Learning and Sharing for Improved k-Coverage in Smart Camera Networks

Arezoo Vejdanparast, Peter R Lewis

Research output: Chapter in Book/Report/Conference proceedingConference contribution

Abstract

In this paper we study the self-adaptive behaviour of smart camera networks. Each Camera is equipped with an adjustable zoom lens in order to improve the coverage redundancy formalised ask-coverage across all moving objects under two perspectives: i) learning the movement patterns of the objects captured by a reinforcement learning algorithm at an individual camera level, and ii) utilising a decentralised coordination strategy by enabling an inter-camera communication among the neighbours. Given the dynamic nature of the problem, the first contribution of the paper is to show how learning an environmental constraint such as the movement pattern of the objects leads to a dynamic zoom selection behaviour that significantly improves k-coverage across the network. In our second contribution we show that the speed of convergence of the learning approach can be improved by applying a knowledge-sharing scheme. This is achieved by employing an inter-camera communication strategy across the network. The results indicate that enabling a knowledge-sharing scheme retains the high performance of pure reinforcement learning approaches. It also leads to a considerably faster convergence to the maximum possible k-coverage in learning approaches across the majority of test scenarios.
Original languageEnglish
Title of host publication2019 IEEE 4th International Workshops on Foundations and Applications of Self* Systems (FAS*W)
PublisherIEEE
Pages80-85
Number of pages6
ISBN (Electronic)978-1-7281-2406-3
ISBN (Print)978-1-7281-2407-0
DOIs
Publication statusPublished - 8 Aug 2019
Event2019 IEEE 4th International Workshops on Foundations and Applications of Self* Systems (FAS*W) - Umea, Sweden
Duration: 16 Jun 201920 Jun 2019

Conference

Conference2019 IEEE 4th International Workshops on Foundations and Applications of Self* Systems (FAS*W)
Period16/06/1920/06/19

Fingerprint

Sharing
Coverage
Camera
Cameras
Knowledge Sharing
Reinforcement learning
Reinforcement Learning
Adaptive Behavior
Communication
Speed of Convergence
Moving Objects
Learning algorithms
Decentralized
Redundancy
Lens
Learning Algorithm
Lenses
High Performance
Learning
Scenarios

Keywords

  • Decentralised coordination
  • Decentralised k-coverage
  • Dynamic reconfiguration
  • Online reinforcement learning
  • Smart camera networks

Cite this

Vejdanparast, A., & Lewis, P. R. (2019). Learning and Sharing for Improved k-Coverage in Smart Camera Networks. In 2019 IEEE 4th International Workshops on Foundations and Applications of Self* Systems (FAS*W) (pp. 80-85). [8791959] IEEE. https://doi.org/10.1109/FAS-W.2019.00033
Vejdanparast, Arezoo ; Lewis, Peter R. / Learning and Sharing for Improved k-Coverage in Smart Camera Networks. 2019 IEEE 4th International Workshops on Foundations and Applications of Self* Systems (FAS*W). IEEE, 2019. pp. 80-85
@inproceedings{31cf1bb5bf414442873f12e249734189,
title = "Learning and Sharing for Improved k-Coverage in Smart Camera Networks",
abstract = "In this paper we study the self-adaptive behaviour of smart camera networks. Each Camera is equipped with an adjustable zoom lens in order to improve the coverage redundancy formalised ask-coverage across all moving objects under two perspectives: i) learning the movement patterns of the objects captured by a reinforcement learning algorithm at an individual camera level, and ii) utilising a decentralised coordination strategy by enabling an inter-camera communication among the neighbours. Given the dynamic nature of the problem, the first contribution of the paper is to show how learning an environmental constraint such as the movement pattern of the objects leads to a dynamic zoom selection behaviour that significantly improves k-coverage across the network. In our second contribution we show that the speed of convergence of the learning approach can be improved by applying a knowledge-sharing scheme. This is achieved by employing an inter-camera communication strategy across the network. The results indicate that enabling a knowledge-sharing scheme retains the high performance of pure reinforcement learning approaches. It also leads to a considerably faster convergence to the maximum possible k-coverage in learning approaches across the majority of test scenarios.",
keywords = "Decentralised coordination, Decentralised k-coverage, Dynamic reconfiguration, Online reinforcement learning, Smart camera networks",
author = "Arezoo Vejdanparast and Lewis, {Peter R}",
year = "2019",
month = "8",
day = "8",
doi = "10.1109/FAS-W.2019.00033",
language = "English",
isbn = "978-1-7281-2407-0",
pages = "80--85",
booktitle = "2019 IEEE 4th International Workshops on Foundations and Applications of Self* Systems (FAS*W)",
publisher = "IEEE",
address = "United States",

}

Vejdanparast, A & Lewis, PR 2019, Learning and Sharing for Improved k-Coverage in Smart Camera Networks. in 2019 IEEE 4th International Workshops on Foundations and Applications of Self* Systems (FAS*W)., 8791959, IEEE, pp. 80-85, 2019 IEEE 4th International Workshops on Foundations and Applications of Self* Systems (FAS*W), 16/06/19. https://doi.org/10.1109/FAS-W.2019.00033

Learning and Sharing for Improved k-Coverage in Smart Camera Networks. / Vejdanparast, Arezoo; Lewis, Peter R.

2019 IEEE 4th International Workshops on Foundations and Applications of Self* Systems (FAS*W). IEEE, 2019. p. 80-85 8791959.

Research output: Chapter in Book/Report/Conference proceedingConference contribution

TY - GEN

T1 - Learning and Sharing for Improved k-Coverage in Smart Camera Networks

AU - Vejdanparast, Arezoo

AU - Lewis, Peter R

PY - 2019/8/8

Y1 - 2019/8/8

N2 - In this paper we study the self-adaptive behaviour of smart camera networks. Each Camera is equipped with an adjustable zoom lens in order to improve the coverage redundancy formalised ask-coverage across all moving objects under two perspectives: i) learning the movement patterns of the objects captured by a reinforcement learning algorithm at an individual camera level, and ii) utilising a decentralised coordination strategy by enabling an inter-camera communication among the neighbours. Given the dynamic nature of the problem, the first contribution of the paper is to show how learning an environmental constraint such as the movement pattern of the objects leads to a dynamic zoom selection behaviour that significantly improves k-coverage across the network. In our second contribution we show that the speed of convergence of the learning approach can be improved by applying a knowledge-sharing scheme. This is achieved by employing an inter-camera communication strategy across the network. The results indicate that enabling a knowledge-sharing scheme retains the high performance of pure reinforcement learning approaches. It also leads to a considerably faster convergence to the maximum possible k-coverage in learning approaches across the majority of test scenarios.

AB - In this paper we study the self-adaptive behaviour of smart camera networks. Each Camera is equipped with an adjustable zoom lens in order to improve the coverage redundancy formalised ask-coverage across all moving objects under two perspectives: i) learning the movement patterns of the objects captured by a reinforcement learning algorithm at an individual camera level, and ii) utilising a decentralised coordination strategy by enabling an inter-camera communication among the neighbours. Given the dynamic nature of the problem, the first contribution of the paper is to show how learning an environmental constraint such as the movement pattern of the objects leads to a dynamic zoom selection behaviour that significantly improves k-coverage across the network. In our second contribution we show that the speed of convergence of the learning approach can be improved by applying a knowledge-sharing scheme. This is achieved by employing an inter-camera communication strategy across the network. The results indicate that enabling a knowledge-sharing scheme retains the high performance of pure reinforcement learning approaches. It also leads to a considerably faster convergence to the maximum possible k-coverage in learning approaches across the majority of test scenarios.

KW - Decentralised coordination

KW - Decentralised k-coverage

KW - Dynamic reconfiguration

KW - Online reinforcement learning

KW - Smart camera networks

UR - https://ieeexplore.ieee.org/document/8791959/

UR - http://www.scopus.com/inward/record.url?scp=85071446663&partnerID=8YFLogxK

U2 - 10.1109/FAS-W.2019.00033

DO - 10.1109/FAS-W.2019.00033

M3 - Conference contribution

SN - 978-1-7281-2407-0

SP - 80

EP - 85

BT - 2019 IEEE 4th International Workshops on Foundations and Applications of Self* Systems (FAS*W)

PB - IEEE

ER -

Vejdanparast A, Lewis PR. Learning and Sharing for Improved k-Coverage in Smart Camera Networks. In 2019 IEEE 4th International Workshops on Foundations and Applications of Self* Systems (FAS*W). IEEE. 2019. p. 80-85. 8791959 https://doi.org/10.1109/FAS-W.2019.00033