Proposal of Cross-modal Biometrics

2014 Fiscal Year Final Research Report

• Project/Area Number: 24650082
• Research Category: Grant-in-Aid for Challenging Exploratory Research
• Allocation Type: Multi-year Fund
• Research Field: Perception information processing/Intelligent robotics
• Research Institution: Osaka University
• Principal Investigator: MAKIHARA Yasushi 大阪大学, 産業科学研究所, 准教授 (90403005)
• Project Period (FY): 2012-04-01 – 2015-03-31
• Keywords: 画像，文章，音声認識など / コンピュータビジョン / パターン認識 / 生体認証

Outline of Final Research Achievements

We formulated cross-modal discriminant analysis and confirmed the effectiveness of the proposed method compared with existing linear discriminant analysis and canonical correlation analysis through simulation experiments. Moreover, to confirm the effectiveness of the proposed method in real problems, we evaluate the performance in cross-modal biometric person authentication. More specifically, we focused on gait modality, which is the main biometric modality for the project integrator, and conducted experiments of cross-modal biometric person authentication against the large intra-class variations such as observation view, walking speed, clothing.

Free Research Field

コンピュータビジョン，パターン認識