Protecting Biometric Features by Periodic Function-Based Transformation and Fuzzy Vault - Transactions on Large-Scale Data-and Knowledge-Centered Systems - page 57

Protecting Biometric Features by Periodic
Function-Based Transformation and Fuzzy Vault

Thu Thi Bao Le (  ) , Tran Khanh Dang, Quynh Chi Truong, and Thi Ai Thao Nguyen

Faculty of Computer Science and Engineering, HCMC University of Technology,
VNUHCM, Ho Chi Minh City, Vietnam
{thule,khanh,tqchi,thaonguyen}@cse.hcmut.edu.vn

Abstract. Biometrics-based authentication is playing an attractive and potential
approach nowadays. However, the end-users do not feel comfortable to use it
once the performance and security are not ensured. Fuzzy vault is one of the
most popular methods for biometric template security. It binds a key with the
biometric template and obtains the helper data. However, the main problem of
fuzzy vault is that it is unable to guarantee the revocability property. In addi-
tion, most of the fuzzy vault schemes are performed on two biometrics
modalities, fingerprints and iris. In previous works, authors suggested some
cancelable transformations attached to a fuzzy vault scheme to overcome these
weaknesses. However, the computational cost of these proposals was quite
large. In this paper, we present a new hybrid scheme of fuzzy vault and periodic
function-based feature transformation for biometric template protection. Our
transformation is not only simpler but also suitable for many kinds of biomet-
rics modalities. The newly proposed fuzzy vault scheme guarantees the revoca-
bility property with an acceptable error rate.

Keywords: Biometrics · Fuzzy vault · Biometrics template protection · Cancel-
able transformations · Face recognition

1 Introduction

As we all know, the traditional authentication schemes usually based on something
the user has (such as: smart card) or something the user knows (such as: password,
PIN). However, those techniques have several limitations. For example, they cannot
distinguish between an authorized user and those who know the correct password [1].
So, we have to choose a strong password and always to keep it in mind.
In recent years, with the rapid development of technologies, biometrics-based au-
thentication systems are becoming potential, because biometrics is literally stuck to an
individual, it can prevent the use of several identities by a single individual. The term
biometric (from the Greek for bio=life, metric=degree) refers to authentication by
means of biological (more accurately, physiological or behavioral) features (such as:
face, voice, fingerprint…) [2]. Using biometrics can overcome above limitations. But,
it still raises some security and privacy concerns [1]. For example, biometrics is se-
cure but not secret, because voice, face… can be easily recorded and may be misused