Dictionaries for Sparse Representation Modeling

Dictionaries for Sparse Representation Modeling

ID:39772531

大小:467.58 KB

页数:13页

时间:2019-07-11

Dictionaries for Sparse Representation Modeling_第1页
Dictionaries for Sparse Representation Modeling_第2页
Dictionaries for Sparse Representation Modeling_第3页
Dictionaries for Sparse Representation Modeling_第4页
Dictionaries for Sparse Representation Modeling_第5页
资源描述:

《Dictionaries for Sparse Representation Modeling》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、INVITEDPAPERDictionariesforSparseRepresentationModelingDigitalsamplingcandisplaysignals,anditshouldbepossibletoexposealargepartofthedesiredsignalinformationwithonlyalimitedsignalsample.ByRonRubinstein,StudentMemberIEEE,AlfredM.Bruckstein,MemberIEEE,andMichaelElad,SeniorMemberIEEEABSTRAC

2、T

3、Sparseandredundantrepresentationmodelingoftime.Thisrepresentation,whileconvenientforthepur-dataassumesanabilitytodescribesignalsaslinearcombina-posesofdisplayorplayback,ismostlyinefficientforanal-tionsofafewatomsfromapre-specifieddictionary.Assuch,ysistasks.Signalprocessingtechniquesc

4、ommonlyrequirethechoiceofthedictionarythatsparsifiesthesignalsiscrucialmoremeaningfulrepresentationswhichcapturetheusefulforthesuccessofthismodel.Ingeneral,thechoiceofapropercharacteristicsofthesignalVforrecognition,therepre-dictionarycanbedoneusingoneoftwoways:i)buildingasentationshoul

5、dhighlightsalientfeatures;fordenoising,sparsifyingdictionarybasedonamathematicalmodelofthetherepresentationshouldefficientlyseparatesignalanddata,orii)learningadictionarytoperformbestonatrainingnoise;andforcompression,therepresentationshouldset.Inthispaperwedescribetheevolutionofthesetw

6、ocapturealargepartofthesignalwithonlyafewcoeffi-paradigms.Asmanifestationsofthefirstapproach,wecovercients.Interestingly,inmanycasestheseseeminglydiffer-topicssuchaswavelets,waveletpackets,contourlets,andentgoalsalign,sharingacoredesireforsimplification.curvelets,allaimingtoexploit1-Dan

7、d2-DmathematicalmodelsRepresentingasignalinvolvesthechoiceofadictionary,forconstructingeffectivedictionariesforsignalsandimages.whichisthesetofelementarysignalsVoratomsVusedtoDictionarylearningtakesadifferentroute,attachingthedecomposethesignal.Whenthedictionaryformsabasis,dictionarytoa

8、setofexamplesitissupposedtoserve.Fromeverysignalisuniquelyrepresentedasthelinearcombi-theseminalworkofFieldandOlshausen,throughtheMOD,thenationofthedictionaryatoms.InthesimplestcasetheK-SVD,theGeneralizedPCAandothers,thispapersurveysthedictionaryisorthogonal,andtherepresentatio

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。