well as abstract shapes (i.e., procedurally generated `nonsense' objects). After constructing over 2000 unique image sets, we administer these tasks to human participants, collecting 35K trials of behavioral data from over 500 participants. This includes explicit choice behaviors as well as intermediate measures, such as reaction time and gaze data. We then evaluate the performance of common vision models (e.g., DINOv2, MAE, CLIP). We find that humans outperform all models by a wide margin. Using a multi-scale evaluation approach, we identify underlying similarities and differences between models and humans: while human-model performance is correlated, humans allocate more time/processing on challenging trials. All images, data, and code can be accessed via our project page.

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。

↑