Abstract
In this paper an approach is described to estimate 3D pose using a part based stochastic method. A proposed representation of the human body is explored defined over joints that employs full conditional models learnt between connected joints. This representation is compared against a popular alternative defined over parts using approximated limb conditionals. It is shown that using full limb conditionals results in a model that is far more representative of the original training data. Furthermore, it is demonstrated that Expectation Maximization is suitable for estimating 3D pose and better convergence is achieved when using full limb conditionals. To demonstrate the efficacy of the proposed method it is applied to the domain of 3D pose estimation using a single monocular image. Quantitative results are provided using the HumanEva dataset which confirm that the proposed method outperforms that of the competing part based model. In this work just a single model is learnt to represent all actions contained in the dataset which is applied to all subjects viewed from differing angles.
This is a preview of subscription content, log in via an institution to check access.
Access this chapter
Tax calculation will be finalised at checkout
Purchases are for personal use only
Preview
Unable to display preview. Download preview PDF.
References
Felzenswalb, P.F., Huttenlocher, D.P.: Pictorial Structures for Object Recogntion. International Journal on Computer Vision 61, 55–79 (2005)
Sigal, L., Bhatia, S., Roth, S., Black, M.J., Isard, M.: Tracking Loose-Limbed People. In: CVPR, pp. 421–428 (2005)
Andriluka, M., Roth, S., Schiele, B.: Pictorial Structures Revisited: People Detection and Articualted Pose Estimation. In: CVPR, pp. 1–8 (2009)
Ramanan, D.: Learning to Parse Images of Articulated Bodies. In: NIPS, pp. 1129–1136 (2006)
Sigal, L., Balan, A., Black, M.: HumanEva: Synchronized Video and Motion Capture Dataset and Baseline Algorithm for Evaluation of Articulated Human Motion. International Journal of Computer Vision 87, 4–27 (2009)
Isard, M.: PAMPAS: Real-Valued Graphical Models for Computer Vision. In: CVPR, pp. 613–620 (2003)
Hua, G., Wu, Y.: Variational Maximum a Posteriori by Annealed Mean Field Analysis. IEEE Transaction on Pattern Analysis and Machine Intelligience 27, 1747–1761 (2005)
Deutscher, J., Davidson, A., Reid, I.: Automatic Partitioning of High Dimensional Search Space associated with Articulated Body Motion Capture. In: CVPR, pp. 669–676 (2001)
Deutscher, J., Blake, A., Reid, I.: Articualted Body Motion Capture by Annealed Particle Filtering. In: CVPR, pp. 126–133 (2000)
Bishop, C.M.: Pattern Recognition and Machine Learning. Springer, Heidelberg (2006)
Dalal, N., Triggs, B.: Histogram of Orientated Gradients for Human Detection. In: CVPR, pp. 886–893 (2005)
Ferrari, V., Marin-Jimenez, M., Zisserman, A.: Progressive Search Space Reduction for Human Pose Estimation. In: CVPR, pp. 1–8 (2008)
Gao, J., Shi, J.: Multiple Frame Motion Inference using Belief Propagation. In: IEEE Conference on Automatic Face and Gesture Recognition, pp. 875–880 (2004)
Bernier, O., Cheung-Mon-Chan, P.: Real-Time 3D Articualted Pose Tracking using Particle filter and Belief Propagation on Factor Graphs. In: BMVC, pp. 27–36 (2006)
Author information
Authors and Affiliations
Editor information
Editors and Affiliations
Rights and permissions
Copyright information
© 2010 Springer-Verlag Berlin Heidelberg
About this paper
Cite this paper
Daubney, B., Xie, X. (2010). Estimating 3D Pose via Stochastic Search and Expectation Maximization. In: Perales, F.J., Fisher, R.B. (eds) Articulated Motion and Deformable Objects. AMDO 2010. Lecture Notes in Computer Science, vol 6169. Springer, Berlin, Heidelberg. https://doi.org/10.1007/978-3-642-14061-7_7
Download citation
DOI: https://doi.org/10.1007/978-3-642-14061-7_7
Publisher Name: Springer, Berlin, Heidelberg
Print ISBN: 978-3-642-14060-0
Online ISBN: 978-3-642-14061-7
eBook Packages: Computer ScienceComputer Science (R0)