資源描述：
《Deformable Part Models are Convolutional Neural Networks》由會(huì)員上傳分享，免費(fèi)在線閱讀，更多相關(guān)內(nèi)容在學(xué)術(shù)論文-天天文庫(kù)。
1、DeformablePartModelsareConvolutionalNeuralNetworksRossGirshick1ForrestIandola2TrevorDarrell2JitendraMalik21MicrosoftResearch2UCBerkeleyrbg@microsoft.comfforresti,trevor,malikg@eecs.berkeley.eduAbstractCNN.Inotherwords,deformablepartmodelsareconvo-lutionalneuralnetworks.Ourconstructionrelieso
2、nanewDeformablepartmodels(DPMs)andconvolutionalneu-networklayer,distancetransformpooling,whichgeneral-ralnetworks(CNNs)aretwowidelyusedtoolsforvi-izesmaxpooling.sualrecognition.Theyaretypicallyviewedasdistinctap-DPMstypicallyoperateonascale-spacepyramidofgra-proaches:DPMsaregraphicalmodels(M
3、arkovrandomdientorientationfeaturemaps(HOG[5]).Butwenow?elds),whileCNNsare“black-box”non-linearclassi?ers.knowthatforobjectdetectionthisfeaturerepresentationisInthispaper,weshowthataDPMcanbeformulatedasasuboptimalcomparedtofeaturescomputedbydeepcon-CNN,thusprovidingasynthesisofthetwoideas.Ou
4、rcon-volutionalnetworks[17].Asasecondinnovation,were-structioninvolvesunrollingtheDPMinferencealgorithmplaceHOGwithfeatureslearnedbyafully-convolutionalandmappingeachsteptoanequivalentCNNlayer.Fromnetwork.This“front-end”networkgeneratesapyramidofthisperspective,itisnaturaltoreplacethestandar
5、dim-deepfeatures,analogoustoaHOGfeaturepyramid.WeagefeaturesusedinDPMswithalearnedfeatureextractor.callthefullmodelaDeepPyramidDPM.WecalltheresultingmodelaDeepPyramidDPMandex-WeexperimentallyvalidateDeepPyramidDPMsbyperimentallyvalidateitonPASCALVOCobjectdetection.measuringobjectdetectionper
6、formanceonPASCALVOCWe?ndthatDeepPyramidDPMssigni?cantlyoutperform[9].SincetraditionalDPMshavebeentunedforHOGfea-DPMsbasedonhistogramsoforientedgradientsfeaturesturesovermanyyears,we?rstanalyzethedifferencesbe-(HOG)andslightlyoutperformsacomparableversionoftweenHOGfeaturepyramidsanddeepfeatur
7、epyramids.therecentlyintroducedR-CNNdetectionsystem,whilerun-WethenselectagoodmodelstructureandtrainaDeep-ningsigni?cantlyfaster.PyramidDPMthatsigni?cantlyoutperformsthebestHOG-basedDPMs.Whilewedon’texpectourapproachtoout-performa?ne-tunedR-CNNdete
