Dimension reduction by local principal component analysis
Neural Computation
Pattern Recognition with Fuzzy Objective Function Algorithms
Pattern Recognition with Fuzzy Objective Function Algorithms
Enhancing security and privacy in biometrics-based authentication systems
IBM Systems Journal - End-to-end security
Cancelable Biometric Filters for Face Recognition
ICPR '04 Proceedings of the Pattern Recognition, 17th International Conference on (ICPR'04) Volume 3 - Volume 03
Robust Distance Measures for Face-Recognition Supporting Revocable Biometric Tokens.
FGR '06 Proceedings of the 7th International Conference on Automatic Face and Gesture Recognition
| Hi-index | 0.00 |
In order to resolve non-revocability of biometrics, changeable biometrics has been recently introduced. Changeable biometrics transforms an original biometric template into a changeable template in a non-invertible manner. The transformed changeable template does not reveal the original biometric template, so that secure concealment of biometric data is possible using a changeable transformation. Changeable biometrics has been applied to face recognition, and there are several changeable face recognition methods. However, previous changeable face transformations cannot provide face images that can be recognized by humans. Hence, 'human inspection', a particular property of face biometrics, cannot be provided by previous changeable face biometric methods. In this paper, we propose a face image synthesis method which allows human inspection of changeable face templates. The proposed face synthesis method is based on subspace modeling of face space, and the face space is obtained by projection method such as principle component analysis (PCA) or independent component analysis (ICA). The face space is modeled by fuzzy C-means clustering and partition-based PCA. Using the proposed method, human-recognizable faces can be synthesized while providing inter-class discrimination and intra-class similarity.