Efficient particle filtering using RANSAC with application to 3D face tracking

Le Lu, Xiangtian Dai, Gregory Hager

Research output: Contribution to journalArticle

Abstract

Particle filtering is a very popular technique for sequential state estimation. However, in high-dimensional cases where the state dynamics are complex or poorly modeled, thousands of particles are usually required for real applications. This paper presents a hybrid sampling solution that combines RANSAC and particle filtering. In this approach, RANSAC provides proposal particles that, with high probability, represent the observation likelihood. Both conditionally independent RANSAC sampling and boosting-like conditionally dependent RANSAC sampling are explored. We show that the use of RANSAC-guided sampling reduces the necessary number of particles to dozens for a full 3D tracking problem. This method is particularly advantageous when state dynamics are poorly modeled. We show empirically that the sampling efficiency (in terms of likelihood) is much higher with the use of RANSAC. The algorithm has been applied to the problem of 3D face pose tracking with changing expression. We demonstrate the validity of our approach with several video sequences acquired in an unstructured environment.

Original languageEnglish (US)
Pages (from-to)581-592
Number of pages12
JournalImage and Vision Computing
Volume24
Issue number6
DOIs
StatePublished - Jun 1 2006

Fingerprint

Sampling
State estimation

Keywords

  • Particle filtering
  • Random projection
  • RANSAC
  • Robust 3D face tracking

ASJC Scopus subject areas

  • Computer Vision and Pattern Recognition
  • Signal Processing
  • Electrical and Electronic Engineering

Cite this

Efficient particle filtering using RANSAC with application to 3D face tracking. / Lu, Le; Dai, Xiangtian; Hager, Gregory.

In: Image and Vision Computing, Vol. 24, No. 6, 01.06.2006, p. 581-592.

Research output: Contribution to journalArticle

@article{98e4cf0b5b7946f49ddfaff81490a765,
title = "Efficient particle filtering using RANSAC with application to 3D face tracking",
abstract = "Particle filtering is a very popular technique for sequential state estimation. However, in high-dimensional cases where the state dynamics are complex or poorly modeled, thousands of particles are usually required for real applications. This paper presents a hybrid sampling solution that combines RANSAC and particle filtering. In this approach, RANSAC provides proposal particles that, with high probability, represent the observation likelihood. Both conditionally independent RANSAC sampling and boosting-like conditionally dependent RANSAC sampling are explored. We show that the use of RANSAC-guided sampling reduces the necessary number of particles to dozens for a full 3D tracking problem. This method is particularly advantageous when state dynamics are poorly modeled. We show empirically that the sampling efficiency (in terms of likelihood) is much higher with the use of RANSAC. The algorithm has been applied to the problem of 3D face pose tracking with changing expression. We demonstrate the validity of our approach with several video sequences acquired in an unstructured environment.",
keywords = "Particle filtering, Random projection, RANSAC, Robust 3D face tracking",
author = "Le Lu and Xiangtian Dai and Gregory Hager",
year = "2006",
month = "6",
day = "1",
doi = "10.1016/j.imavis.2005.08.003",
language = "English (US)",
volume = "24",
pages = "581--592",
journal = "Image and Vision Computing",
issn = "0262-8856",
publisher = "Elsevier Limited",
number = "6",

}

TY - JOUR

T1 - Efficient particle filtering using RANSAC with application to 3D face tracking

AU - Lu, Le

AU - Dai, Xiangtian

AU - Hager, Gregory

PY - 2006/6/1

Y1 - 2006/6/1

N2 - Particle filtering is a very popular technique for sequential state estimation. However, in high-dimensional cases where the state dynamics are complex or poorly modeled, thousands of particles are usually required for real applications. This paper presents a hybrid sampling solution that combines RANSAC and particle filtering. In this approach, RANSAC provides proposal particles that, with high probability, represent the observation likelihood. Both conditionally independent RANSAC sampling and boosting-like conditionally dependent RANSAC sampling are explored. We show that the use of RANSAC-guided sampling reduces the necessary number of particles to dozens for a full 3D tracking problem. This method is particularly advantageous when state dynamics are poorly modeled. We show empirically that the sampling efficiency (in terms of likelihood) is much higher with the use of RANSAC. The algorithm has been applied to the problem of 3D face pose tracking with changing expression. We demonstrate the validity of our approach with several video sequences acquired in an unstructured environment.

AB - Particle filtering is a very popular technique for sequential state estimation. However, in high-dimensional cases where the state dynamics are complex or poorly modeled, thousands of particles are usually required for real applications. This paper presents a hybrid sampling solution that combines RANSAC and particle filtering. In this approach, RANSAC provides proposal particles that, with high probability, represent the observation likelihood. Both conditionally independent RANSAC sampling and boosting-like conditionally dependent RANSAC sampling are explored. We show that the use of RANSAC-guided sampling reduces the necessary number of particles to dozens for a full 3D tracking problem. This method is particularly advantageous when state dynamics are poorly modeled. We show empirically that the sampling efficiency (in terms of likelihood) is much higher with the use of RANSAC. The algorithm has been applied to the problem of 3D face pose tracking with changing expression. We demonstrate the validity of our approach with several video sequences acquired in an unstructured environment.

KW - Particle filtering

KW - Random projection

KW - RANSAC

KW - Robust 3D face tracking

UR - http://www.scopus.com/inward/record.url?scp=33744998365&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=33744998365&partnerID=8YFLogxK

U2 - 10.1016/j.imavis.2005.08.003

DO - 10.1016/j.imavis.2005.08.003

M3 - Article

AN - SCOPUS:33744998365

VL - 24

SP - 581

EP - 592

JO - Image and Vision Computing

JF - Image and Vision Computing

SN - 0262-8856

IS - 6

ER -