Using deep neural networks to address the how of object recognition

Author:

Ayzenberg VladislavORCID,Blauch NicholasORCID,Behrmann Marlene

Abstract

Recently, we drew on evidence from neuroimaging, computational modelling, and neuropsychology to suggest that global shape information may be computed in the dorsal visual pathway and transmitted to the ventral pathway to support object recognition. In their commentary, Goodale and Milner argued that this framework was inconsistent with the response profile of a single patient, RV, who had bilateral dorsal lesions. They found that RV could discriminate simple geometric figures from one another but had difficulty grasping these very same stimuli – a finding consistent with their characterization of the dorsal pathway as a system for processing visuomotor information. Goodale and Milner further argued that because the stimuli they used were matched along many visual dimensions, RV could only have discriminated them on the basis of global shape. Given that RV had lesions to the dorsal pathway, Goodale and Milner took their results to mean that the dorsal pathway is not involved in global shape perception. Here we evaluated this assumption by asking whether deep neural network models (DNNs), which have difficulty representing global shape, could also discriminate between these stimuli.

Publisher

Center for Open Science

Cited by 1 articles. 订阅此论文施引文献 订阅此论文施引文献,注册后可以免费订阅5篇论文的施引文献,订阅后可以查看论文全部施引文献

同舟云学术

1.学者识别学者识别

2.学术分析学术分析

3.人才评估人才评估

"同舟云学术"是以全球学者为主线,采集、加工和组织学术论文而形成的新型学术文献查询和分析系统,可以对全球学者进行文献检索和人才价值评估。用户可以通过关注某些学科领域的顶尖人物而持续追踪该领域的学科进展和研究前沿。经过近期的数据扩容,当前同舟云学术共收录了国内外主流学术期刊6万余种,收集的期刊论文及会议论文总量共计约1.5亿篇,并以每天添加12000余篇中外论文的速度递增。我们也可以为用户提供个性化、定制化的学者数据。欢迎来电咨询!咨询电话:010-8811{复制后删除}0370

www.globalauthorid.com

TOP

Copyright © 2019-2024 北京同舟云网络信息技术有限公司
京公网安备11010802033243号  京ICP备18003416号-3