Search

Biometrics is widely used in access control systems, but it has a serious drawback in case of stealing or leakage, therefore, in the recent past, biometric template security has obtained increasing interest and many solutions have been investigated to secure biometric templates. Cancelable biometrics, which consists in the application of an intentional and repeatable modification to the original biometric template, is one the most promising approach which can satisfy the required properties of non-reversibility, accuracy, diversity and reusability. One of the most promising cancellable biometrics formulation is known as BioHashing and is based on the combination of the biometric template with user-specific tokenised random numbers (generated by an hash key) to produce a non-invertible hash code. The Hashed code is irreproducible without presenting both the biometric data and the hash key, it is irreversible and it obtains extremely good performance when the hash key is kept secret. Unfortunately, its performance strongly degrades when an impostor steals the hash key and tries to authenticate as a genuine user. In the literature, several methods have been recently proposed to solve this drawback, including an improved version of BioHashing and a method named BioPhasor. In this chapter we discuss the existing literature about various biometric template protection techniques and propose some solutions to answer to one of the main lacks of existing approaches: the loss of accuracy with respect to their analogous non-protected biometric systems. Our solutions are based on the use of ensembles of classifiers in an improved version of the BioPhasor hashing and in a method for the fusion between the BioPhasor and the Improved BioHashing approaches. Moreover we propose a cohort normalization method for both protected and unprotected matchers. Our results, obtained on different datasets and biometrics, show that the proposed solutions outperform the base BioPhasor method, in particular in the worst testing hypothesis when always an impostor steals all the personal information of a genuine user.

In this work an on-line signature authentication system based on an ensemble of local, regional, and global matchers is presented. Specifically, the following matching approaches are taken into account: the fusion of two local methods employing Dynamic Time Warping, a Hidden Markov Model based approach where each signature is described by means of its regional properties, and a Linear Programming Descriptor classifier trained by global features. Moreover, a template protection scheme employing the BioHashing and the BioConvolving approaches, two well known template protection techniques for biometric recognition, is discussed. The reported experimental results, evaluated on the public MCYT signature database, show that our best ensemble obtains an impressive Equal Error Rate of 3%, when only five genuine signatures are acquired for each user during enrollment. Moreover, when the proposed protected system is taken into account, the Equal Error Rate achieved in the worst case scenario, that is,when an "impostor" is able to steal the hash keys, is equal to 4.51%, whereas an Equal Error Rate ∼0 can be obtained when nobody steals the hash keys. © 2009 Elsevier Ltd. All rights reserved.

Verification based on tokenised pseudo-random numbers and user specific biometric feature has received much attention. In this paper, we propose a BioHashing system for automatic face recognition based on Fisher-based Feature Transform, a supervised transform for dimensionality reduction that has been proved to be very effective for the face recognition task. Since the dimension of the Fisher-based transformed space is bounded by the number of classes - 1, we use random subspace to create K feature spaces to be concatenated in a new higher dimensional space, in order to obtain a big and reliable ''BioHash code''.

In this work, we propose a multi-modal method that combines the scores of selected fingerprint matchers with the score obtained by a palm authenticator where the palm features are combined with pseudo-random numbers. We use a random subspace of AdaBoost.M1 to combine the scores of the best fingerprint matchers automatically selected among the competitors of FVC2004, and we study the performance when a different number of competitors are involved in the fusion. Moreover a deep study is carried out to design a hybrid system based on the combination of palm image features and a personal key. In conclusion, the aim of this work is to design a multi-modal biometric system and to analyze the benefits and the limits of fusion approaches in order to boost the performance of hybrid system in the worst testing hypothesis when an ''impostor'' steal the personal key of the user A before he tries to authenticate as A. The experimental results reported in this paper confirm that using a multi-modal ensemble of matchers it is possible to overcome some of the limitations of each single matcher leading to a considerable performance improvement.

Given the recent explosion of interest in human authentication, verification based on tokenized pseudo-random numbers and the user specific biometric feature (BioHashing) has received much attention. These methods have significant functional advantages over sole biometrics i.e. zero equal error rate. The main drawback of the base BioHashing method proposed in the literature relies in exhibiting low performance when an ''impostor''B steals the pseudo-random numbers of A and he tries to authenticate as A. In this paper, we introduce some ideas to improve the base BioHashing approach in order to maintain a very low equal error rate when nobody steals the Hash key, and to reach good performance also when an ''impostor'' steals the Hash key.