返回
Multi-Stage Multi-Recursive-Input Fully Convolutional Networks for Neuronal Boundary Detection

Multi-Stage Multi-Recursive-Input Fully Convolutional Networks for Neuronal Boundary Detection

ID：40721769

大小：1.31 MB

页数：10页

时间：2019-08-06

Multi-Stage Multi-Recursive-Input Fully Convolutional Networks for Neuronal Boundary Detection_第1页
Multi-Stage Multi-Recursive-Input Fully Convolutional Networks for Neuronal Boundary Detection_第2页
Multi-Stage Multi-Recursive-Input Fully Convolutional Networks for Neuronal Boundary Detection_第3页
Multi-Stage Multi-Recursive-Input Fully Convolutional Networks for Neuronal Boundary Detection_第4页
Multi-Stage Multi-Recursive-Input Fully Convolutional Networks for Neuronal Boundary Detection_第5页
Multi-Stage Multi-Recursive-Input Fully Convolutional Networks for Neuronal Boundary Detection_第6页
Multi-Stage Multi-Recursive-Input Fully Convolutional Networks for Neuronal Boundary Detection_第7页
Multi-Stage Multi-Recursive-Input Fully Convolutional Networks for Neuronal Boundary Detection_第8页
Multi-Stage Multi-Recursive-Input Fully Convolutional Networks for Neuronal Boundary Detection_第9页
Multi-Stage Multi-Recursive-Input Fully Convolutional Networks for Neuronal Boundary Detection_第10页
资源描述：

《Multi-Stage Multi-Recursive-Input Fully Convolutional Networks for Neuronal Boundary Detection》由会员上传分享，免费在线阅读，更多相关内容在学术论文-天天文库。

1、Multi-stageMulti-recursive-inputFullyConvolutionalNetworksforNeuronalBoundaryDetectionWeiShen1,2,BinWang1,YuanJiang1∗,YanWang2,AlanYuille21KeyLaboratoryofSpecialtyFiberOpticsandOpticalAccessNetworks,ShanghaiUniversity2DepartmentofComputerScience,JohnsHopkinsUniversitywei.shen@t.shu.edu

2、.cn,{wangbin418,jy9387}@outlook.com,{wyanny.9,alan.l.yuille}@gmail.comAbstractInthefieldofconnectomics,neuroscientistsseektoi-dentifycorticalconnectivitycomprehensively.NeuronalboundarydetectionfromtheElectronMicroscopy(EM)im-agesisoftendonetoassisttheautomaticreconstructionofneuronalci

3、rcuit.ButthesegmentationofEMimagesisachallengingproblem,asitrequiresthedetectortobeable(a)(b)(c)todetectbothfilament-likethinandblob-likethickmem-Figure1.Neuronalstructuresegmentation:anEMimage(a)andthegroundtruthsforitsneuronalboundarydetectionresult(b)andbrane,whilesuppressingtheambig

4、uousintracellularstruc-segmentationresult(c),respectively.ture.Inthispaper,weproposemulti-stagemulti-recursive-inputfullyconvolutionalnetworkstoaddressthisproblem.Themultiplerecursiveinputsforonestage,i.e.,themulti-riousandevenimpractical[13],whichdrivesthedemandplesideoutputswithdiffe

5、rentreceptivefieldsizeslearnedforefficientautomatedneuronalcircuitreconstructionap-fromthelowerstage,providemulti-scalecontextualbound-proaches.aryinformationfortheconsecutivelearning.Thisdesignisbiologically-plausible,asitlikesahumanvisualsystemSerialsectionEMproducesastackof2Dimagesbyc

6、ut-tocomparedifferentpossiblesegmentationsolutionstoad-tingsectionsofbraintissue.Duetotheanisotropicresolu-dresstheambiguousboundaryissue.Ourmulti-stagenet-tionsofin-planeandout-of-plane,mostneuronalcircuitre-worksaretrainedend-to-end.Itachievespromisingre-constructionapproachesfollowt

7、hefollowingpipeline:(1)sultsontwopublicavailableEMsegmentationdatasets,neuronalboundarydetectiononeach2Dimage,(2)neu-themousepiriformcortexdatasetandtheISBI2012EMronalstructuresegmentationbasedonthe2Dboundarydataset.map,and(3)linkingtheneuronalsegmentsacross2Dim-agesintoa3Dreconstruc

当前文档最多预览五页，下载文档查看全文

此文档下载收益归作者所有

当前文档最多预览五页，下载文档查看全文
温馨提示：
1. 部分包含数学公式或PPT动画的文件，查看预览时可能会显示错乱或异常，文件下载后无此问题，请放心下载。
2. 本文档由用户上传，版权归属用户，天天文库负责整理代发布。如果您对本文档版权有争议请及时联系客服。
3. 下载前请仔细阅读文档内容，确认文档内容符合您的需求后进行下载，若出现内容与标题不符可向本站投诉处理。
4. 下载文档时可能由于网络波动等原因无法下载或下载错误，付费完成后未能成功下载的用户请联系客服处理。