Presentation 2017-10-12
Improvement of Accuracy by Machine Learning for Personal Authentication using High Frequency Intra-Body Propagation Characteristics
Shun Onoda, Takahiro Yoshida, Seiichiro Hangai,
PDF Download Page PDF download Page Link
Abstract(in Japanese) (See Japanese page)
Abstract(in English) As one of biometrics that can perform continuous personal authentication only by handling equipment, the personal authentication method using intra-body-frequency characteristics between two fingers have been researching in our laboratory. However, in our previous study, the authentication accuracy with verification method using the Manhattan distance for spectra of the intra-body-frequency characteristics (pass-through / reflection) measured by VNA was very low, e.g. the equal error rate (EER) of the verification using reflection characteristic S22 by the nine subjects was 25.1%. Therefore, in this study, we applied logistic regression, which is one of machine learning method, to the verification in order to improve the verification performance. As a result, the 5.8% EER was archived by applying the logistic regression, that was 19.3 points improvement. It was found that the logistic regression was also effective in the verification using intra-body-propagation characteristics.
Keyword(in Japanese) (See Japanese page)
Keyword(in English) Biometrics / intra-body-propagation characteristics / logistic regression
Paper # BioX2017-26
Date of Issue 2017-10-05 (BioX)

Conference Information
Committee BioX
Conference Date 2017/10/12(2days)
Place (in Japanese) (See Japanese page)
Place (in English) Nobumoto Ohama Memorial Hall
Topics (in Japanese) (See Japanese page)
Topics (in English) Biometrics, etc.
Chair Kazuhiko Sumi(AGU)
Vice Chair Hiroshi Takano(Toyama Pref. Univ.) / Hitoshi Imaoka(NEC)
Secretary Hiroshi Takano(AIST) / Hitoshi Imaoka(Fujitsu Labs.)
Assistant Masatsugu Ichino(Univ. of Electro-Comm.) / Naoyuki Takada(Secom) / Norihiro Okui(KDDI Research)

Paper Information
Registration To Technical Committee on Biometrics
Language JPN
Title (in Japanese) (See Japanese page)
Sub Title (in Japanese) (See Japanese page)
Title (in English) Improvement of Accuracy by Machine Learning for Personal Authentication using High Frequency Intra-Body Propagation Characteristics
Sub Title (in English)
Keyword(1) Biometrics
Keyword(2) intra-body-propagation characteristics
Keyword(3) logistic regression
1st Author's Name Shun Onoda
1st Author's Affiliation Tokyo University of Science(TUS)
2nd Author's Name Takahiro Yoshida
2nd Author's Affiliation Tokyo University of Science(TUS)
3rd Author's Name Seiichiro Hangai
3rd Author's Affiliation Tokyo University of Science(TUS)
Date 2017-10-12
Paper # BioX2017-26
Volume (vol) vol.117
Number (no) BioX-236
Page pp.pp.7-10(BioX),
#Pages 4
Date of Issue 2017-10-05 (BioX)