Five points to check when comparing visual perception in humans and machines.

Five points to check when comparing visual perception in humans and machines.
复制标题

在比较人类和机器的视觉感知时要检查的五点。

DOI:
10.1167/jov.21.3.16
复制
发表时间:
2021-03-01
期刊:
影响因子:
1.8
通讯作者:
Bethge M
Bethge M
中科院分区:
医学4区
文献类型:
--
作者:
Funke CM;Borowski J;Stosio K;Brendel W;Wallis TSA;Bethge M

文献摘要

参考文献

被引文献

相似文献

随着机器在复杂识别任务中的表现上升到人类水平,越来越多的工作是针对比较人类和机器的信息处理。这些研究是一个令人兴奋的机会,可以通过研究一个系统来了解另一个系统。在这里,我们提出了关于如何设计、实施和解释实验的想法,以便在比较人类和机器感知时充分支持机制的研究。我们通过三个案例研究来演示和应用这些想法。第一个案例研究显示了人为偏见如何影响对结果的解释,以及几种分析工具可以帮助克服这种人为参考点。在第二个案例研究中,我们强调了视觉推理任务中必要机制和充分机制之间的区别。因此,我们表明,与之前的建议相反,反馈机制可能不是问题任务所必需的。第三个案例研究强调了调整实验条件的重要性。我们发现,在调整实验以使人类和机器之间的条件更公平时,先前观察到的物体识别差异并不成立。在介绍人类和机器视觉推理比较研究的清单时,我们希望强调如何克服设计和推理中的潜在陷阱。
With the rise of machines to human-level performance in complex recognition tasks, a growing amount of work is directed toward comparing information processing in humans and machines. These studies are an exciting chance to learn about one system by studying the other. Here, we propose ideas on how to design, conduct, and interpret experiments such that they adequately support the investigation of mechanisms when comparing human and machine perception. We demonstrate and apply these ideas through three case studies. The first case study shows how human bias can affect the interpretation of results and that several analytic tools can help to overcome this human reference point. In the second case study, we highlight the difference between necessary and sufficient mechanisms in visual reasoning tasks. Thereby, we show that contrary to previous suggestions, feedback mechanisms might not be necessary for the tasks in question. The third case study highlights the importance of aligning experimental conditions. We find that a previously observed difference in object recognition does not hold when adapting the experiment to make conditions more equitable between humans and machines. In presenting a checklist for comparative studies of visual reasoning in humans and machines, we hope to highlight how to overcome potential pitfalls in design and inference.
DOI: 10.1073/pnas.1109168108
发表时间: 2011-10-25
影响因子: 11.1
作者:
Fleuret, Francois;Li, Ting;Geman, Donald
通讯作者: Geman, Donald
DOI: 10.1007/bf00344251
发表时间: 1980-01-01
影响因子: 1.9
作者:
FUKUSHIMA, K
通讯作者: FUKUSHIMA, K
DOI: 10.7554/elife.55978
发表时间: 2020-09-02
期刊: ELIFE
影响因子: 7.7
作者:
Dujmovic, Marin;Malhotra, Gaurav;Bowers, Jeffrey S.
通讯作者: Bowers, Jeffrey S.
DOI: 10.1371/journal.pcbi.1006897
发表时间: 2019-04-01
影响因子: 4.3
作者:
Cadena, Santiago A.;Denfield, George H.;Ecker, Alexander S.
通讯作者: Ecker, Alexander S.
DOI: 10.1016/j.neuron.2012.01.010
发表时间: 2012-02-09
期刊: Neuron
影响因子: 16.2
作者:
DiCarlo JJ;Zoccolan D;Rust NC
通讯作者: Rust NC