版權(quán)說(shuō)明:本文檔由用戶提供并上傳,收益歸屬內(nèi)容提供方,若內(nèi)容存在侵權(quán),請(qǐng)進(jìn)行舉報(bào)或認(rèn)領(lǐng)
文檔簡(jiǎn)介
SharingdetailofImageNetClassificationwithDeepCNNs林木得OutlineOverviewGoalDatasetModelMotivationArchitectureResultsPartIBasicProblemsActivationFunctionLossFunctionLearningMethodPartIIModelFeaturesReLUNonlinearityTrainingonMultipleGPUsLocalResponseNormalizationOverlappoolingReduceOverfittingDataAugmentationDropoutPartIIIMainphasesPreprocessInitializationStochasticgradientdescentTestReferencesOverviewGoalDatasetModelResultsGoalImageclassificationClassify
theImageNetLSVRC-2010contestimagesinto1000differentclasses.DataSetroughly1.2milliontrainingimages50,000validationimages150,000testingimagesModelMotivation利用自然圖像性質(zhì)
stationarity
of
statistics
locality
of
pixel
independencies模擬神經(jīng)網(wǎng)絡(luò)工作機(jī)理
receptivefieldModelArchitectureResultsTesterrorinILSVR-2010testsetResultsTesterrorinILSVR-2012testsetsPartIBasicProblemsActivationFunctionLostFunctionLearningMethodActivationFunctionForalllayersexceptoutputlayer: RectifiedLinearUnit(ReLU)TobeconfirmedForoutputlayer:
ReLUandLossFunctionmultinomiallogisticregressionobjective:
tobeconfirmed
LearningMethodGradientDescentTobemorespecific,StochasticGradientDescentwithbatchof128images.PartIIModelFeaturesReLUNonlinearityTrainingonMultipleGPUsLocalResponseNormalizationOverlappoolingReduceOverfittingDataAugmentationDropoutReLUNonlinearityStandardactivationfunction:f(x)=tanh(x)orf(x)=(1+ex)-1
Newinthispaper:
RectifiedLinearUnit(ReLU):
f(x)=max(0,x)
CIFAR-10PerformancecompariseTrainingonMultipleGPUsputshalfofthekernels(orneurons)oneachGPUtheGPUscommunicateonlyincertainlayers.readfromandwritetooneanother’smemorydirectly,Withouthostmachinememoryreducesourtop-1andtop-5errorratesby1.7%and1.2%LocalResponseNormalizationOnvalidationset
k=2,n=5,alpha=10-4,andbeta=0.75
In
realneurons,
橫向抑制reducesourtop-1andtop-5errorratesby1.4%and1.2%,respectively.OverlappoolingTraditionally,
non-overlappoolingNewinthispaper:Overlappoolings=2andz=3.educesthetop-1andtop-5errorratesby0.4%and0.3%,respectivelyWhypooling:
1,reducenumberofneuron 2,translateinvarianceOverallarchitectureOverallArchitectureNeuronineachlayers:224x224x3,55x55x96,27x27x256,13x13x394,13x13x394,13x13x256,4096,4096,1000.Almost:650,000neuronsParameterineachlayers:11x11x3x96,5x5x48x256,3x3x256x384,3x3x192x384,3x3x192x256,43264x4096,4096x4096,4096x1000Almost:60millionparametersReduceOverfittingReduceoverfittingisthemostimportantproblemforthismodelDataArgumentationgeneratingimagetranslationsandhorizontalreflec-tions.Train:Afactorof2048moreimagesTest:5x2imagesaveragepredictalteringtheintensitiesoftheRGBchannelsintrainingimages.toeachRGBimagepixelIxy=[IR,IG,IB]Tweaddthefollowingquantity:xyxyxyreducesthetop-1errorratebyover1%.
ReduceOverfittingDropoutMotivation:
Tooexpensivetocombinemanyabovemodelsthattakes5daystotrain
ReduceOverfittingDropoutHOW:
train:settingtozerotheoutputofeachhiddenneuronwithprobability0.5inthefirst2fully-connectlayers.
test:usealltheneuronsbutmultiplytheiroutputsby0.5ReduceOverfittingDropoutCost:
roughlydoublesthenumberofiterationsrequiredtoconverge
PartIIIMainphasesPreprocessInitializationStochasticgradientdescentTestPreprocessdown-sampledtheimagestoafixedresolutionof256x256rescaledtheimagesuchthattheshortersidewasoflength256croppedoutthecentral256x256patchfromtheresultingimagesubtractingthemeanactivityoverthetrainingsetfromeachpixel.Thustrainnetworkonthe(centered)rawRGBvaluesofthepixels.Initializationinitializedtheweightsineachlayerfromazero-meanGaussiandistributionwithstandardde-viation0.01.initializedtheneuronbiasesinthesecond,fourth,andfifth
convolutionallayers,aswellasinthefully-connectedhiddenlayers,withtheconstant1
initializedtheneuronbiasesintheremaininglayerswiththeconstant0learningratewasinitializedat0.01Stochasticgradientdescentwithabatchsizeof128examplesdecayof0.0005Updaterulesdividethelearningrateby10whenthevalidationerrorratestoppedimprovingwiththecurrentlearningrate.learningratereducedthreetimespriortotermination90cyclesthrough1.2millionimages
,took5to6daysTestAttesttime,thenetworkmakesapredictionbyextracting5x2224x224patchesaswellastheirhorizontalreflections(hencetenpatchesinall),andaveragingthepredictionsmadebythenetwork’ssoftmaxlayeronthetenpatches.Attesttime,weusealltheneuronsbutmultiplytheiroutputsby0.5
inthefirsttwofully-connectedlayers.References1,ImageNetClassifi
溫馨提示
- 1. 本站所有資源如無(wú)特殊說(shuō)明,都需要本地電腦安裝OFFICE2007和PDF閱讀器。圖紙軟件為CAD,CAXA,PROE,UG,SolidWorks等.壓縮文件請(qǐng)下載最新的WinRAR軟件解壓。
- 2. 本站的文檔不包含任何第三方提供的附件圖紙等,如果需要附件,請(qǐng)聯(lián)系上傳者。文件的所有權(quán)益歸上傳用戶所有。
- 3. 本站RAR壓縮包中若帶圖紙,網(wǎng)頁(yè)內(nèi)容里面會(huì)有圖紙預(yù)覽,若沒(méi)有圖紙預(yù)覽就沒(méi)有圖紙。
- 4. 未經(jīng)權(quán)益所有人同意不得將文件中的內(nèi)容挪作商業(yè)或盈利用途。
- 5. 人人文庫(kù)網(wǎng)僅提供信息存儲(chǔ)空間,僅對(duì)用戶上傳內(nèi)容的表現(xiàn)方式做保護(hù)處理,對(duì)用戶上傳分享的文檔內(nèi)容本身不做任何修改或編輯,并不能對(duì)任何下載內(nèi)容負(fù)責(zé)。
- 6. 下載文件中如有侵權(quán)或不適當(dāng)內(nèi)容,請(qǐng)與我們聯(lián)系,我們立即糾正。
- 7. 本站不保證下載資源的準(zhǔn)確性、安全性和完整性, 同時(shí)也不承擔(dān)用戶因使用這些下載資源對(duì)自己和他人造成任何形式的傷害或損失。
最新文檔
- 2024年酒店VIP會(huì)員卡充值服務(wù)具體合同樣本版B版
- 二零二五年度新型城市配送兼職司機(jī)服務(wù)協(xié)議2篇
- 光伏發(fā)電項(xiàng)目合同
- 工業(yè)機(jī)器人聯(lián)合開(kāi)發(fā)合作協(xié)議
- IT系統(tǒng)安全維護(hù)指南
- 2025年度電子商務(wù)平臺(tái)內(nèi)容審核與服務(wù)合同2篇
- 2025年度煤炭裝卸作業(yè)質(zhì)量保證合同3篇
- 智能家居設(shè)備制造合同
- 2025年度生態(tài)園林施工分包合同模板(景觀設(shè)計(jì))3篇
- 2024年采購(gòu)精英專(zhuān)屬福利協(xié)議樣本版
- 零碳智慧園區(qū)解決方案
- 2025年林權(quán)抵押合同范本
- 2024年北師大版四年級(jí)數(shù)學(xué)上學(xué)期學(xué)業(yè)水平測(cè)試 期末卷(含答案)
- 2024年高考物理一輪復(fù)習(xí)講義(新人教版):第七章動(dòng)量守恒定律
- 浙江省寧波市慈溪市2023-2024學(xué)年高三上學(xué)期語(yǔ)文期末測(cè)試試卷
- 草學(xué)類(lèi)專(zhuān)業(yè)生涯發(fā)展展示
- 法理學(xué)課件馬工程
- 《玉米種植技術(shù)》課件
- 第47屆世界技能大賽江蘇省選拔賽計(jì)算機(jī)軟件測(cè)試項(xiàng)目技術(shù)工作文件
- 2023年湖北省公務(wù)員錄用考試《行測(cè)》答案解析
- M200a電路分析(電源、藍(lán)牙、FM)
評(píng)論
0/150
提交評(píng)論