Classifying bio-inspired model of point-light human motion using Echo State networks
DC Element | Wert | Sprache |
---|---|---|
dc.contributor.author | Tanisaro, P. | |
dc.contributor.author | Lehman, C. | |
dc.contributor.author | Sütfeld, L. | |
dc.contributor.author | Pipa, G. | |
dc.contributor.author | Heidemann, G. | |
dc.contributor.editor | Verschure, P.F. | |
dc.contributor.editor | Lintas, A. | |
dc.contributor.editor | Villa, A.E. | |
dc.contributor.editor | Rovetta, S. | |
dc.date.accessioned | 2021-12-23T16:34:18Z | - |
dc.date.available | 2021-12-23T16:34:18Z | - |
dc.date.issued | 2017 | |
dc.identifier.isbn | 9783319685991 | |
dc.identifier.issn | 03029743 | |
dc.identifier.uri | https://osnascholar.ub.uni-osnabrueck.de/handle/unios/18089 | - |
dc.description | Conference of 26th International Conference on Artificial Neural Networks, ICANN 2017 ; Conference Date: 11 September 2017 Through 14 September 2017; Conference Code:202259 | |
dc.description.abstract | We introduce a feature extraction scheme from a biologically inspired model using receptive fields (RFs) to point-light human motion patterns to form an action descriptor. The Echo State Network (ESN) which also has a biological plausibility is chosen for classification. We demonstrate the efficiency and robustness of applying the proposed feature extraction technique with ESN by constraining the test data based on arbitrary untrained viewpoints, in combination with unseen subjects under the following conditions: (i) lower sub-sampling frame rates to simulate data sequence loss, (ii) remove key points to simulate occlusion, and (iii) include untrained movements such as drunkard's walk. © Springer International Publishing AG 2017. | |
dc.description.sponsorship | ENNS | |
dc.language.iso | en | |
dc.publisher | Springer Verlag | |
dc.relation.ispartof | Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics) | |
dc.subject | Bio-inspired model | |
dc.subject | Biological motion | |
dc.subject | Biological motion perception | |
dc.subject | Echo state network | |
dc.subject | Echo state networks | |
dc.subject | Extraction | |
dc.subject | Feature extraction | |
dc.subject | Learning systems | |
dc.subject | Motion capture | |
dc.subject | Motion recognition | |
dc.subject | Motion recognition, Motion estimation | |
dc.subject | Neural networks, Bio-inspired Models | |
dc.title | Classifying bio-inspired model of point-light human motion using Echo State networks | |
dc.type | conference paper | |
dc.identifier.doi | 10.1007/978-3-319-68600-4_11 | |
dc.identifier.scopus | 2-s2.0-85034236748 | |
dc.identifier.url | https://www.scopus.com/inward/record.uri?eid=2-s2.0-85034236748&doi=10.1007%2f978-3-319-68600-4_11&partnerID=40&md5=67257dbba10550c99da9be6fa9892bfc | |
dc.description.volume | 10613 LNCS | |
dc.description.startpage | 84 | |
dc.description.endpage | 91 | |
dcterms.isPartOf.abbreviation | Lect. Notes Comput. Sci. | |
crisitem.author.dept | Institut für Kognitionswissenschaft | - |
crisitem.author.dept | Institut für Kognitionswissenschaft | - |
crisitem.author.deptid | institute28 | - |
crisitem.author.deptid | institute28 | - |
crisitem.author.orcid | 0000-0002-3416-2652 | - |
crisitem.author.parentorg | FB 08 - Humanwissenschaften | - |
crisitem.author.parentorg | FB 08 - Humanwissenschaften | - |
crisitem.author.grandparentorg | Universität Osnabrück | - |
crisitem.author.grandparentorg | Universität Osnabrück | - |
crisitem.author.netid | PiGo340 | - |
crisitem.author.netid | HeGu645 | - |
Seitenaufrufe
5
Letzte Woche
0
0
Letzter Monat
1
1
geprüft am 18.05.2024