气候变化领域集成服务门户(CCPortal): Unsupervised neural network models of the ventral visual stream

CCPortal

DOI	10.1073/pnas.2014196118
	Unsupervised neural network models of the ventral visual stream
	Zhuang C.; Yan S.; Nayebi A.; Schrimpf M.; Frank M.C.; DiCarlo J.J.; Yamins D.L.K.
发表日期	2021
ISSN	00278424
卷号	118 期号:3
英文摘要	Deep neural networks currently provide the best quantitative models of the response patterns of neurons throughout the primate ventral visual stream. However, such networks have remained implausible as a model of the development of the ventral stream, in part because they are trained with supervised methods requiring many more labels than are accessible to infants during development. Here, we report that recent rapid progress in unsupervised learning has largely closed this gap. We find that neural network models learned with deep unsupervised contrastive embedding methods achieve neural prediction accuracy in multiple ventral visual cortical areas that equals or exceeds that of models derived using today’s best supervised methods and that the mapping of these neural network models’ hidden layers is neuroanatomically consistent across the ventral stream. Strikingly, we find that these methods produce brain-like representations even when trained solely with real human child developmental data collected from head-mounted cameras, despite the fact that these datasets are noisy and limited. We also find that semisupervised deep contrastive embeddings can leverage small numbers of labeled examples to produce representations with substantially improved error-pattern consistency to human behavior. Taken together, these results illustrate a use of unsupervised learning to provide a quantitative model of a multiarea cortical brain system and present a strong candidate for a biologically plausible computational theory of primate sensory learning. © 2021 National Academy of Sciences. All rights reserved.
英文关键词	Deep neural networks; Unsupervised algorithms; Ventral visual stream
语种	英语
scopus关键词	adult; Article; brain mapping; computer model; controlled study; deep neural network; developmental stage; embedding; human; infant; learning; learning algorithm; nerve potential; nonhuman; prediction; primate; priority journal; social behavior; task performance; visual cortex
来源期刊	Proceedings of the National Academy of Sciences of the United States of America
文献类型	期刊论文
条目标识符	http://gcip.llas.ac.cn/handle/2XKMVOVA/180991
作者单位	Department of Psychology, Stanford University, Stanford, CA 94305, United States; Department of Computer Science, The University of Texas at Austin, Austin, TX 78712, United States; Neurosciences PhD Program, Stanford University, Stanford, CA 94305, United States; Brain and Cognitive Sciences, Massachusetts Institute of Technology, Cambridge, MA 02139, United States; Department of Computer Science, Stanford University, Stanford, CA 94305, United States; Wu Tsai Neurosciences Institute, Stanford University, Stanford, CA 94305, United States
推荐引用方式 GB/T 7714	Zhuang C.,Yan S.,Nayebi A.,et al. Unsupervised neural network models of the ventral visual stream[J],2021,118(3).
APA	Zhuang C..,Yan S..,Nayebi A..,Schrimpf M..,Frank M.C..,...&Yamins D.L.K..(2021).Unsupervised neural network models of the ventral visual stream.Proceedings of the National Academy of Sciences of the United States of America,118(3).
MLA	Zhuang C.,et al."Unsupervised neural network models of the ventral visual stream".Proceedings of the National Academy of Sciences of the United States of America 118.3(2021).