Domain adaptation for target classification using micro-Doppler spectra in radar networks

Peter Svenningsson; Francesco Fioranelli; Alexander Yarovoy

Domain adaptation for target classification using micro-Doppler spectra in radar networks

Peter Svenningsson, Francesco Fioranelli, Alexander Yarovoy

Microwave Sensing, Signals & Systems

Research output: Chapter in Book/Conference proceedings/Edited volume › Conference contribution › Scientific › peer-review

25 Downloads (Pure)

Abstract

In this paper, the classification of human activity from micro-Doppler spectrograms measured by a radar network is considered. To cope with differences between the training and test datasets due to changes in the set of participants, signal-to-noise ratio and polarimetry, domain adaptation is proposed. To realize this, linear mapping between the two domains is assumed and estimated by one of two methods, expectation-maximization or empirical estimates of statistical moments. The performance of the methods is evaluated on experimental data measured by a multi-static radar network. The proposed methods increase the classification accuracy by 5–15 percentiles on the recorded dataset.

Original language	English
Title of host publication	2021 IEEE 24th International Conference on Information Fusion (FUSION)
Subtitle of host publication	Proceedings
Publisher	IEEE
Pages	1-8
Number of pages	8
ISBN (Electronic)	978-1-7377497-1-4
ISBN (Print)	978-1-6654-1427-2
Publication status	Published - 2021
Event	2021 IEEE 24th International Conference on Information Fusion (FUSION) - Hybrid at Sun City, South Africa Duration: 1 Nov 2021 → 4 Nov 2021 Conference number: 24th

Conference

Conference	2021 IEEE 24th International Conference on Information Fusion (FUSION)
Abbreviated title	Fusion 2021
Country/Territory	South Africa
City	Hybrid at Sun City
Period	1/11/21 → 4/11/21

Bibliographical note

Green Open Access added to TU Delft Institutional Repository 'You share, we take care!' - Taverne project https://www.openaccess.nl/en/you-share-we-take-care

Otherwise as indicated in the copyright section: the publisher is the copyright holder of this work and the author uses the Dutch legislation to make this work public.

Keywords

radar target classification
micro-Doppler signature
domain adaptation
multi-static radar network

Access to Document

Domain_adaptation_for_target_classification_using_micro-Doppler_spectra_in_radar_networksFinal published version, 553 KB

https://ieeexplore.ieee.org/document/9626930

Cite this

@inproceedings{7940305717094b7fa426441a52d78fc8,

title = "Domain adaptation for target classification using micro-Doppler spectra in radar networks",

abstract = "In this paper, the classification of human activity from micro-Doppler spectrograms measured by a radar network is considered. To cope with differences between the training and test datasets due to changes in the set of participants, signal-to-noise ratio and polarimetry, domain adaptation is proposed. To realize this, linear mapping between the two domains is assumed and estimated by one of two methods, expectation-maximization or empirical estimates of statistical moments. The performance of the methods is evaluated on experimental data measured by a multi-static radar network. The proposed methods increase the classification accuracy by 5–15 percentiles on the recorded dataset.",

keywords = "radar target classification, micro-Doppler signature, domain adaptation, multi-static radar network",

author = "Peter Svenningsson and Francesco Fioranelli and Alexander Yarovoy",

note = "Green Open Access added to TU Delft Institutional Repository 'You share, we take care!' - Taverne project https://www.openaccess.nl/en/you-share-we-take-care Otherwise as indicated in the copyright section: the publisher is the copyright holder of this work and the author uses the Dutch legislation to make this work public.; 2021 IEEE 24th International Conference on Information Fusion (FUSION), Fusion 2021 ; Conference date: 01-11-2021 Through 04-11-2021",

year = "2021",

language = "English",

isbn = "978-1-6654-1427-2",

pages = "1--8",

booktitle = "2021 IEEE 24th International Conference on Information Fusion (FUSION)",

publisher = "IEEE",

address = "United States",

}

Svenningsson, P, Fioranelli, F & Yarovoy, A 2021, Domain adaptation for target classification using micro-Doppler spectra in radar networks. in 2021 IEEE 24th International Conference on Information Fusion (FUSION): Proceedings., 9626930, IEEE, pp. 1-8, 2021 IEEE 24th International Conference on Information Fusion (FUSION), Hybrid at Sun City, South Africa, 1/11/21. <https://ieeexplore.ieee.org/document/9626930>

Domain adaptation for target classification using micro-Doppler spectra in radar networks. / Svenningsson, Peter; Fioranelli, Francesco ; Yarovoy, Alexander.
2021 IEEE 24th International Conference on Information Fusion (FUSION): Proceedings. IEEE, 2021. p. 1-8 9626930.

Research output: Chapter in Book/Conference proceedings/Edited volume › Conference contribution › Scientific › peer-review

TY - GEN

T1 - Domain adaptation for target classification using micro-Doppler spectra in radar networks

AU - Svenningsson, Peter

AU - Fioranelli, Francesco

AU - Yarovoy, Alexander

N1 - Conference code: 24th

PY - 2021

Y1 - 2021

N2 - In this paper, the classification of human activity from micro-Doppler spectrograms measured by a radar network is considered. To cope with differences between the training and test datasets due to changes in the set of participants, signal-to-noise ratio and polarimetry, domain adaptation is proposed. To realize this, linear mapping between the two domains is assumed and estimated by one of two methods, expectation-maximization or empirical estimates of statistical moments. The performance of the methods is evaluated on experimental data measured by a multi-static radar network. The proposed methods increase the classification accuracy by 5–15 percentiles on the recorded dataset.

AB - In this paper, the classification of human activity from micro-Doppler spectrograms measured by a radar network is considered. To cope with differences between the training and test datasets due to changes in the set of participants, signal-to-noise ratio and polarimetry, domain adaptation is proposed. To realize this, linear mapping between the two domains is assumed and estimated by one of two methods, expectation-maximization or empirical estimates of statistical moments. The performance of the methods is evaluated on experimental data measured by a multi-static radar network. The proposed methods increase the classification accuracy by 5–15 percentiles on the recorded dataset.

KW - radar target classification

KW - micro-Doppler signature

KW - domain adaptation

KW - multi-static radar network

M3 - Conference contribution

SN - 978-1-6654-1427-2

SP - 1

EP - 8

BT - 2021 IEEE 24th International Conference on Information Fusion (FUSION)

PB - IEEE

T2 - 2021 IEEE 24th International Conference on Information Fusion (FUSION)

Y2 - 1 November 2021 through 4 November 2021

ER -

Domain adaptation for target classification using micro-Doppler spectra in radar networks

Abstract

Conference

Bibliographical note

Keywords

Access to Document

Fingerprint

Cite this