Abstract
This paper proposes a novel probabilistic graphical model, called MTHMM-P, for partitioning general nonrigid motion into piecewise elastic motion, so as to achieve nonrigid motion estimation without the need of any a priori shape model. To this end, two interrelated sub-problems have to be addressed: partitioning whole motion sequence into several coherent pieces and estimating elastic motion inside each one. The proposed MTHMM-P jointly formulates these two sub-problems. By means of a competition-cooperation partition mechanism and joint inference algorithm, the MTHMM-P can automatically determine the number of pieces as well as adjust the span of each piece by adapting to the input sequence, and at the same time achieve the estimation of piecewise elastic motion. Experiments on the motion of face and body show the capability of the MTHMM-P.
This research was supported in part by the National Natural Science Foundation of China under Grant No. 61003098.
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
References
Huang, T.: Modeling, analysis, and visualization of nonrigid object motion. In: ICPR. (1990)
Aggarwal, J., Cai, Q., Liao, W., Sabata, B.: Nonrigid motion analysis: articulated and elastic motion. CVIU 70(2), 142–156 (1998)
Duta, N., Jain, A., Dubuisson-Jolly, M.: Automatic construction of 2D shape models. PAMI 23(5), 433–446 (2001)
Chui, H., Rangarajan, A., Zhang, J., Leonard, C.: Unsupervised learning of an atlas from unlabeled point-sets. PAMI 26(2), 160–172 (2004)
Di, H., Tao, L., Xu, G.: A mixture of transformed hidden Markov models for elastic motion estimation. PAMI 31(10), 1817–1830 (2009)
Beauchemin, S., Barron, J.: The computation of optical flow. ACM Comput. Surv. 27(3), 433–467 (1995)
Sun, D., Roth, S., Black, M.J.: A quantitative analysis of current practices in optical flow estimation and the principles behind them. IJCV 106(2), 115–137 (2014)
Revaud, J., Weinzaepfel, P., Harchaoui, Z., Schmid, C.: EpicFlow: Edge-preserving interpolation of correspondences for optical flow. In: CVPR (2015)
Bao, L., Yang, Q, Jin, H.: Fast edge-preserving patchmatch for large displacement optical flow. In: CVPR (2014)
Torresani, L., Yang, D., Alexander, E., Bregler, C.: Tracking and modeling non-rigid objects with rank constraints. In: CVPR (2001)
Wang, H., Schmid, C.: Action recognition with improved trajectories. In: ICCV (2013)
Wang, H., Klaser, A., Schmid, C., Liu, C.L.: Dense trajectories and motion boundary descriptors for action recognition. IJCV 103(1), 60–79 (2013)
Sundaram, N., Brox, T., Keutzer, K.: Dense point trajectories by GPU-accelerated large displacement optical flow. In: Daniilidis, K., Maragos, P., Paragios, N. (eds.) ECCV 2010, Part I. LNCS, vol. 6311, pp. 438–451. Springer, Heidelberg (2010)
Rubinstein, M., Liu, C., Freeman, W.T.: Towards longer long-range motion trajectories. In: BMVC (2012)
Frey, B., Jojic, N.: A comparison of algorithms for inference and learning in probabilistic graphical models. PAMI 27(9), 1392–1416 (2005)
Rabiner, L.: A tutorial on hidden Markov models and selected applications in speech recognition. Proc. IEEE 77(2), 257–286 (1989)
Author information
Authors and Affiliations
Corresponding author
Editor information
Editors and Affiliations
Rights and permissions
Copyright information
© 2015 Springer International Publishing Switzerland
About this paper
Cite this paper
Di, H., Tao, L., Xu, G. (2015). Adaptive Piecewise Elastic Motion Estimation. In: Huang, DS., Jo, KH., Hussain, A. (eds) Intelligent Computing Theories and Methodologies. ICIC 2015. Lecture Notes in Computer Science(), vol 9226. Springer, Cham. https://doi.org/10.1007/978-3-319-22186-1_21
Download citation
DOI: https://doi.org/10.1007/978-3-319-22186-1_21
Published:
Publisher Name: Springer, Cham
Print ISBN: 978-3-319-22185-4
Online ISBN: 978-3-319-22186-1
eBook Packages: Computer ScienceComputer Science (R0)