版权说明:本文档由用户提供并上传,收益归属内容提供方,若内容存在侵权,请进行举报或认领
文档简介
SharingdetailofImageNetClassificationwithDeepCNNs林木得OutlineOverviewGoalDatasetModelMotivationArchitectureResultsPartIBasicProblemsActivationFunctionLossFunctionLearningMethodPartIIModelFeaturesReLUNonlinearityTrainingonMultipleGPUsLocalResponseNormalizationOverlappoolingReduceOverfittingDataAugmentationDropoutPartIIIMainphasesPreprocessInitializationStochasticgradientdescentTestReferencesOverviewGoalDatasetModelResultsGoalImageclassificationClassify
theImageNetLSVRC-2010contestimagesinto1000differentclasses.DataSetroughly1.2milliontrainingimages50,000validationimages150,000testingimagesModelMotivation利用自然图像性质
stationarity
of
statistics
locality
of
pixel
independencies模拟神经网络工作机理
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,
横向抑制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
温馨提示
- 1. 本站所有资源如无特殊说明,都需要本地电脑安装OFFICE2007和PDF阅读器。图纸软件为CAD,CAXA,PROE,UG,SolidWorks等.压缩文件请下载最新的WinRAR软件解压。
- 2. 本站的文档不包含任何第三方提供的附件图纸等,如果需要附件,请联系上传者。文件的所有权益归上传用户所有。
- 3. 本站RAR压缩包中若带图纸,网页内容里面会有图纸预览,若没有图纸预览就没有图纸。
- 4. 未经权益所有人同意不得将文件中的内容挪作商业或盈利用途。
- 5. 人人文库网仅提供信息存储空间,仅对用户上传内容的表现方式做保护处理,对用户上传分享的文档内容本身不做任何修改或编辑,并不能对任何下载内容负责。
- 6. 下载文件中如有侵权或不适当内容,请与我们联系,我们立即纠正。
- 7. 本站不保证下载资源的准确性、安全性和完整性, 同时也不承担用户因使用这些下载资源对自己和他人造成任何形式的伤害或损失。
最新文档
- 上海市数据中心租赁协议
- 2023火电厂烟气脱硝技术导则
- 《食品安全》课件-食品中常见的双糖
- X射线影像增强器项目可行性报告
- 渔业病害防治技术手册
- 糖尿病积极生活
- 住宅小区物业管理规定
- 商业地产开发比例趋势
- 物业管理优化:房地产客户情况说明
- 高血压患者心理护理:改善生活品质
- 牧场饲草料管理制度
- GB/T 34960.3-2017信息技术服务治理第3部分:绩效评价
- GB/T 15089-2001机动车辆及挂车分类
- 2022年北京市大兴区生态环境系统事业单位招聘笔试试题及答案
- 宁夏回族自治区2022年7月普通高中学业水平测试
- 2023年国投新疆罗布泊钾盐有限责任公司招聘笔试题库及答案解析
- 2022年中考物理真题选及详解-计算题
- 解剖学泌尿系统理论知识考核试题及答案
- 电梯维保服务质量年度考核表
- 炼钢厂环境保护培训课件
- 药品生产质量管理论文
评论
0/150
提交评论