参考文献

[1]Pan,Y.F.,Hou,X.,Liu,C.L.:Textlocalizationinnaturalsceneimagesbasedonconditionalrandomfield.In:Proc.ICDAR.(2009)

[2]JADERBERG,M.,SIMONYAN,K.,VEDALDI,A.,ANDZISSERMAN,A.2014.

Readingtextinthewildwithconvolutionalneuralnetworks.arXivpreprintarXiv:1412.1842.

[3]J.Matas,O.Chum,M.Urban,andT.Pajdla,“RobustWideBaselineStereo

fromMaximallyStableExtremalRegions,”Proc.BritishMachineVision

Conf.,pp.384-393,2002.

[4]Neumann,L.,Matas,J.:Real-timescenetextlocalizationandrecognition.In:Proc.CVPR.(2012).

[5]X.-C.Yin,X.Yin,K.Huang,andH.-W.Hao,“Robusttextdetection

innaturalsceneimages,”PatternAnalysisandMachineIntelligence,

IEEETransactionson,vol.36,no.5,pp.970–983,May2014.

[6]M.-C.Sung,B.Jun,H.ChoandD.Kim,“SceneTextDetectionwithRobustCharacterCandidateExtractionMethod”In:Proc.ICDAR.(2015).

[7]YannLeCun,YoshuaBengio&GeoffreyHinton,Deeplearning

doi:10.1038/nature14539

[8]Krizhevsky,A.,Sutskever,I.&Hinton,G.ImageNetclassificationwithdeep

convolutionalneuralnetworks.InProc.AdvancesinNeuralInformation

ProcessingSystems251090–1098(2012).

[9]Bengio,Y.,Lamblin,P.,Popovici,D.&Larochelle,H.Greedylayer-wisetrainingofdeepnetworks.InProc.AdvancesinNeuralInformationProcessingSystems19153–160(2006).

[10]Hubel,D.H.&Wiesel,T.N.Receptivefields,binocularinteraction,andfunctionalarchitectureinthecat’svisualcortex.J.Physiol.160,106–154(1962).

54

[11]Rumelhart,D.E.,Hinton,G.E.&Williams,R.J.Learningrepresentationsbyback-propagatingerrors.Nature323,533–536(1986).

[12]Hinton,G.E.,Osindero,S.&Teh,Y.-W.Afastlearningalgorithmfordeepbeliefnets.NeuralComp.18,1527–1554(2006).

[13]Hinton,G.E.&Salakhutdinov,R.Reducingthedimensionalityofdatawithneuralnetworks.Science313,504–507(2006).

[14]LeCun,Y.etal.Handwrittendigitrecognitionwithaback-propagationnetwork.InProc.AdvancesinNeuralInformationProcessingSystems396–404(1990).

[15]AndrewNg.etal.UnsupervisedFeatureLearningandDeepLearningTutorialhttp://ufldl.stanford.edu/tutorial/.

[16]Glorot,X.,Bordes,A.&Bengio.Y.Deepsparserectifierneuralnetworks.InProc.14thInternationalConferenceonArtificialIntelligenceandStatistics315–323(2011).

[17]LeCun,Y.,Bottou,L.,Bengio,Y.&Haffner,P.Gradient-basedlearningappliedtodocumentrecognition.Proc.IEEE86,2278–2324(1998).

[18]Srivastava,N.,Hinton,G.,Krizhevsky,A.,Sutskever,I.&Salakhutdinov,R.Dropout:asimplewaytopreventneuralnetworksfromoverfitting.J.MachineLearningRes.15,1929–1958(2014).

[19]Dauphin,Y.etal.Identifyingandattackingthesaddlepointproblemin

high-dimensionalnon-convexoptimization.InProc.AdvancesinNeuralInformationProcessingSystems272933–2941(2014).

[20]Choromanska,A.,Henaff,M.,Mathieu,M.,Arous,G.B.&LeCun,Y.Thelosssurfaceofmultilayernetworks.InProc.ConferenceonAIandStatisticshttp://

arxiv.org/abs/1412.0233(2014).

[21]DavidNist′erandHenrikStew′enius.LinearTimeMaximallyStableExtremalRegions.In:Proc.ECCV.(2008).

[22]MichaelD.andHorstB.,EfficientMaximallyStableExtremalRegion(MSER)Tracking.In:Proc.CVPR.(2006).

[23]ChristianWolfandJean-MichelJolion.Objectcount/AreaGraphsfortheEvaluationofObjectDetectionandSegmentationAlgorithms,In:Proc.ICDAR.(2006).

55

[24]Jia,YangqingandShelhamer,Evan.etal.Caffe:ConvolutionalArchitectureforFastFeatureEmbedding.arXivpreprintarXiv:1408.5093

[25]AlessandroZamberletti,LuciaNoce,andIgnazioGallo.TextLocalizationbasedonFastFeaturePyramidsandMulti-resolutionMaximallyStableExtremalRegions.In:Proc.ICDAR.(2015).

[26]WeilinHuang,YuQiao,andXiaoouTang.RobustSceneTextDetectionwithConvolutionNeuralNetworkInducedMSERTrees.In:Proc.ECCV.(2014).

[27]M.Ranzato,Y.Boureau,andY.LeCun.“Sparsefeaturelearningfordeepbeliefnetworks,”inNeuralInformationProcessingSystems,2007.

[28]Marc'AurelioRanzato.“NEURALNETSFORVISIONCVPR2012TutorialonDeepLearningPartIII”.Google.

[29]AnnaZhu,GuoyouWang,YangboDong.Detectingnaturalscenestextviaautoimagepartition,two-stagegroupingandtwo-layerclassification.InPatternRecognitionLetters,2015.

[30]D.Karatzas,F.Shafait,S.Uchida,M.Iwamura,L.Gomez,S.Robles,J.Mas,D.

Fernandez,J.Almazan,L.P.delasHeras,"ICDAR2013RobustReadingCompetition",InProc.12thInternationalConferenceofDocumentAnalysisandRecognition,2013,IEEECPS,pp.1115-1124.

[31]JiaDeng,WeiDong,RichardSocher,Li-JiaLi,KaiLiandLiFei-Fei,"ImageNet:ALarge-ScaleHierarchicalImageDatabase",InProc.CVPR.(2009).

[32]马然.基于深度学习的自然场景文本识别系统的设计与实现[D].吉林大学.(2015).

[33]李嘉.基于卷积神经网络的自然场景下的车牌检测[D].郑州大学.(2012).

56