A One-Class Classification Method for Human Gait Authentication Using Micro-Doppler Signatures

Haoran  Ji; Chunping  Hou; Yang  Yang; Francesco Fioranelli; Yue  Lang

doi:10.1109/LSP.2021.3122344

A One-Class Classification Method for Human Gait Authentication Using Micro-Doppler Signatures

Haoran Ji, Chunping Hou, Yang Yang, Francesco Fioranelli, Yue Lang

Microwave Sensing, Signals & Systems

Research output: Contribution to journal › Article › Scientific › peer-review

5 Citations (Scopus)

30 Downloads (Pure)

Abstract

In this letter, a radar-based gait authentication method is proposed. We focus on the overfitting problem on the target category caused by limited training data in authentication models and propose a one-class classification model to alleviate this problem. The effectiveness of such model is verified by establishing a radar-based gait dataset, which is composed of gait micro-Doppler spectrograms derived from nine human subjects. The experimental results demonstrate that, under the condition of limited training data, the performances of an authentication model degrade because misclassification of the non-target samples easily occurs. The proposed method effectively avoids this risk, performing the other existing authentication and one-class classification methods on the metric Equal Error Rate.

Original language	English
Article number	9585408
Pages (from-to)	2182-2186
Number of pages	5
Journal	IEEE Signal Processing Letters
Volume	28
DOIs	https://doi.org/10.1109/LSP.2021.3122344
Publication status	Published - 2021

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

Gait authentication
micro-Doppler radar
one class classification

Access to Document

10.1109/LSP.2021.3122344

A_One-Class_Classification_Method_for_Human_Gait_Authentication_Using_Micro-Doppler_SignaturesFinal published version, 434 KB

Cite this

@article{39f7d5f77ecc43c2956e046bf3b2a289,

title = "A One-Class Classification Method for Human Gait Authentication Using Micro-Doppler Signatures",

abstract = "In this letter, a radar-based gait authentication method is proposed. We focus on the overfitting problem on the target category caused by limited training data in authentication models and propose a one-class classification model to alleviate this problem. The effectiveness of such model is verified by establishing a radar-based gait dataset, which is composed of gait micro-Doppler spectrograms derived from nine human subjects. The experimental results demonstrate that, under the condition of limited training data, the performances of an authentication model degrade because misclassification of the non-target samples easily occurs. The proposed method effectively avoids this risk, performing the other existing authentication and one-class classification methods on the metric Equal Error Rate.",

keywords = "Gait authentication, micro-Doppler radar, one class classification",

author = "Haoran Ji and Chunping Hou and Yang Yang and Francesco Fioranelli and Yue Lang",

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.",

year = "2021",

doi = "10.1109/LSP.2021.3122344",

language = "English",

volume = "28",

pages = "2182--2186",

journal = "IEEE Signal Processing Letters",

issn = "1558-2361",

publisher = "Institute of Electrical and Electronics Engineers (IEEE)",

}

TY - JOUR

T1 - A One-Class Classification Method for Human Gait Authentication Using Micro-Doppler Signatures

AU - Ji, Haoran

AU - Hou, Chunping

AU - Yang, Yang

AU - Fioranelli, Francesco

AU - Lang, Yue

N1 - 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.

PY - 2021

Y1 - 2021

N2 - In this letter, a radar-based gait authentication method is proposed. We focus on the overfitting problem on the target category caused by limited training data in authentication models and propose a one-class classification model to alleviate this problem. The effectiveness of such model is verified by establishing a radar-based gait dataset, which is composed of gait micro-Doppler spectrograms derived from nine human subjects. The experimental results demonstrate that, under the condition of limited training data, the performances of an authentication model degrade because misclassification of the non-target samples easily occurs. The proposed method effectively avoids this risk, performing the other existing authentication and one-class classification methods on the metric Equal Error Rate.

AB - In this letter, a radar-based gait authentication method is proposed. We focus on the overfitting problem on the target category caused by limited training data in authentication models and propose a one-class classification model to alleviate this problem. The effectiveness of such model is verified by establishing a radar-based gait dataset, which is composed of gait micro-Doppler spectrograms derived from nine human subjects. The experimental results demonstrate that, under the condition of limited training data, the performances of an authentication model degrade because misclassification of the non-target samples easily occurs. The proposed method effectively avoids this risk, performing the other existing authentication and one-class classification methods on the metric Equal Error Rate.

KW - Gait authentication

KW - micro-Doppler radar

KW - one class classification

UR - http://www.scopus.com/inward/record.url?scp=85118535561&partnerID=8YFLogxK

U2 - 10.1109/LSP.2021.3122344

DO - 10.1109/LSP.2021.3122344

M3 - Article

SN - 1558-2361

VL - 28

SP - 2182

EP - 2186

JO - IEEE Signal Processing Letters

JF - IEEE Signal Processing Letters

M1 - 9585408

ER -

A One-Class Classification Method for Human Gait Authentication Using Micro-Doppler Signatures

Abstract

Bibliographical note

Keywords

Access to Document

Other files and links

Fingerprint

Cite this