Learning motion correlation for tracking articulated human body with a Rao-Blackwellised particle filter

Xinyu Xu, Baoxin Li

Research output: Chapter in Book/Report/Conference proceedingConference contribution

39 Citations (Scopus)

Abstract

Inference in 3D articulated human body tracking is challenging due to the high dimensionality and nonlinearity of the parameter-space. We propose a particle filter with Rao-Blackwellisation which marginalizes part of the state variables by exploiting the correlation between the right-side and the left-side joint Euler angles. The correlation is naturally induced by the symmetric and repetitive patterns in specific human activities. A novel algorithm is proposed to learn the correlation from the training data using Partial Least Square regression. The learned correlation is then used as motion prior in designing the Rao-Blackwellised particle filter, which estimates only one group of state variables using the Monte Carlo method, leaving the other group being exactly computed through an analytical filter that utilizes the learned motion correlation. We evaluate the effectiveness of the motion correlation for 3D articulated human body tracking. The accuracy of the proposed 3D tracker is quantitatively assessed based on the distance between the true and the estimated marker positions. Extensive experiments with multi-camera walking sequences from the HumanEva-I/II data set show that (i) the proposed tracker achieves significantly lower estimation error than both the Annealed Particle Filter and the standard Particle Filter; and (ii) the learned motion correlation generalizes well to motion performed by subjects other than the training subject.

Original languageEnglish (US)
Title of host publicationProceedings of the IEEE International Conference on Computer Vision
DOIs
StatePublished - 2007
Event2007 IEEE 11th International Conference on Computer Vision, ICCV - Rio de Janeiro, Brazil
Duration: Oct 14 2007Oct 21 2007

Other

Other2007 IEEE 11th International Conference on Computer Vision, ICCV
CountryBrazil
CityRio de Janeiro
Period10/14/0710/21/07

Fingerprint

Error analysis
Monte Carlo methods
Cameras
Experiments

ASJC Scopus subject areas

  • Software
  • Computer Vision and Pattern Recognition

Cite this

Learning motion correlation for tracking articulated human body with a Rao-Blackwellised particle filter. / Xu, Xinyu; Li, Baoxin.

Proceedings of the IEEE International Conference on Computer Vision. 2007. 4408951.

Research output: Chapter in Book/Report/Conference proceedingConference contribution

Xu, X & Li, B 2007, Learning motion correlation for tracking articulated human body with a Rao-Blackwellised particle filter. in Proceedings of the IEEE International Conference on Computer Vision., 4408951, 2007 IEEE 11th International Conference on Computer Vision, ICCV, Rio de Janeiro, Brazil, 10/14/07. https://doi.org/10.1109/ICCV.2007.4408951
Xu, Xinyu ; Li, Baoxin. / Learning motion correlation for tracking articulated human body with a Rao-Blackwellised particle filter. Proceedings of the IEEE International Conference on Computer Vision. 2007.
@inproceedings{342283e7a7fa48edb1fabb9ea0a68527,
title = "Learning motion correlation for tracking articulated human body with a Rao-Blackwellised particle filter",
abstract = "Inference in 3D articulated human body tracking is challenging due to the high dimensionality and nonlinearity of the parameter-space. We propose a particle filter with Rao-Blackwellisation which marginalizes part of the state variables by exploiting the correlation between the right-side and the left-side joint Euler angles. The correlation is naturally induced by the symmetric and repetitive patterns in specific human activities. A novel algorithm is proposed to learn the correlation from the training data using Partial Least Square regression. The learned correlation is then used as motion prior in designing the Rao-Blackwellised particle filter, which estimates only one group of state variables using the Monte Carlo method, leaving the other group being exactly computed through an analytical filter that utilizes the learned motion correlation. We evaluate the effectiveness of the motion correlation for 3D articulated human body tracking. The accuracy of the proposed 3D tracker is quantitatively assessed based on the distance between the true and the estimated marker positions. Extensive experiments with multi-camera walking sequences from the HumanEva-I/II data set show that (i) the proposed tracker achieves significantly lower estimation error than both the Annealed Particle Filter and the standard Particle Filter; and (ii) the learned motion correlation generalizes well to motion performed by subjects other than the training subject.",
author = "Xinyu Xu and Baoxin Li",
year = "2007",
doi = "10.1109/ICCV.2007.4408951",
language = "English (US)",
booktitle = "Proceedings of the IEEE International Conference on Computer Vision",

}

TY - GEN

T1 - Learning motion correlation for tracking articulated human body with a Rao-Blackwellised particle filter

AU - Xu, Xinyu

AU - Li, Baoxin

PY - 2007

Y1 - 2007

N2 - Inference in 3D articulated human body tracking is challenging due to the high dimensionality and nonlinearity of the parameter-space. We propose a particle filter with Rao-Blackwellisation which marginalizes part of the state variables by exploiting the correlation between the right-side and the left-side joint Euler angles. The correlation is naturally induced by the symmetric and repetitive patterns in specific human activities. A novel algorithm is proposed to learn the correlation from the training data using Partial Least Square regression. The learned correlation is then used as motion prior in designing the Rao-Blackwellised particle filter, which estimates only one group of state variables using the Monte Carlo method, leaving the other group being exactly computed through an analytical filter that utilizes the learned motion correlation. We evaluate the effectiveness of the motion correlation for 3D articulated human body tracking. The accuracy of the proposed 3D tracker is quantitatively assessed based on the distance between the true and the estimated marker positions. Extensive experiments with multi-camera walking sequences from the HumanEva-I/II data set show that (i) the proposed tracker achieves significantly lower estimation error than both the Annealed Particle Filter and the standard Particle Filter; and (ii) the learned motion correlation generalizes well to motion performed by subjects other than the training subject.

AB - Inference in 3D articulated human body tracking is challenging due to the high dimensionality and nonlinearity of the parameter-space. We propose a particle filter with Rao-Blackwellisation which marginalizes part of the state variables by exploiting the correlation between the right-side and the left-side joint Euler angles. The correlation is naturally induced by the symmetric and repetitive patterns in specific human activities. A novel algorithm is proposed to learn the correlation from the training data using Partial Least Square regression. The learned correlation is then used as motion prior in designing the Rao-Blackwellised particle filter, which estimates only one group of state variables using the Monte Carlo method, leaving the other group being exactly computed through an analytical filter that utilizes the learned motion correlation. We evaluate the effectiveness of the motion correlation for 3D articulated human body tracking. The accuracy of the proposed 3D tracker is quantitatively assessed based on the distance between the true and the estimated marker positions. Extensive experiments with multi-camera walking sequences from the HumanEva-I/II data set show that (i) the proposed tracker achieves significantly lower estimation error than both the Annealed Particle Filter and the standard Particle Filter; and (ii) the learned motion correlation generalizes well to motion performed by subjects other than the training subject.

UR - http://www.scopus.com/inward/record.url?scp=50649121574&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=50649121574&partnerID=8YFLogxK

U2 - 10.1109/ICCV.2007.4408951

DO - 10.1109/ICCV.2007.4408951

M3 - Conference contribution

BT - Proceedings of the IEEE International Conference on Computer Vision

ER -