測距センサを使った低次元データによる非装着型の歩容認証

• Item type: SIG Technical Reports(1)
• 公開日: 2021-10-21
• タイトル: 測距センサを使った低次元データによる非装着型の歩容認証
• タイトル / 言語: en
• タイトル: Non-Wearable Gait Identification with Low-Dimensional Time Series Using Distance Sensor
• 言語: jpn
• キーワード / 主題Scheme: Other
• キーワード / 主題: 計測・認証
• 資源タイプ / 資源タイプ識別子: http://purl.org/coar/resource_type/c_18gh
• 資源タイプ: technical report
• 著者所属: 慶應義塾大学大学院理工学研究科
• 著者所属: 慶應義塾大学大学院理工学研究科
• 著者所属(英): en
• 著者所属(英): Graduate School of Science and Technology, Keio University
• 著者所属(英): en
• 著者所属(英): Graduate School of Science and Technology, Keio University
• 著者名: 宗形, 篤恭; 杉浦, 裕太
• 著者名(英): Atsuya, Munakata; Yuta, Sugiura
• 論文抄録 / 内容記述タイプ: Other
• 論文抄録 / 内容記述: Gait recognition is a biometric authentication based on the individuality of a person’s walking pattern. There are two main methods: one uses image sensors, and the other uses an accelerometer. The image sensor method is superior in that it is non-wearable and does not require attention or cooperation from the user. However, the information acquired by the image sensor is high dimensional and contains secondary information, which poses a problem in terms of privacy protection. In this study, we propose a system to identify individuals by extracting their walking patterns from one-dimensional time-series data acquired by Time-of-Flight (ToF) distance sensors. The system was evaluated by Leave-One-Out cross-validation, and the average identification accuracy was 92.0% for 5 users and 75.0% for 10 users.
• 論文抄録(英) / 内容記述タイプ: Other
• 論文抄録(英) / 内容記述: Gait recognition is a biometric authentication based on the individuality of a person’s walking pattern. There are two main methods: one uses image sensors, and the other uses an accelerometer. The image sensor method is superior in that it is non-wearable and does not require attention or cooperation from the user. However, the information acquired by the image sensor is high dimensional and contains secondary information, which poses a problem in terms of privacy protection. In this study, we propose a system to identify individuals by extracting their walking patterns from one-dimensional time-series data acquired by Time-of-Flight (ToF) distance sensors. The system was evaluated by Leave-One-Out cross-validation, and the average identification accuracy was 92.0% for 5 users and 75.0% for 10 users.
• 書誌レコードID / 収録物識別子タイプ: NCID
• 書誌レコードID / 収録物識別子: AA12049625
• 書誌情報: 研究報告エンタテインメントコンピューティング（EC）; 巻 2021-EC-61, 号 5, p. 1-4, 発行日 2021-10-21
• ISSN / 収録物識別子タイプ: ISSN
• ISSN / 収録物識別子: 2188-8914
• Notice: SIG Technical Reports are nonrefereed and hence may later appear in any journals, conferences, symposia, etc.
• 出版者 / 言語: ja
• 出版者: 情報処理学会