A Heavy-Tailed Noise Tolerant Labeled Multi-Bernoulli Filter

Wanying Zhang; Feng Yang; Yan Liang; Zhentao Liu

doi:10.23919/ICIF.2018.8455322

A Heavy-Tailed Noise Tolerant Labeled Multi-Bernoulli Filter

Wanying Zhang, Feng Yang, Yan Liang, Zhentao Liu

School of Automation

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

3 Scopus citations

Abstract

The well-known labeled multi-Bernoulli (LMB) filter for multi-target tracking in clutters works well only under the Gaussian noise assumptions. Since this Gaussian assumption can hardly hold in practice, we present the problem of the LMB with heavy-tailed non-Gaussian measurement noise. Through modeling the measurement noise as Student's t distribution, a heavy-tailed measurement noise tolerant LMB (TLMB) is derived in the framework of variational Bayesian inference for the joint estimation of the target state together with the unknown scale matrix and degree of freedom (dof) of the Student's t distribution. Simulations on multi-target tracking in clutter with unreliable sensor demonstrate the effectiveness and superiority of the proposed TLMB.

Original language	English
Title of host publication	2018 21st International Conference on Information Fusion, FUSION 2018
Publisher	Institute of Electrical and Electronics Engineers Inc.
Pages	2461-2467
Number of pages	7
ISBN (Print)	9780996452762
DOIs	https://doi.org/10.23919/ICIF.2018.8455322
State	Published - 5 Sep 2018
Event	21st International Conference on Information Fusion, FUSION 2018 - Cambridge, United Kingdom Duration: 10 Jul 2018 → 13 Jul 2018

Publication series

Name	2018 21st International Conference on Information Fusion, FUSION 2018

Conference

Conference	21st International Conference on Information Fusion, FUSION 2018
Country/Territory	United Kingdom
City	Cambridge
Period	10/07/18 → 13/07/18

Keywords

heavy-tailed measurement noise
labeled multi-Bernoulli filter
multi-target tracking
S-tudent's distribution
variational Bayesian

Access to Document

10.23919/ICIF.2018.8455322

Cite this

Zhang, W., Yang, F., Liang, Y., & Liu, Z. (2018). A Heavy-Tailed Noise Tolerant Labeled Multi-Bernoulli Filter. In 2018 21st International Conference on Information Fusion, FUSION 2018 (pp. 2461-2467). Article 8455322 (2018 21st International Conference on Information Fusion, FUSION 2018). Institute of Electrical and Electronics Engineers Inc.. https://doi.org/10.23919/ICIF.2018.8455322

@inproceedings{fb557ef2a2df469fa35f7977928f2e35,

title = "A Heavy-Tailed Noise Tolerant Labeled Multi-Bernoulli Filter",

abstract = "The well-known labeled multi-Bernoulli (LMB) filter for multi-target tracking in clutters works well only under the Gaussian noise assumptions. Since this Gaussian assumption can hardly hold in practice, we present the problem of the LMB with heavy-tailed non-Gaussian measurement noise. Through modeling the measurement noise as Student's t distribution, a heavy-tailed measurement noise tolerant LMB (TLMB) is derived in the framework of variational Bayesian inference for the joint estimation of the target state together with the unknown scale matrix and degree of freedom (dof) of the Student's t distribution. Simulations on multi-target tracking in clutter with unreliable sensor demonstrate the effectiveness and superiority of the proposed TLMB.",

keywords = "heavy-tailed measurement noise, labeled multi-Bernoulli filter, multi-target tracking, S-tudent's distribution, variational Bayesian",

author = "Wanying Zhang and Feng Yang and Yan Liang and Zhentao Liu",

note = "Publisher Copyright: {\textcopyright} 2018 ISIF; 21st International Conference on Information Fusion, FUSION 2018 ; Conference date: 10-07-2018 Through 13-07-2018",

year = "2018",

month = sep,

day = "5",

doi = "10.23919/ICIF.2018.8455322",

language = "英语",

isbn = "9780996452762",

series = "2018 21st International Conference on Information Fusion, FUSION 2018",

publisher = "Institute of Electrical and Electronics Engineers Inc.",

pages = "2461--2467",

booktitle = "2018 21st International Conference on Information Fusion, FUSION 2018",

}

Zhang, W, Yang, F, Liang, Y & Liu, Z 2018, A Heavy-Tailed Noise Tolerant Labeled Multi-Bernoulli Filter. in 2018 21st International Conference on Information Fusion, FUSION 2018., 8455322, 2018 21st International Conference on Information Fusion, FUSION 2018, Institute of Electrical and Electronics Engineers Inc., pp. 2461-2467, 21st International Conference on Information Fusion, FUSION 2018, Cambridge, United Kingdom, 10/07/18. https://doi.org/10.23919/ICIF.2018.8455322

A Heavy-Tailed Noise Tolerant Labeled Multi-Bernoulli Filter. / Zhang, Wanying; Yang, Feng; Liang, Yan et al.
2018 21st International Conference on Information Fusion, FUSION 2018. Institute of Electrical and Electronics Engineers Inc., 2018. p. 2461-2467 8455322 (2018 21st International Conference on Information Fusion, FUSION 2018).

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

TY - GEN

T1 - A Heavy-Tailed Noise Tolerant Labeled Multi-Bernoulli Filter

AU - Zhang, Wanying

AU - Yang, Feng

AU - Liang, Yan

AU - Liu, Zhentao

PY - 2018/9/5

Y1 - 2018/9/5

N2 - The well-known labeled multi-Bernoulli (LMB) filter for multi-target tracking in clutters works well only under the Gaussian noise assumptions. Since this Gaussian assumption can hardly hold in practice, we present the problem of the LMB with heavy-tailed non-Gaussian measurement noise. Through modeling the measurement noise as Student's t distribution, a heavy-tailed measurement noise tolerant LMB (TLMB) is derived in the framework of variational Bayesian inference for the joint estimation of the target state together with the unknown scale matrix and degree of freedom (dof) of the Student's t distribution. Simulations on multi-target tracking in clutter with unreliable sensor demonstrate the effectiveness and superiority of the proposed TLMB.

AB - The well-known labeled multi-Bernoulli (LMB) filter for multi-target tracking in clutters works well only under the Gaussian noise assumptions. Since this Gaussian assumption can hardly hold in practice, we present the problem of the LMB with heavy-tailed non-Gaussian measurement noise. Through modeling the measurement noise as Student's t distribution, a heavy-tailed measurement noise tolerant LMB (TLMB) is derived in the framework of variational Bayesian inference for the joint estimation of the target state together with the unknown scale matrix and degree of freedom (dof) of the Student's t distribution. Simulations on multi-target tracking in clutter with unreliable sensor demonstrate the effectiveness and superiority of the proposed TLMB.

KW - heavy-tailed measurement noise

KW - labeled multi-Bernoulli filter

KW - multi-target tracking

KW - S-tudent's distribution

KW - variational Bayesian

UR - http://www.scopus.com/inward/record.url?scp=85054103686&partnerID=8YFLogxK

U2 - 10.23919/ICIF.2018.8455322

DO - 10.23919/ICIF.2018.8455322

M3 - 会议稿件

AN - SCOPUS:85054103686

SN - 9780996452762

T3 - 2018 21st International Conference on Information Fusion, FUSION 2018

SP - 2461

EP - 2467

BT - 2018 21st International Conference on Information Fusion, FUSION 2018

PB - Institute of Electrical and Electronics Engineers Inc.

T2 - 21st International Conference on Information Fusion, FUSION 2018

Y2 - 10 July 2018 through 13 July 2018

ER -

A Heavy-Tailed Noise Tolerant Labeled Multi-Bernoulli Filter

Abstract

Publication series

Conference

Keywords

Access to Document

Other files and links

Fingerprint

Cite this