schliessen

Filtern

 

Bibliotheken

Deep Residual Network Predicts Cortical Representation and Organization of Visual Features for Rapid Categorization

The brain represents visual objects with topographic cortical patterns. To address how distributed visual representations enable object categorization, we established predictive encoding models based on a deep residual network, and trained them to predict cortical responses to natural movies. Using... Full description

Journal Title: Sci Rep 2018, Vol.8(1), pp.3752-3752
Main Author: Wen, Haiguang
Other Authors: Shi, Junxing , Chen, Wei , Liu, Zhongming
Format: Electronic Article Electronic Article
Language: English
Subjects:
ID: ISSN: 2045-2322 ; DOI: 10.1038/s41598-018-22160-9
Zum Text:
SendSend as email Add to Book BagAdd to Book Bag
Staff View
recordid: palgrave_j10.1038/s41598-018-22160-9
title: Deep Residual Network Predicts Cortical Representation and Organization of Visual Features for Rapid Categorization
format: Article
creator:
  • Wen, Haiguang
  • Shi, Junxing
  • Chen, Wei
  • Liu, Zhongming
subjects:
  • Biology
ispartof: Sci Rep, 2018, Vol.8(1), pp.3752-3752
description: The brain represents visual objects with topographic cortical patterns. To address how distributed visual representations enable object categorization, we established predictive encoding models based on a deep residual network, and trained them to predict cortical responses to natural movies. Using this predictive model, we mapped human cortical representations to 64,000 visual objects from 80 categories with high throughput and accuracy. Such representations covered both the ventral and dorsal pathways, reflected multiple levels of object features, and preserved semantic relationships between categories. In the entire visual cortex, object representations were organized into three clusters of categories: biological objects, non-biological objects, and background scenes. In a finer scale specific to each cluster, object representations revealed sub-clusters for further categorization. Such hierarchical clustering of category representations was mostly contributed by cortical representations of object features from middle to high levels. In summary, this study demonstrates a useful computational strategy to characterize the cortical organization and representations of visual features for rapid categorization.
language: eng
source:
identifier: ISSN: 2045-2322 ; DOI: 10.1038/s41598-018-22160-9
fulltext: fulltext
issn:
  • 2045-2322
  • 20452322
url: Link


@attributes
ID220769996
RANK0.07
NO1
SEARCH_ENGINEprimo_central_multiple_fe
SEARCH_ENGINE_TYPEPrimo Central Search Engine
LOCALfalse
PrimoNMBib
record
control
sourcerecordid10.1038/s41598-018-22160-9
sourceidpalgrave_j
recordidTN_palgrave_j10.1038/s41598-018-22160-9
sourcesystemOther
pqid2009223019
display
typearticle
titleDeep Residual Network Predicts Cortical Representation and Organization of Visual Features for Rapid Categorization
creatorWen, Haiguang ; Shi, Junxing ; Chen, Wei ; Liu, Zhongming
ispartofSci Rep, 2018, Vol.8(1), pp.3752-3752
identifierISSN: 2045-2322 ; DOI: 10.1038/s41598-018-22160-9
descriptionThe brain represents visual objects with topographic cortical patterns. To address how distributed visual representations enable object categorization, we established predictive encoding models based on a deep residual network, and trained them to predict cortical responses to natural movies. Using this predictive model, we mapped human cortical representations to 64,000 visual objects from 80 categories with high throughput and accuracy. Such representations covered both the ventral and dorsal pathways, reflected multiple levels of object features, and preserved semantic relationships between categories. In the entire visual cortex, object representations were organized into three clusters of categories: biological objects, non-biological objects, and background scenes. In a finer scale specific to each cluster, object representations revealed sub-clusters for further categorization. Such hierarchical clustering of category representations was mostly contributed by cortical representations of object features from middle to high levels. In summary, this study demonstrates a useful computational strategy to characterize the cortical organization and representations of visual features for rapid categorization.
languageeng
source
subjectBiology;
version7
lds50peer_reviewed
links
openurl$$Topenurl_article
openurlfulltext$$Topenurlfull_article
search
creatorcontrib
0Wen, Haiguang
1Shi, Junxing
2Chen, Wei
3Liu, Zhongming
titleDeep Residual Network Predicts Cortical Representation and Organization of Visual Features for Rapid Categorization
descriptionThe brain represents visual objects with topographic cortical patterns. To address how distributed visual representations enable object categorization, we established predictive encoding models based on a deep residual network, and trained them to predict cortical responses to natural movies. Using this predictive model, we mapped human cortical representations to 64,000 visual objects from 80 categories with high throughput and accuracy. Such representations covered both the ventral and dorsal pathways, reflected multiple levels of object features, and preserved semantic relationships between categories. In the entire visual cortex, object representations were organized into three clusters of categories: biological objects, non-biological objects, and background scenes. In a finer scale specific to each cluster, object representations revealed sub-clusters for further categorization. Such hierarchical clustering of category representations was mostly contributed by cortical representations of object features from middle to high levels. In summary, this study demonstrates a useful computational strategy to characterize the cortical organization and representations of visual features for rapid categorization.
general
0English
110.1038/s41598-018-22160-9
2Palgrave Macmillan Journals
sourceidpalgrave_j
recordidpalgrave_j10.1038/s41598-018-22160-9
issn
02045-2322
120452322
rsrctypearticle
creationdate2018
addtitleSci Rep
searchscopepalgrave_j
scopepalgrave_j
lsr30VSR-Enriched:[pages, pqid, orcidid, subject]
sort
titleDeep Residual Network Predicts Cortical Representation and Organization of Visual Features for Rapid Categorization
authorWen, Haiguang ; Shi, Junxing ; Chen, Wei ; Liu, Zhongming
creationdate20181200
facets
frbrgroupid4872967895455976049
frbrtype5
newrecords20180228
languageeng
creationdate2018
collectionPalgrave Macmillan Journals
prefilterarticles
rsrctypearticles
creatorcontrib
0Wen, Haiguang
1Shi, Junxing
2Chen, Wei
3Liu, Zhongming
jtitleSci Rep
toplevelpeer_reviewed
delivery
delcategoryRemote Search Resource
fulltextfulltext
addata
aulast
0Wen
1Shi
2Chen
3Liu
aufirst
0Haiguang
1Junxing
2Wei
3Zhongming
au
0Wen, Haiguang
1Shi, Junxing
2Chen, Wei
3Liu, Zhongming
atitleDeep Residual Network Predicts Cortical Representation and Organization of Visual Features for Rapid Categorization
jtitleSci Rep
risdate201812
volume8
issue1
spage3752
epage3752
issn2045-2322
formatjournal
genrearticle
ristypeJOUR
abstractThe brain represents visual objects with topographic cortical patterns. To address how distributed visual representations enable object categorization, we established predictive encoding models based on a deep residual network, and trained them to predict cortical responses to natural movies. Using this predictive model, we mapped human cortical representations to 64,000 visual objects from 80 categories with high throughput and accuracy. Such representations covered both the ventral and dorsal pathways, reflected multiple levels of object features, and preserved semantic relationships between categories. In the entire visual cortex, object representations were organized into three clusters of categories: biological objects, non-biological objects, and background scenes. In a finer scale specific to each cluster, object representations revealed sub-clusters for further categorization. Such hierarchical clustering of category representations was mostly contributed by cortical representations of object features from middle to high levels. In summary, this study demonstrates a useful computational strategy to characterize the cortical organization and representations of visual features for rapid categorization.
pubNature Publishing Group UK
doi10.1038/s41598-018-22160-9
pages1-17
orcidid0000-0003-3044-3139
date2018-12