Graphical stochastic models for tracking applications with variational message passing inference

F. Trusheim; A. Condurache; A. Mertins

doi:10.1109/IPTA.2016.7820985

Graphical stochastic models for tracking applications with variational message passing inference

F. Trusheim, A. Condurache, A. Mertins

Institute of Signal Processing

Robert Bosch GmbH

Abstract

In this paper we present a novel, highly-adoptable, state-estimation filter based on the framework of graphical stochastical models and variational message passing inference. We evaluate our method on both real and simulated data for tracking applications. Our experimental results show that the proposed approach offers qualitative and computational advantages over established filter methods in practical situations, where the noise within a process is not simply a Gaussian noise, but rather described by a more complex distribution.

Original language	English
Title of host publication	2016 Sixth International Conference on Image Processing Theory, Tools and Applications (IPTA)
Number of pages	6
Publisher	IEEE
Publication date	01.12.2016
Pages	1-6
Article number	7820985
ISBN (Print)	978-1-4673-8911-2
ISBN (Electronic)	978-1-4673-8910-5
DOIs	https://doi.org/10.1109/IPTA.2016.7820985
Publication status	Published - 01.12.2016
Event	6th International Conference on Image Processing Theory, Tools and Applications - Oulu, Finland Duration: 12.12.2016 → 15.12.2016 Conference number: 125997

UN SDGs

This output contributes to the following UN Sustainable Development Goals (SDGs)

Access to Document

10.1109/IPTA.2016.7820985

Cite this

@inproceedings{33a88911e26847c0bce5f45cfe00fda2,

title = "Graphical stochastic models for tracking applications with variational message passing inference",

abstract = "In this paper we present a novel, highly-adoptable, state-estimation filter based on the framework of graphical stochastical models and variational message passing inference. We evaluate our method on both real and simulated data for tracking applications. Our experimental results show that the proposed approach offers qualitative and computational advantages over established filter methods in practical situations, where the noise within a process is not simply a Gaussian noise, but rather described by a more complex distribution.",

author = "F. Trusheim and A. Condurache and A. Mertins",

year = "2016",

month = dec,

day = "1",

doi = "10.1109/IPTA.2016.7820985",

language = "English",

isbn = "978-1-4673-8911-2",

series = "IEEE Proceedings",

publisher = "IEEE",

pages = "1--6",

booktitle = "2016 Sixth International Conference on Image Processing Theory, Tools and Applications (IPTA)",

note = "6th International Conference on Image Processing Theory, Tools and Applications, IPTA 2016 ; Conference date: 12-12-2016 Through 15-12-2016",

}

Trusheim, F, Condurache, A & Mertins, A 2016, Graphical stochastic models for tracking applications with variational message passing inference. in 2016 Sixth International Conference on Image Processing Theory, Tools and Applications (IPTA)., 7820985, IEEE Proceedings, IEEE, pp. 1-6, 6th International Conference on Image Processing Theory, Tools and Applications, Oulu, Finland, 12.12.16. https://doi.org/10.1109/IPTA.2016.7820985

Graphical stochastic models for tracking applications with variational message passing inference. / Trusheim, F.; Condurache, A.; Mertins, A.
2016 Sixth International Conference on Image Processing Theory, Tools and Applications (IPTA). IEEE, 2016. p. 1-6 7820985 (IEEE Proceedings).

Research output: Chapters in Books/Reports/Conference Proceedings › Conference contribution › peer-review

TY - GEN

T1 - Graphical stochastic models for tracking applications with variational message passing inference

AU - Trusheim, F.

AU - Condurache, A.

AU - Mertins, A.

N1 - Conference code: 125997

PY - 2016/12/1

Y1 - 2016/12/1

N2 - In this paper we present a novel, highly-adoptable, state-estimation filter based on the framework of graphical stochastical models and variational message passing inference. We evaluate our method on both real and simulated data for tracking applications. Our experimental results show that the proposed approach offers qualitative and computational advantages over established filter methods in practical situations, where the noise within a process is not simply a Gaussian noise, but rather described by a more complex distribution.

AB - In this paper we present a novel, highly-adoptable, state-estimation filter based on the framework of graphical stochastical models and variational message passing inference. We evaluate our method on both real and simulated data for tracking applications. Our experimental results show that the proposed approach offers qualitative and computational advantages over established filter methods in practical situations, where the noise within a process is not simply a Gaussian noise, but rather described by a more complex distribution.

UR - https://www.scopus.com/record/display.uri?eid=2-s2.0-85013214723&origin=inward&txGid=2bd28c9aff0539b0af2641e5ff5317e1

U2 - 10.1109/IPTA.2016.7820985

DO - 10.1109/IPTA.2016.7820985

M3 - Conference contribution

SN - 978-1-4673-8911-2

T3 - IEEE Proceedings

SP - 1

EP - 6

BT - 2016 Sixth International Conference on Image Processing Theory, Tools and Applications (IPTA)

PB - IEEE

T2 - 6th International Conference on Image Processing Theory, Tools and Applications

Y2 - 12 December 2016 through 15 December 2016

ER -

Graphical stochastic models for tracking applications with variational message passing inference

Abstract

UN SDGs

Access to Document

Other files and links

Fingerprint

Cite this