Structure from Articulated Motion: Accurate and Stable Monocular 3D Reconstruction without Training Data.
articulated structure recovery
human pose estimation
structure from motion
Journal
Sensors (Basel, Switzerland)
ISSN: 1424-8220
Titre abrégé: Sensors (Basel)
Pays: Switzerland
ID NLM: 101204366
Informations de publication
Date de publication:
22 Oct 2019
22 Oct 2019
Historique:
received:
23
09
2019
revised:
15
10
2019
accepted:
15
10
2019
entrez:
27
10
2019
pubmed:
28
10
2019
medline:
28
10
2019
Statut:
epublish
Résumé
Recovery of articulated 3D structure from 2D observations is a challenging computer vision problem with many applications. Current learning-based approaches achieve state-of-the-art accuracy on public benchmarks but are restricted to specific types of objects and motions covered by the training datasets. Model-based approaches do not rely on training data but show lower accuracy on these datasets. In this paper, we introduce a model-based method called
Identifiants
pubmed: 31652665
pii: s19204603
doi: 10.3390/s19204603
pmc: PMC6833108
pii:
doi:
Types de publication
Journal Article
Langues
eng
Sous-ensembles de citation
IM
Subventions
Organisme : Bundesministerium für Bildung und Forschung
ID : 01IW18002
Références
IEEE Trans Pattern Anal Mach Intell. 2008 May;30(5):865-77
pubmed: 18369255
IEEE Trans Pattern Anal Mach Intell. 2011 Jul;33(7):1442-56
pubmed: 21079275
IEEE Trans Pattern Anal Mach Intell. 2016 Aug;38(8):1505-16
pubmed: 27093439
IEEE Trans Pattern Anal Mach Intell. 2019 Jan 14;:null
pubmed: 30640602
IEEE Trans Pattern Anal Mach Intell. 2014 Jul;36(7):1325-39
pubmed: 26353306
Sensors (Basel). 2019 Aug 31;19(17):null
pubmed: 31480461
IEEE Trans Pattern Anal Mach Intell. 2019 Apr;41(4):901-914
pubmed: 29993801
Proc IEEE Int Conf Comput Vis. 2011;:802-809
pubmed: 24002226