文档下载 文档下载导航
设为首页 | 加入收藏
搜索 请输入内容:  
 导航当前位置: 文档下载 > 所有分类 > IT/计算机 > 互联网 > Deep Belief Networks for Image Denoising
侵权投诉

Deep Belief Networks for Image Denoising

iscorruptedbynoise).Asaresult,ifaparticularnodeisanoisenode,itshouldhavehigherrelativeactivity.Ontheotherhand,ifitisanimagenode,itshouldhavelowerrelativeactivity.Thistheoryisjusti edbythefactthattheactivationofimagenodesshould

Deep Belief Networks for Image Denoising

Deep Belief Networks for Image Denoising

besameforbothnoiselessanditscorrespondingnoisyimages.ThisprocessisillustratedinFigure3.

Figure3:Relativeactivityof

thelastlayernodescanbecom-

putedbysubtractingthelast

nodesvaluesconstructedbya

noiselessimageanditscorre-

spondingnoisyimage.

Byperformingaboveoperationforallimagesandaveragingonthevaluesofeachnodeinthelastlayer,averagerelativeactivityofthelastlayernodesiscomputed.Thenodesthatstillhavehighaveragerelativeactivityareconsiderednoisenodes.Nowthatthenoisenodesarediscovered,thenextstepistolowertheiractivity.Sincenoisenodesdonotchangemuchwhencleanimagesarefedtothenetwork,wechoosetheiraveragevalueforallcleanimagesastheirneutralvalues(thevalueswhichmakenodesinactive).Finally,thenoisenodesareinactiveandconsequently,anoiselessimagecanbereconstructedasitisshowninFigure4.

Figure4:Nodesinthelast

layeroftheDBNarepresent-

ingthenoiseandthecontents

oftheinputimage.Byreduc-

ingthenoisynodes(graynodes

inthe gureabove)activity,a

noiselessimagewillbecreated.

Rightsideimageisthenoisy

imageandtheleftsideimageis

noiselessreconstruction.

4ExperimentalResults

MNISTdatasetisadatasetofhandwrittendigitsconsistedof60,000imagesfortrainingand10,000imagesfortest.Tomodelnaturalnoise,weaddedadditivewhiteGaussiannoise(AWGN)toimageswithavarianceof0.20.Therefore,ournewdatasetwasconsistedof120,000noisyandcleanimages

上一页第4页

你可能喜欢

热门文档

相关文档

站点地图 | 文档上传 | 侵权投诉 | 手机版
本站所有资源均来自互联网,本站只负责收集和整理,均不承担任何法律责任,如有侵权等其它行为请联系我们.
文档下载 Copyright 2013 doc.guandang.net All Rights Reserved.  -  email