various applications, such as daily activity monitoring and sports performance analysis. However, when employed for behavior verification, it requires not
only high recognition performance but also robustness against intentional manipulation.
To address this challenge, we propose a dual authenticat...
various applications, such as daily activity monitoring and sports performance analysis. However, when employed for behavior verification, it requires not
only high recognition performance but also robustness against intentional manipulation.
To address this challenge, we propose a dual authentication framework that simultaneously verifies the wear location of the sensor and authenticates the user. Specifically, a
transformer-based model was constructed for wear location authentication, and a dedicated binary classification model was developed for user authentication. Experimental
evaluations demonstrated that the wear location authentication model achieved F1 scores
of 0.90 or higher across all classes, while the wearer authentication model, validated
through k-fold cross-validation on data from 16 users, achieved a low equal error rate of
6.1%. These results highlight the effectiveness of the proposed method and its potential
to provide a tamper-resistant foundation for behavioral authentication systems.
•生体情報パターン認証 (Cheungら
[1]
,Vhaduriら
[2][3]
)
•心拍数をベースにした手法
•心拍数だけで認証できないときは歩行データ、呼吸音も使用
•課題
•ウェアラブルデバイスに対して多数の計測機能を要求
•歩行データ認証 (Leeら
[4]
,Liuら
[5]
)
•歩行動作から個人を認証
•Liuらの研究では環境変化に対する頑健性を示した.
•e.g. 路面の種類,様々なセンサの装着部位や向き
•課題
•所定の位置に装着していることが前提
5
[1]William Cheung and Sudip Vhaduri: ContextDependentImplicit Authentication for Wearable Device Users, Proceedings of the 2020 IEEE 31st Annual International Symposium on Personal, Indoor and Mobile Radio Communications,
pp. 1–7, IEEE, (2020).
[2]Sudip Vhaduriand Christian Poellabauer: MultiModalBiometric-Based Implicit Authentication of Wearable Device Users, IEEE Transactions on Information Forensics and Security, Vol.14, No. 12, pp. 3116–3125, IEEE, (2019).
[3]Sudip Vhaduri, SayantonDibboVhaduriand William Cheung: HIAuth: A Hierarchical Implicit Authentication System for IoT Wearables Using Multiple Biometrics, IEEE Access, Vol. 9, pp. 116395–116406, IEEE, (2021).
[4]Soobin Lee, SeungjaeLee, EunkyoungPark, JongshillLee and In Young Kim: Gait-Based Continuous Authentication Using a Novel Sensor Compensation Algorithm and Geometric Features Extracted From Wearable Sensors, IEEE
Access, Vol. 10, pp. 120122–120135, IEEE, (2022).
[5]Yushi Liu, Kamen Ivanov, JunxianWang, Fuhai Xiong, JiahongWang, Min Wang, Zedong Nie, Lei Wang and Yan Yan: Topological Data Analysis for Robust Gait Biometrics Based on Wearable Sensors, IEEE Transactions on
Consumer Electronics, Vol. 70, No. 2, pp. 4910–4921, IEEE, (2024).
関連研究(装着者認証 )
•歩行中の加速度データ (Kunzeら
[6]
)
•歩行中かどうか検出
•手首,頭部, 左大腿部のポケット, 左胸ポケットの4クラス識別
•課題
•歩行中以外の識別に未対応であること
•識別クラスが少ないこと
•カメラ映像+ IMU(Ruizら
[7]
,Bannisら
[8]
)
•カメラ映像から装着者の姿勢推定
•ネットワーク上のデバイスの仮想 IDと装着者と装着部位を識別
•課題
•カメラに映る空間内のみ使用可能
•プライバシーの観点からの抵抗感
6
[6]Klaus Kunze, Patrick Lukowicz, Harald Junker, Gerhard Tr¨oster: Where am I: Recognizing On-body Positions of Wearable Sensors, In: Thomas Strang, Christian Linnhoff-Popien(eds) Location-and Context-Awareness. LoCA2005.
Lecture Notes in Computer Science, Vol. 3479, Springer, Berlin, Heidelberg, pp. 214–229, (2005).
[7]Carlos Ruiz, ShijiaPan, Hae Young Noh, Pei Zhang: WhereWear: Calibration-free Wearable Device Identification through Ambient Sensing, Proc. The 5th ACM Workshop on Wearable Systems and Applications (WearSys’19), pp.
29–34, (2019).
[8]Adeola Bannis, ShijiaPan, Carlos Ruiz, John Shen, Hae Young Noh, Pei Zhang: IDIoT: Multimodal Framework for Ubiquitous Identification and Assignment of Human-carried Wearable Devices, ACM Trans. Internet Things, Vol. 4, No.
2, Art. 11, 25 pages (2023).
関連研究(装着部位認証 )
•心拍・脈拍の時間差 (Yoshidaら
[9]
)
•心電図センサと脈拍センサの同時計測
•波形の時間差を用いて装着部位を推定
•特定の動作を強要する必要がないという優位性
•課題
•心電図センサ,脈拍センサの装着が前提
•日常的な使用には不向き
7
[9]Kazuki Yoshida, Kazuya Murao: Estimating load positions of wearable devices based on difference in pulse wave arrival time, Proc. ACM Int. Symp. Wearable Computers (ISWC ’19), pp. 234–243, 10 pages, London, United Kingdom
(2019).
関連研究(装着部位認証 )