Distributed Expectation Propagation for Multi-Object Tracking over Sensor Networks

Lily Li; Runze Gan; James R. Hopgood; Michael E. Davies; S.J. Godsill

Distributed Expectation Propagation for Multi-Object Tracking over Sensor Networks

Lily Li^*, Runze Gan, James R. Hopgood, Michael E. Davies, S.J. Godsill

^*Corresponding author for this work

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

Abstract

In this paper, we present a novel distributed expectation propagation algorithm for multiple sensors, multiple objects tracking in cluttered environments. The proposed framework enables each sensor to operate locally while collaboratively exchanging moment estimates with other sensors, thus eliminating the need to transmit all data to a central processing node. Specifically, we introduce a fast and parallelisable Rao-Blackwellised Gibbs sampling scheme to approximate the tilted distributions, which enhances the accuracy and efficiency of expectation propagation updates. Results demonstrate that the proposed algorithm improves both communication and inference efficiency for multi-object tracking tasks with dynamic sensor connectivity and varying clutter levels.

Original language	English
Title of host publication	23rd IEEE Statistical Signal Processing Workshop (SSP 2025)
Publisher	Institute of Electrical and Electronics Engineers
Publication status	Accepted/In press - 3 Apr 2025

Embargoed Document

Accepted Author Manuscript
Accepted author manuscript, 296 KB
Licence: Creative Commons: Attribution (CC-BY)

1 Conference contribution

PiVoT: Poisson Measurements-based Variational Multi-object Detection and Tracking
Gan, R., Li, L., Hopgood, J. R., Davies, M. E. & Godsill, S. J., 30 Apr 2025, (Accepted/In press) 28th International Conference on Information Fusion. International Society of Information Fusion
Research output: Chapter in Book/Report/Conference proceeding › Conference contribution

Cite this

@inproceedings{7cb93df6c9b149d5af0c3ecd3ec14bfc,

title = "Distributed Expectation Propagation for Multi-Object Tracking over Sensor Networks",

abstract = "In this paper, we present a novel distributed expectation propagation algorithm for multiple sensors, multiple objects tracking in cluttered environments. The proposed framework enables each sensor to operate locally while collaboratively exchanging moment estimates with other sensors, thus eliminating the need to transmit all data to a central processing node. Specifically, we introduce a fast and parallelisable Rao-Blackwellised Gibbs sampling scheme to approximate the tilted distributions, which enhances the accuracy and efficiency of expectation propagation updates. Results demonstrate that the proposed algorithm improves both communication and inference efficiency for multi-object tracking tasks with dynamic sensor connectivity and varying clutter levels.",

author = "Lily Li and Runze Gan and Hopgood, {James R.} and Davies, {Michael E.} and S.J. Godsill",

year = "2025",

month = apr,

day = "3",

language = "English",

booktitle = "23rd IEEE Statistical Signal Processing Workshop (SSP 2025)",

publisher = "Institute of Electrical and Electronics Engineers",

address = "United States",

}

TY - GEN

T1 - Distributed Expectation Propagation for Multi-Object Tracking over Sensor Networks

AU - Li, Lily

AU - Gan, Runze

AU - Hopgood, James R.

AU - Davies, Michael E.

AU - Godsill, S.J.

PY - 2025/4/3

Y1 - 2025/4/3

N2 - In this paper, we present a novel distributed expectation propagation algorithm for multiple sensors, multiple objects tracking in cluttered environments. The proposed framework enables each sensor to operate locally while collaboratively exchanging moment estimates with other sensors, thus eliminating the need to transmit all data to a central processing node. Specifically, we introduce a fast and parallelisable Rao-Blackwellised Gibbs sampling scheme to approximate the tilted distributions, which enhances the accuracy and efficiency of expectation propagation updates. Results demonstrate that the proposed algorithm improves both communication and inference efficiency for multi-object tracking tasks with dynamic sensor connectivity and varying clutter levels.

AB - In this paper, we present a novel distributed expectation propagation algorithm for multiple sensors, multiple objects tracking in cluttered environments. The proposed framework enables each sensor to operate locally while collaboratively exchanging moment estimates with other sensors, thus eliminating the need to transmit all data to a central processing node. Specifically, we introduce a fast and parallelisable Rao-Blackwellised Gibbs sampling scheme to approximate the tilted distributions, which enhances the accuracy and efficiency of expectation propagation updates. Results demonstrate that the proposed algorithm improves both communication and inference efficiency for multi-object tracking tasks with dynamic sensor connectivity and varying clutter levels.

M3 - Conference contribution

BT - 23rd IEEE Statistical Signal Processing Workshop (SSP 2025)

PB - Institute of Electrical and Electronics Engineers

ER -

Distributed Expectation Propagation for Multi-Object Tracking over Sensor Networks

Abstract

Embargoed Document

Fingerprint

Research output

PiVoT: Poisson Measurements-based Variational Multi-object Detection and Tracking

Cite this