000271108 001__ 271108
000271108 005__ 20240923102401.0
000271108 0247_ $$2doi$$a10.1007/s00348-024-03768-2
000271108 0247_ $$2ISSN$$a0723-4864
000271108 0247_ $$2ISSN$$a1432-1114
000271108 037__ $$aDZNE-2024-00980
000271108 082__ $$a530
000271108 1001_ $$aLagemann, Christian$$b0
000271108 245__ $$aChallenges of deep unsupervised optical flow estimation for particle-image velocimetry data
000271108 260__ $$aHeidelberg [u.a.]$$bSpringer$$c2024
000271108 3367_ $$2DRIVER$$aarticle
000271108 3367_ $$2DataCite$$aOutput Types/Journal article
000271108 3367_ $$0PUB:(DE-HGF)16$$2PUB:(DE-HGF)$$aJournal Article$$bjournal$$mjournal$$s1727079782_25599
000271108 3367_ $$2BibTeX$$aARTICLE
000271108 3367_ $$2ORCID$$aJOURNAL_ARTICLE
000271108 3367_ $$00$$2EndNote$$aJournal Article
000271108 520__ $$aIn recent years, several algorithms have been proposed that leverage deep learning techniques within the analysis workflow of particle-image velocimetry (PIV) measurements. This emerging body of work has shown that deep learning has the potential to match or outperform state-of-the-art classical algorithms in terms of efficiency, accuracy, and spatial resolution. However, the huge diversity in dynamic flows and varying particle-image conditions require PIV processing schemes to have high generalization capabilities to unseen flow and lighting conditions. If these conditions vary strongly compared to the training data, the performance of fully supervised PIV tools can degrade substantially. In contrast, unsupervised learning ameliorates the need for comprehensive labeled training data and can permit a much wider range of data to be used during training. Therefore, unsupervised deep learning could improve inference capability for challenging real-world use cases. However, design of an unsupervised loss objective is non-trivial and requires application-specific consideration. Motivated by the foregoing, in this paper we study unsupervised deep learning for PIV processing, systematically investigating key components of losses and accommodating regularizers and deriving a proxy loss. The resulting algorithm, named Unsupervised Recurrent All-Pairs Field Transforms for PIV (URAFT-PIV), is unsupervised and meant specifically for PIV applications. We investigate performance under varying image and lighting conditions in synthetic and experimental data, with a breadth and depth going well beyond currently available empirical results. These results shed new light on deep learning for PIV processing and in particular on the scope for unsupervised learning in this domain.
000271108 536__ $$0G:(DE-HGF)POF4-354$$a354 - Disease Prevention and Healthy Aging (POF4-354)$$cPOF4-354$$fPOF IV$$x0
000271108 588__ $$aDataset connected to CrossRef, Journals: pub.dzne.de
000271108 7001_ $$0P:(DE-2719)9001044$$aLagemann, Kai$$b1$$udzne
000271108 7001_ $$0P:(DE-2719)2811372$$aMukherjee, Sach$$b2$$udzne
000271108 7001_ $$aSchröder, Wolfgang$$b3
000271108 773__ $$0PERI:(DE-600)1476361-8$$a10.1007/s00348-024-03768-2$$gVol. 65, no. 3, p. 30$$n3$$p30$$tExperiments in fluids$$v65$$x0723-4864$$y2024
000271108 8564_ $$uhttps://pub.dzne.de/record/271108/files/DZNE-2024-00980%20SUP.pdf
000271108 8564_ $$uhttps://pub.dzne.de/record/271108/files/DZNE-2024-00980_Restricted.pdf
000271108 8564_ $$uhttps://pub.dzne.de/record/271108/files/DZNE-2024-00980%20SUP.pdf?subformat=pdfa$$xpdfa
000271108 8564_ $$uhttps://pub.dzne.de/record/271108/files/DZNE-2024-00980_Restricted.pdf?subformat=pdfa$$xpdfa
000271108 909CO $$ooai:pub.dzne.de:271108$$pVDB
000271108 9101_ $$0I:(DE-588)1065079516$$6P:(DE-2719)9001044$$aDeutsches Zentrum für Neurodegenerative Erkrankungen$$b1$$kDZNE
000271108 9101_ $$0I:(DE-588)1065079516$$6P:(DE-2719)2811372$$aDeutsches Zentrum für Neurodegenerative Erkrankungen$$b2$$kDZNE
000271108 9131_ $$0G:(DE-HGF)POF4-354$$1G:(DE-HGF)POF4-350$$2G:(DE-HGF)POF4-300$$3G:(DE-HGF)POF4$$4G:(DE-HGF)POF$$aDE-HGF$$bGesundheit$$lNeurodegenerative Diseases$$vDisease Prevention and Healthy Aging$$x0
000271108 9141_ $$y2024
000271108 915__ $$0StatID:(DE-HGF)3002$$2StatID$$aDEAL Springer$$d2023-10-21$$wger
000271108 915__ $$0StatID:(DE-HGF)3002$$2StatID$$aDEAL Springer$$d2023-10-21$$wger
000271108 915__ $$0StatID:(DE-HGF)0100$$2StatID$$aJCR$$bEXP FLUIDS : 2022$$d2023-10-21
000271108 915__ $$0StatID:(DE-HGF)0200$$2StatID$$aDBCoverage$$bSCOPUS$$d2023-10-21
000271108 915__ $$0StatID:(DE-HGF)0300$$2StatID$$aDBCoverage$$bMedline$$d2023-10-21
000271108 915__ $$0StatID:(DE-HGF)0600$$2StatID$$aDBCoverage$$bEbsco Academic Search$$d2023-10-21
000271108 915__ $$0StatID:(DE-HGF)0030$$2StatID$$aPeer Review$$bASC$$d2023-10-21
000271108 915__ $$0StatID:(DE-HGF)0199$$2StatID$$aDBCoverage$$bClarivate Analytics Master Journal List$$d2023-10-21
000271108 915__ $$0StatID:(DE-HGF)0113$$2StatID$$aWoS$$bScience Citation Index Expanded$$d2023-10-21
000271108 915__ $$0StatID:(DE-HGF)0150$$2StatID$$aDBCoverage$$bWeb of Science Core Collection$$d2023-10-21
000271108 915__ $$0StatID:(DE-HGF)0160$$2StatID$$aDBCoverage$$bEssential Science Indicators$$d2023-10-21
000271108 915__ $$0StatID:(DE-HGF)1160$$2StatID$$aDBCoverage$$bCurrent Contents - Engineering, Computing and Technology$$d2023-10-21
000271108 915__ $$0StatID:(DE-HGF)9900$$2StatID$$aIF < 5$$d2023-10-21
000271108 9201_ $$0I:(DE-2719)1013030$$kAG Mukherjee$$lStatistics and Machine Learning$$x0
000271108 980__ $$ajournal
000271108 980__ $$aVDB
000271108 980__ $$aI:(DE-2719)1013030
000271108 980__ $$aUNRESTRICTED