가변 얼굴 영상의 합성 방법에 관한 연구

Cancelable biometrics has been introduced for secure usage of biometric data in biometric recognition system. Cancelable template is a revocably and non-invertibly transformed version of original biometric data. Normally, the face biometric data can be interpreted by machine and human. However, due to non-invertible transform, cancelable template cannot be understood by human. In this paper, we propose a method about human interpretable face image synthesis. Because, face image is synthesized using revocable cancelable template, synthesized face image is also revocable. In other words, we can simply regenerate face image with newly issued cancelable template.