000851272 001__ 851272
000851272 005__ 20210129234827.0
000851272 0247_ $$2doi$$a10.1109/IGARSS.2018.8518671
000851272 0247_ $$2Handle$$a2128/19910
000851272 037__ $$aFZJ-2018-04967
000851272 1001_ $$0P:(DE-HGF)0$$aLange, Julius$$b0
000851272 1112_ $$aIEEE International Geoscience and Remote Sensing Symposium (IGARSS)$$cValencia$$d2018-07-22 - 2018-07-27$$gIGARSS 2018$$wSpain
000851272 245__ $$aThe Influence of Sampling Methods on Pixel-Wise Hyperspectral Image Classification with 3D Convolutional Neural Networks
000851272 260__ $$c2018
000851272 29510 $$aIGARSS 2018 - 2018 IEEE International Geoscience and Remote Sensing Symposium
000851272 300__ $$a2087 - 2090
000851272 3367_ $$2ORCID$$aCONFERENCE_PAPER
000851272 3367_ $$033$$2EndNote$$aConference Paper
000851272 3367_ $$2BibTeX$$aINPROCEEDINGS
000851272 3367_ $$2DRIVER$$aconferenceObject
000851272 3367_ $$2DataCite$$aOutput Types/Conference Paper
000851272 3367_ $$0PUB:(DE-HGF)8$$2PUB:(DE-HGF)$$aContribution to a conference proceedings$$bcontrib$$mcontrib$$s1541429841_25759
000851272 3367_ $$0PUB:(DE-HGF)7$$2PUB:(DE-HGF)$$aContribution to a book$$mcontb
000851272 520__ $$aSupervised image classification is one of the essential techniques for generating semantic maps from remotely sensed images.The lack of labeled ground truth datasets, due to the inherent time effort and cost involved in collecting training samples, has led to the practice of training and validating new classifiers within a single image. In line with that, the dominant approach for the division of the available ground truth into disjoint training and test sets is random sampling. This paper discusses the problems that arise when this strategy is adopted in conjunction with spectral-spatial and pixel-wise classifiers such as 3D Convolutional Neural Networks (3D CNN). It is shown that a random sampling scheme leads to a violation of the independence assumption and to the illusion that global knowledge is extracted from the training set.To tackle this issue, two improved sampling strategies based on the Density-Based Clustering Algorithm (DBSCAN) are proposed. They minimize the violation of the train and test samples independence assumption and thus ensure an honest estimation of the generalization capabilities of the classifier.
000851272 536__ $$0G:(DE-HGF)POF3-512$$a512 - Data-Intensive Science and Federated Computing (POF3-512)$$cPOF3-512$$fPOF III$$x0
000851272 536__ $$0G:(EU-Grant)754304$$aDEEP-EST - DEEP - Extreme Scale Technologies (754304)$$c754304$$fH2020-FETHPC-2016$$x1
000851272 536__ $$0G:(EU-Grant)763558$$aSIMDAS - Upgrade of CaSToRC into a Center of Excellence in Simulation and Data Science (763558)$$c763558$$fH2020-WIDESPREAD-04-2017-TeamingPhase1$$x2
000851272 7001_ $$0P:(DE-Juel1)171343$$aCavallaro, Gabriele$$b1
000851272 7001_ $$0P:(DE-Juel1)162390$$aGötz, Markus$$b2
000851272 7001_ $$0P:(DE-HGF)0$$aErnir, Erlingsson$$b3
000851272 7001_ $$0P:(DE-Juel1)132239$$aRiedel, Morris$$b4
000851272 773__ $$a10.1109/IGARSS.2018.8518671
000851272 8564_ $$uhttps://juser.fz-juelich.de/record/851272/files/IGARSS_Lange.pdf$$yOpenAccess
000851272 8564_ $$uhttps://juser.fz-juelich.de/record/851272/files/IGARSS_Lange.gif?subformat=icon$$xicon$$yOpenAccess
000851272 8564_ $$uhttps://juser.fz-juelich.de/record/851272/files/IGARSS_Lange.jpg?subformat=icon-1440$$xicon-1440$$yOpenAccess
000851272 8564_ $$uhttps://juser.fz-juelich.de/record/851272/files/IGARSS_Lange.jpg?subformat=icon-180$$xicon-180$$yOpenAccess
000851272 8564_ $$uhttps://juser.fz-juelich.de/record/851272/files/IGARSS_Lange.jpg?subformat=icon-640$$xicon-640$$yOpenAccess
000851272 8564_ $$uhttps://juser.fz-juelich.de/record/851272/files/IGARSS_Lange.pdf?subformat=pdfa$$xpdfa$$yOpenAccess
000851272 909CO $$ooai:juser.fz-juelich.de:851272$$pdnbdelivery$$pec_fundedresources$$pVDB$$pdriver$$popen_access$$popenaire
000851272 9101_ $$0I:(DE-588b)5008462-8$$6P:(DE-Juel1)171343$$aForschungszentrum Jülich$$b1$$kFZJ
000851272 9101_ $$0I:(DE-588b)5008462-8$$6P:(DE-Juel1)162390$$aForschungszentrum Jülich$$b2$$kFZJ
000851272 9101_ $$0I:(DE-HGF)0$$6P:(DE-HGF)0$$aHáskóli Íslands$$b3
000851272 9101_ $$0I:(DE-588b)5008462-8$$6P:(DE-Juel1)132239$$aForschungszentrum Jülich$$b4$$kFZJ
000851272 9131_ $$0G:(DE-HGF)POF3-512$$1G:(DE-HGF)POF3-510$$2G:(DE-HGF)POF3-500$$3G:(DE-HGF)POF3$$4G:(DE-HGF)POF$$aDE-HGF$$bKey Technologies$$lSupercomputing & Big Data$$vData-Intensive Science and Federated Computing$$x0
000851272 9141_ $$y2018
000851272 915__ $$0StatID:(DE-HGF)0510$$2StatID$$aOpenAccess
000851272 9201_ $$0I:(DE-Juel1)JSC-20090406$$kJSC$$lJülich Supercomputing Center$$x0
000851272 980__ $$acontrib
000851272 980__ $$aVDB
000851272 980__ $$aUNRESTRICTED
000851272 980__ $$acontb
000851272 980__ $$aI:(DE-Juel1)JSC-20090406
000851272 9801_ $$aFullTexts