2022 Fiscal Year Final Research Report
A Novel Electromagnetic Liveness Detection System against High-Level Fingerprint Spoof Attacks using Machine Learning for Fingerprint Authentication Systems
| Project/Area Number |
20H04189
|
|---|
| Research Category |
Grant-in-Aid for Scientific Research (B)
|
| Allocation Type | Single-year Grants |
|---|
| Section | 一般 |
|---|
| Review Section |
Basic Section 60070:Information security-related
|
|---|
| Research Institution | Ritsumeikan University |
|---|
Principal Investigator |
Maeda Tadahiko 立命館大学, 情報理工学部, 教授 (40351324)
|
|---|
| Project Period (FY) |
2020-04-01 – 2023-03-31
|
| Keywords | 生体認証 / 指紋認証 / 生体検知 / CSRR / 機械学習 / ニューラスネットワーク / Autoencoder / バイオメトリックス |
|---|
| Outline of Final Research Achievements |
In recent years, spoofing attacks against fingerprint authentication technologies caused by fake fingerprints have been reported and raised severe social security issues. This report describes systematic countermeasures for fingerprint authentication systems. A multi-ring liveness detection sensor was proposed to improve the detection accuracy, and measurements were carried out for human and fake fingers to evaluate the effectiveness of the proposed sensor. Also, a modified C-type sensor and a coplanar-type sensor were designed and fabricated for the measurements. Besides, we have proposed a novel two-step judgment method to improve detection accuracy and assessed the accuracy to demonstrate the effectiveness of the proposed algorithm. We have also developed and evaluated unique person-specific templates generated with the sensor’s scattering responses for the frequency range from 9.5 GHz through 14 GHz using a neural network to enhance the overall detection accuracy for the system.
|
| Free Research Field |
ワイヤレスシステム
|
| Academic Significance and Societal Importance of the Research Achievements |
人体指紋の撮像画像を判定基準として使用する指紋認証方式は，多要素認証を構成する重要要素技術の一つであり，情報社会のセキュリティを支える技術として期待されている．一方，日常生活においては素手で活動することが一般的であり，外部に露出している人体指紋情報を判定情報源とする判定方式は，本人が意識しない状況下で指紋撮像や残留指紋から指紋が複製される本質的な脆弱性が存在している．本研究では外部から遠隔では取得することが出来ない生体電磁応答特性を指紋認証に適用することに注目し，独自の生体検知センサ構造に機械学習アルゴリズムを組み合わせることで，高度偽装物に対抗するための指紋認証高度化要素技術を検討した．
|
