REPAINT:KnowledgeTransferinDeepReinforcementLearningYunzheTao1SahikaGenc1JonathanChung1TaoSun1SunilMallya1Abstractimproveperformanceonothertasks.AcceleratinglearningprocessesforcomplextasksTransfer...
Zero-ShotKnowledgeDistillationfromaDecision-BasedBlack-BoxModelZiWang1Abstractresource-limiteddevicessuchasmobilephonesanddrones(Moskalenkoetal.,2018).Recently,alargenumberofap-Knowledgedistillatio...
ReasoningOverVirtualKnowledgeBasesWithOpenPredicateRelationsHaitianSun1PatVerga2BhuwanDhingra2RuslanSalakhutdinov1WilliamW.Cohen2AbstractAnalternativetoextractinginformationtoaugmentexistingKBsreli...
KnowledgeEnhancedMachineLearningPipelineagainstDiverseAdversarialAttacksNeziheMerveGu¨rel1XiangyuQi2LukaRimanic1CeZhang1BoLi3Abstract1.IntroductionDespitethegreatsuccessesachievedbydeepneu-Recents...
Data-FreeKnowledgeDistillationforHeterogeneousFederatedLearningZhuangdiZhu1JunyuanHong1JiayuZhou1Abstractprivacy-preservinglearningscheme,FLhasshownitspo-tentialtofacilitatereal-worldapplications,i...
LearningCompoundTaskswithoutTask-specificKnowledgeviaImitationandSelf-supervisedLearningSang-HyunLee12Seung-WooSeo1AbstractCompoundTasksub-task3Mostreal-worldtasksarecompoundtasksthatContinuousLate...
Feature-map-levelOnlineAdversarialKnowledgeDistillationInseopChung1SeongUkPark1JanghoKim1NojunKwak1Abstractsuchasmobileorembeddedsystems.Toovercomethisissue,manyresearcheshavebeenconductedtodevelop...
TowardsUnderstandingKnowledgeDistillationMaryPhuong1ChristophH.Lampert1Abstractdistillation-basedtraininghasbeenconfirmedseveraltimes:theoptimizationstepisgenerallymorewell-behavedthanKnowledgedist...
PredicateExchange:InferencewithDeclarativeKnowledgeZennaTavares1JavierBurroni2EdgarMinasyan3ArmandoSolarLezama1RajeshRanganath4Abstracttributionoverthreedimensionalscenes.SomeformsofKnowledge,sucha...
Zero-ShotKnowledgeDistillationinDeepNetworksGauravKumarNayak1KondaReddyMopuri2VaisakhShaj3R.VenkateshBabu1AnirbanChakraborty1Abstractlentperformance,buttheycanbehugeandcomputationallyexpensive.Henc...
RepresentationLearningonGraphswithJumpingKnowledgeNetworksKeyuluXu1ChengtaoLi1YonglongTian1TomohiroSonobe2Ken-ichiKawarabayashi2StefanieJegelka1Abstract2017;Grover&Leskovec,2016;Tangetal.,2015).Rec...
KnowledgeTransferwithJacobianMatchingSurajSrinivas1Franc¸oisFleuret1Abstracttoeasilyexplorethespaceofneuralnetworkarchitectures,whichcanbeusedforneuralnetworkarchitecturesearch,Classicaldistillati...
CanonicalTensorDecompositionforKnowledgeBaseCompletionTimothéeLacroix12NicolasUsunier1GuillaumeObozinski2Abstractetal.,2014),aparticularcaseofCPwhichsharesthefac-torsforthesubjectandobjectmodes,wa...
Know-Evolve:DeepTemporalReasoningforDynamicKnowledgeGraphsRakshitTrivedi1HanjunDai1YichenWang1LeSong1AbstractFigure1.SampletemporalKnowledgesubgraphbetweenpersons,organizationsandcountries.Theavail...