3D Human Pose Tracking Priors using Geodesic Mixture Models

Edgar Simo Serra, Carme Torras, Francesc Moreno-Noguer

Research output: Contribution to journalArticle

9 Citations (Scopus)

Abstract

We present a novel approach for learning a finite mixture model on a Riemannian manifold in which Euclidean metrics are not applicable and one needs to resort to geodesic distances consistent with the manifold geometry. For this purpose, we draw inspiration on a variant of the expectation-maximization algorithm, that uses a minimum message length criterion to automatically estimate the optimal number of components from multivariate data lying on an Euclidean space. In order to use this approach on Riemannian manifolds, we propose a formulation in which each component is defined on a different tangent space, thus avoiding the problems associated with the loss of accuracy produced when linearizing the manifold with a single tangent space. Our approach can be applied to any type of manifold for which it is possible to estimate its tangent space. Additionally, we consider using shrinkage covariance estimation to improve the robustness of the method, especially when dealing with very sparsely distributed samples. We evaluate the approach on a number of situations, going from data clustering on manifolds to combining pose and kinematics of articulated bodies for 3D human pose tracking. In all cases, we demonstrate remarkable improvement compared to several chosen baselines.

Original languageEnglish
Pages (from-to)1-21
Number of pages21
JournalInternational Journal of Computer Vision
DOIs
Publication statusAccepted/In press - 2016 Aug 24

Fingerprint

Kinematics
Geometry

Keywords

  • 3D human pose
  • Human kinematics
  • Mixture modelling
  • Probabilistic priors
  • Riemannian manifolds

ASJC Scopus subject areas

  • Software
  • Computer Vision and Pattern Recognition
  • Artificial Intelligence

Cite this

3D Human Pose Tracking Priors using Geodesic Mixture Models. / Simo Serra, Edgar; Torras, Carme; Moreno-Noguer, Francesc.

In: International Journal of Computer Vision, 24.08.2016, p. 1-21.

Research output: Contribution to journalArticle

Simo Serra, Edgar ; Torras, Carme ; Moreno-Noguer, Francesc. / 3D Human Pose Tracking Priors using Geodesic Mixture Models. In: International Journal of Computer Vision. 2016 ; pp. 1-21.
@article{a87825f2b59c4585bb4498a2ded3de62,
title = "3D Human Pose Tracking Priors using Geodesic Mixture Models",
abstract = "We present a novel approach for learning a finite mixture model on a Riemannian manifold in which Euclidean metrics are not applicable and one needs to resort to geodesic distances consistent with the manifold geometry. For this purpose, we draw inspiration on a variant of the expectation-maximization algorithm, that uses a minimum message length criterion to automatically estimate the optimal number of components from multivariate data lying on an Euclidean space. In order to use this approach on Riemannian manifolds, we propose a formulation in which each component is defined on a different tangent space, thus avoiding the problems associated with the loss of accuracy produced when linearizing the manifold with a single tangent space. Our approach can be applied to any type of manifold for which it is possible to estimate its tangent space. Additionally, we consider using shrinkage covariance estimation to improve the robustness of the method, especially when dealing with very sparsely distributed samples. We evaluate the approach on a number of situations, going from data clustering on manifolds to combining pose and kinematics of articulated bodies for 3D human pose tracking. In all cases, we demonstrate remarkable improvement compared to several chosen baselines.",
keywords = "3D human pose, Human kinematics, Mixture modelling, Probabilistic priors, Riemannian manifolds",
author = "{Simo Serra}, Edgar and Carme Torras and Francesc Moreno-Noguer",
year = "2016",
month = "8",
day = "24",
doi = "10.1007/s11263-016-0941-2",
language = "English",
pages = "1--21",
journal = "International Journal of Computer Vision",
issn = "0920-5691",
publisher = "Springer Netherlands",

}

TY - JOUR

T1 - 3D Human Pose Tracking Priors using Geodesic Mixture Models

AU - Simo Serra, Edgar

AU - Torras, Carme

AU - Moreno-Noguer, Francesc

PY - 2016/8/24

Y1 - 2016/8/24

N2 - We present a novel approach for learning a finite mixture model on a Riemannian manifold in which Euclidean metrics are not applicable and one needs to resort to geodesic distances consistent with the manifold geometry. For this purpose, we draw inspiration on a variant of the expectation-maximization algorithm, that uses a minimum message length criterion to automatically estimate the optimal number of components from multivariate data lying on an Euclidean space. In order to use this approach on Riemannian manifolds, we propose a formulation in which each component is defined on a different tangent space, thus avoiding the problems associated with the loss of accuracy produced when linearizing the manifold with a single tangent space. Our approach can be applied to any type of manifold for which it is possible to estimate its tangent space. Additionally, we consider using shrinkage covariance estimation to improve the robustness of the method, especially when dealing with very sparsely distributed samples. We evaluate the approach on a number of situations, going from data clustering on manifolds to combining pose and kinematics of articulated bodies for 3D human pose tracking. In all cases, we demonstrate remarkable improvement compared to several chosen baselines.

AB - We present a novel approach for learning a finite mixture model on a Riemannian manifold in which Euclidean metrics are not applicable and one needs to resort to geodesic distances consistent with the manifold geometry. For this purpose, we draw inspiration on a variant of the expectation-maximization algorithm, that uses a minimum message length criterion to automatically estimate the optimal number of components from multivariate data lying on an Euclidean space. In order to use this approach on Riemannian manifolds, we propose a formulation in which each component is defined on a different tangent space, thus avoiding the problems associated with the loss of accuracy produced when linearizing the manifold with a single tangent space. Our approach can be applied to any type of manifold for which it is possible to estimate its tangent space. Additionally, we consider using shrinkage covariance estimation to improve the robustness of the method, especially when dealing with very sparsely distributed samples. We evaluate the approach on a number of situations, going from data clustering on manifolds to combining pose and kinematics of articulated bodies for 3D human pose tracking. In all cases, we demonstrate remarkable improvement compared to several chosen baselines.

KW - 3D human pose

KW - Human kinematics

KW - Mixture modelling

KW - Probabilistic priors

KW - Riemannian manifolds

UR - http://www.scopus.com/inward/record.url?scp=84983376012&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=84983376012&partnerID=8YFLogxK

U2 - 10.1007/s11263-016-0941-2

DO - 10.1007/s11263-016-0941-2

M3 - Article

AN - SCOPUS:84983376012

SP - 1

EP - 21

JO - International Journal of Computer Vision

JF - International Journal of Computer Vision

SN - 0920-5691

ER -