Learning switching linear models of human motion

Vladimir Pavlovic; James M. Rehg; John MacCormick

Learning switching linear models of human motion

Vladimir Pavlovic, James M. Rehg, John MacCormick

Research output: Chapter in Book/Report/Conference proceeding › Conference contribution

Abstract

The human figure exhibits complex and rich dynamic behavior that is both nonlinear and time-varying. Effective models of human dynamics can be learned from motion capture data using switching linear dynamic system (SLDS) models. We present results for human motion synthesis, classification, and visual tracking using learned SLDS models. Since exact inference in SLDS is intractable, we present three approximate inference algorithms and compare their performance. In particular, a new variational inference algorithm is obtained by casting the SLDS model as a Dynamic Bayesian Network. Classification experiments show the superiority of SLDS over conventional HMM's for our problem domain.

Original language	English (US)
Title of host publication	Advances in Neural Information Processing Systems 13 - Proceedings of the 2000 Conference, NIPS 2000
Publisher	Neural information processing systems foundation
ISBN (Print)	0262122413, 9780262122412
State	Published - 2001
Externally published	Yes
Event	14th Annual Neural Information Processing Systems Conference, NIPS 2000 - Denver, CO, United States Duration: Nov 27 2000 → Dec 2 2000

Publication series

Name	Advances in Neural Information Processing Systems
ISSN (Print)	1049-5258

Other

Other	14th Annual Neural Information Processing Systems Conference, NIPS 2000
Country/Territory	United States
City	Denver, CO
Period	11/27/00 → 12/2/00

All Science Journal Classification (ASJC) codes

Computer Networks and Communications
Information Systems
Signal Processing

Cite this

@inproceedings{6ec941cc5d5f4ffca3f7539613f67653,

title = "Learning switching linear models of human motion",

abstract = "The human figure exhibits complex and rich dynamic behavior that is both nonlinear and time-varying. Effective models of human dynamics can be learned from motion capture data using switching linear dynamic system (SLDS) models. We present results for human motion synthesis, classification, and visual tracking using learned SLDS models. Since exact inference in SLDS is intractable, we present three approximate inference algorithms and compare their performance. In particular, a new variational inference algorithm is obtained by casting the SLDS model as a Dynamic Bayesian Network. Classification experiments show the superiority of SLDS over conventional HMM's for our problem domain.",

author = "Vladimir Pavlovic and Rehg, {James M.} and John MacCormick",

year = "2001",

language = "English (US)",

isbn = "0262122413",

series = "Advances in Neural Information Processing Systems",

publisher = "Neural information processing systems foundation",

booktitle = "Advances in Neural Information Processing Systems 13 - Proceedings of the 2000 Conference, NIPS 2000",

note = "14th Annual Neural Information Processing Systems Conference, NIPS 2000 ; Conference date: 27-11-2000 Through 02-12-2000",

}

Pavlovic, V, Rehg, JM & MacCormick, J 2001, Learning switching linear models of human motion. in Advances in Neural Information Processing Systems 13 - Proceedings of the 2000 Conference, NIPS 2000. Advances in Neural Information Processing Systems, Neural information processing systems foundation, 14th Annual Neural Information Processing Systems Conference, NIPS 2000, Denver, CO, United States, 11/27/00.

Learning switching linear models of human motion. / Pavlovic, Vladimir; Rehg, James M.; MacCormick, John.

Advances in Neural Information Processing Systems 13 - Proceedings of the 2000 Conference, NIPS 2000. Neural information processing systems foundation, 2001. (Advances in Neural Information Processing Systems).

Research output: Chapter in Book/Report/Conference proceeding › Conference contribution

TY - GEN

T1 - Learning switching linear models of human motion

AU - Pavlovic, Vladimir

AU - Rehg, James M.

AU - MacCormick, John

PY - 2001

Y1 - 2001

N2 - The human figure exhibits complex and rich dynamic behavior that is both nonlinear and time-varying. Effective models of human dynamics can be learned from motion capture data using switching linear dynamic system (SLDS) models. We present results for human motion synthesis, classification, and visual tracking using learned SLDS models. Since exact inference in SLDS is intractable, we present three approximate inference algorithms and compare their performance. In particular, a new variational inference algorithm is obtained by casting the SLDS model as a Dynamic Bayesian Network. Classification experiments show the superiority of SLDS over conventional HMM's for our problem domain.

AB - The human figure exhibits complex and rich dynamic behavior that is both nonlinear and time-varying. Effective models of human dynamics can be learned from motion capture data using switching linear dynamic system (SLDS) models. We present results for human motion synthesis, classification, and visual tracking using learned SLDS models. Since exact inference in SLDS is intractable, we present three approximate inference algorithms and compare their performance. In particular, a new variational inference algorithm is obtained by casting the SLDS model as a Dynamic Bayesian Network. Classification experiments show the superiority of SLDS over conventional HMM's for our problem domain.

UR - http://www.scopus.com/inward/record.url?scp=84898993749&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=84898993749&partnerID=8YFLogxK

M3 - Conference contribution

AN - SCOPUS:84898993749

SN - 0262122413

SN - 9780262122412

T3 - Advances in Neural Information Processing Systems

BT - Advances in Neural Information Processing Systems 13 - Proceedings of the 2000 Conference, NIPS 2000

PB - Neural information processing systems foundation

T2 - 14th Annual Neural Information Processing Systems Conference, NIPS 2000

Y2 - 27 November 2000 through 2 December 2000

ER -

Learning switching linear models of human motion

Abstract

Publication series

Other

All Science Journal Classification (ASJC) codes

Other files and links

Fingerprint

Cite this