Deterministic 3D human pose estimation using rigid structure

Valmadre, Jack and Lucey, Simon (2010). Deterministic 3D human pose estimation using rigid structure. In: Kostas Daniilidis, Petros Maragos and Nikos Paragios, Computer Vision: ECCV 2010. 11th European Conference on Computer Vision: Proceedings. ECCV 2010: 11th European Conference on Computer Vision, Hersonissos, Heraklion, Crete, Greece, (467-480). 5-11 September, 2010. doi:10.1007/978-3-642-15558-1_34


Author Valmadre, Jack
Lucey, Simon
Title of paper Deterministic 3D human pose estimation using rigid structure
Conference name ECCV 2010: 11th European Conference on Computer Vision
Conference location Hersonissos, Heraklion, Crete, Greece
Conference dates 5-11 September, 2010
Proceedings title Computer Vision: ECCV 2010. 11th European Conference on Computer Vision: Proceedings   Check publisher's open access policy
Journal name Lecture Notes in Computer Science   Check publisher's open access policy
Place of Publication Heidelberg, Germany
Publisher Springer
Publication Year 2010
Sub-type Fully published paper
DOI 10.1007/978-3-642-15558-1_34
ISBN 9783642155574
9783642155581
ISSN 0302-9743
1611-3349
Editor Kostas Daniilidis
Petros Maragos
Nikos Paragios
Volume 6313
Issue III
Start page 467
End page 480
Total pages 14
Language eng
Formatted Abstract/Summary
This paper explores a method, first proposed by Wei and Chai, for estimating 3D human pose from several frames of uncalibrated 2D point correspondences containing projected body joint locations. In their work Wei and Chai boldly claimed that, through the introduction of rigid constraints to the torso and hip, camera scales, bone lengths and absolute depths could be estimated from a finite number of frames (i.e. ≥ 5). In this paper we show this claim to be false, demonstrating in principle one can never estimate these parameters in a finite number of frames. Further, we demonstrate their approach is only valid for rigid sub-structures of the body (e.g. torso). Based on this analysis we propose a novel approach using deterministic structure from motion based on assumptions of rigidity in the body's torso. Our approach provides notably more accurate estimates and is substantially faster than Wei and Chai's approach, and unlike the original, can be solved as a deterministic least-squares problem.
Subjects 1700 Computer Science
2614 Theoretical Computer Science
Keyword Human pose estimation
Structure from motion
Q-Index Code E1
Q-Index Status Provisional Code
Institutional Status UQ

 
Versions
Version Filter Type
Citation counts: Scopus Citation Count Cited 10 times in Scopus Article | Citations
Google Scholar Search Google Scholar
Created: Thu, 28 Nov 2013, 01:07:09 EST by System User on behalf of School of Information Technol and Elec Engineering