Ana F. Sequeira

Biometric systems based on iris are vulnerable to several attacks, particularly direct attacks consisting on the presen-tation of a fake iris to the sensor. The development of iris liveness detection techniques is crucial for the deployment of iris biometric applications in daily life specially in the mobile biometric field. The 1st Mobile Iris Liveness Detec-tion Competition (MobILive) was organized in the context of IJCB2014 in order to record recent advances in iris live-ness detection. The goal for (MobILive) was to contribute to the state of the art of this particular subject. This com-petition covered the most common and simple spoofing at-tack in which printed images from an authorized user are presented to the sensor by a non-authorized user in order to obtain access. The benchmark dataset was the MobBIOfake database which is composed by a set of 800 iris images and its corresponding fake copies (obtained from printed images of the original ones captured with the same handhe...

Abstract—The new challenges concerning iris recognition are based on the need to support less constrained image acquisition environments. The proposed algorithm aims to achieve robust iris segmentation, based on combined information from iris center and iris contour, with such images. A 5.72 % mean error was obtained for the outer contour segmentation while the pupillary region presented a 9.36 % mean pixel misclassification. I.

The development of presentation attack detection (PAD) methods has become a high level concern in biometric security. As in other pattern recognition tasks, the use of deep learning is increasingly common. However, it is still doubtful if handcrafted features should be discarded. This work focused on the comparison of using handcrafted features and deep learning techniques at the feature extraction level in a face PAD method. Handcrafted features were based on Local Binary Patterns, while a Convolutional Neural Network based on VGG-16 was used for deep feature extraction. A Support Vector Machine was used for binary classification after dimensionality reduction using Principal Component Analysis. The methods were tested using the NUAA database, and the results show that handcrafted feature extraction still offer better results, with 3.1% APCER and 25.2% BPCER.

This work presents a novel multimodal database comprising 3D face, 2D face, thermal face, visible iris, finger and hand veins, voice and anthropometrics. This dataset will constitute a valuable resource to the field with its number and variety of biometric traits. Acquired in the context of the EU PROTECT project, the dataset allows several combinations of biometric traits and envisages applications such as border control. Based upon the results of the unimodal data, a fusion scheme was applied to ascertain the recognition potential of combining these biometric traits in a multimodal approach. Due to the variability on the discriminative power of the traits, a leave the n-best out fusion technique was applied to obtain different recognition results.

Biometrics represents a return to a traditional way of identifying someone relying on what that person is instead of what that person knows or owns. Even though the significant amount of research that has been done in this field, there is still much to do as new emerging scenarios of application appear everyday. Biometric recognition systems are no longer restricted to forensic investigation or control management of employees. They have been gaining a visibility and applicability in daily use devices which reinforces their usability in all aspects of our day to day life. With this spread of biometric applications, nowadays commonly found in our laptops, our smart phones, some bank management services and airport custom services, a necessity for improved security also is rising. The importance of protecting our identity and our data has become crucial as our devices are filled with sensible information of many kinds. Therefore the presentation attack or liveness detection methods as ...

Fingerprint presentation attack detection (PAD) methods present a stunning performance in current literature. However, the fingerprint PAD generalisation problem is still an open challenge requiring the development of methods able to cope with sophisticated and unseen attacks as our eventual intruders become more capable. This work addresses this problem by applying a regularisation technique based on an adversarial training and representation learning specifically designed to to improve the PAD generalisation capacity of the model to an unseen attack. In the adopted approach, the model jointly learns the representation and the classifier from the data, while explicitly imposing invariance in the high-level representations regarding the type of attacks for a robust PAD. The application of the adversarial training methodology is evaluated in two different scenarios: i) a handcrafted feature extraction method combined with a Multilayer Perceptron (MLP); and ii) an end-to-end solution ...

Despite the high performance of current presentation attack detection (PAD) methods, the robustness to unseen attacks is still an under addressed challenge. This work approaches the problem by enforcing the learning of the bona fide presentations while making the model less dependent on the presentation attack instrument species (PAIS). The proposed model comprises an encoder, mapping from input features to latent representations, and two classifiers operating on these underlying representations: (i) the task-classifier, for predicting the class labels (as bona fide or attack); and (ii) the species-classifier, for predicting the PAIS. In the learning stage, the encoder is trained to help the task-classifier while trying to fool the species-classifier. Plus, an additional training objective enforcing the similarity of the latent distributions of different species is added leading to a ‘PAI-species’-independent model. The experimental results demonstrated that the proposed regularisat...

The spread of biometric applications in mobile devices handled by untrained users opened the door to sources of noise in mobile iris recognition such as larger extent of rotation in the capture and more off-angle imagery not found so extensively in more constrained acquisition settings. As a result of the limitations of the methods in handling such large degrees of freedom there is often an increase in segmentation errors. In this work, a new near-infrared iris dataset captured with a mobile device is evaluated to analyse, in particular, the rotation observed in images and its impact on segmentation and biometric recognition accuracy. For this study a (manually annotated) ground truth segmentation was used which will be published in tandem with the paper. Similarly to most research challenges in biometrics and computer vision in general, deep learning techniques are proving to outperform classical methods in segmentation methods. The utilization of parameterized CNN-based iris segme...

In recent years many authors have recognized that the path forward, regarding iris recognition, is the development of iris recognition systems that can work independently of the conditions under which iris images are acquired. Recent works have tried to achieve robust and unconstrained iris recognition in order to develop real-world applicable methods. In this work, the problem of unconstrained iris recognition is referred and some results of an approach to the problem of fusing color information to enhance the performance of an iris authentication system are briefly presented.

In recent years, deep neural networks have shown significant progress in computer vision due to their large generalization capacity; however, the overfitting problem ubiquitously threatens the learning process of these highly nonlinear architectures. Dropout is a recent solution to mitigate overfitting that has witnessed significant success in various classification applications. Recently, many efforts have been made to improve the Standard dropout using an unsupervised merit-based semantic selection of neurons in the latent space. However, these studies do not consider the task-relevant information quality and quantity and the diversity of the latent kernels. To solve the challenge of dropping less informative neurons in deep learning, we propose an efficient end-to-end dropout algorithm that selects the most informative neurons with the highest correlation with the target output considering the sparsity in its selection procedure. First, to promote activation diversity, we devise ...

This work presents a novel dual-spectrum database containing both iris and periocular images synchronously captured from a distance and within a realistic indoor environment. This database was used in the 1st Cross-Spectrum Iris/Periocular Recognition Competition (Cross-Eyed 2016). This competition aimed at recording recent advances in cross- spectrum iris and periocular recognition. Six submissions were evaluated for cross-spectrum periocular recognition, and three for iris recognition. The submitted algorithms are briefly introduced. Detailed results are reported in this paper, and comparison of the results is discussed.

Fingerprint liveness detection methods have been developed as an attempt to overcome the vulnerability of fingerprint biometric systems to spoofing attacks. Traditional approaches have been quite optimistic about the behavior of the intruder assuming the use
of a previously known material. This assumption has led to the use of supervised techniques to estimate the performance of the methods, using both live and spoof samples to train
the predictive models and evaluate each type of fake samples individually. Additionally,
the background was often included in the sample representation, completely distorting the decision process. Therefore, we propose that an automatic segmentation step should be performed to isolate the fingerprint from the background and truly decide on the liveness of the fingerprint and not on the characteristics of the background. Also, we argue that
one cannot aim to model the fake samples completely since the material used by the intruder is unknown beforehand. We approach the design by modeling the distribution
of the live samples and predicting as fake the samples very unlikely according to that
model. Our experiments compare the performance of the supervised approaches with the semi-supervised ones that rely solely on the live samples. The results obtained differ
from the ones obtained by the more standard approaches which reinforces our conviction that the results in the literature are misleadingly estimating the true vulnerability of the
biometric system.

Biometric recognition is nowadays a mature technology with several applications.
However, biometric systems based on fingerprint are vulnerable
to direct attacks consisting on the presentation of a fake fingerprint to
the sensor. This work focuses on fingerprint liveness detection methods
as an attempt to overcome that vulnerability. Two methods from the stateof-
the-art in iris liveness detection were tested with fingerprint databases
containing different kinds of fake samples. One aim of the work was to
investigate how these iris techniques would perform with fingerprint fake
samples. The other purpose was to diversify the classification scenario
by broaden the classification task from being made within each type of
samples to being made in sets mixing the types of fake samples.

Biometric systems based on iris are vulnerable to several attacks, particularly direct attacks consisting on the presentation of a fake iris to the sensor. The development of iris liveness detection techniques is crucial for the deployment of iris biometric applications in daily life specially in the mobile biometric field. The 1st Mobile Iris Liveness Detection Competition (MobILive) was organized in the context of IJCB2014 in order to record recent advances in iris liveness detection. The goal for (MobILive) was to contribute to the state of the art of this particular subject. This competition covered the most common and simple spoofing attack in which printed images from an authorized user are presented to the sensor by a non-authorized user in order to obtain access. The benchmark dataset was the MobBIOfake database which is composed by a set of 800 iris images and its corresponding fake copies (obtained from printed images of the original ones captured with the same handheld device and in similar conditions). In this paper we present a brief description of the methods and the results achieved by the six participants in the competition.

Biometric systems based on iris are vulnerable to direct attacks consisting on the presentation of a fake iris to the sensor (a printed or a contact lenses iris image, among others). The mobile biometrics scenario stresses the importance of assessing the security issues. The application of countermeasures against this type of attacking scheme is the problem addressed in the present paper. Widening a previous work, several stateof-
the-art iris liveness detection methods were implemented and
adapted to a less-constrained scenario. The proposed method
combines a feature selection step prior to the use of state-of-theart classifiers to perform the classification based upon the “best features”. Five well known existing databases for iris liveness purposes (Biosec, Clarkson, NotreDame and Warsaw) and a recently published database, MobBIOfake, with real and fake images captured in the mobile scenario were tested. The results obtained suggest that the automated segmentation step does not degrade significantly the results.

Biometrics represents a return to a natural way of identification: testing someone by what (s)he is, instead of relying on something (s)he owns or knows seems likely to be the way forward. Biometric systems that include multiple sources of information are known as multimodal. Such systems are generally regarded as an
alternative to fight a variety of problems all unimodal systems stumble upon. One of the main challenges found in the development of biometric recognition systems is the shortage of publicly available databases acquired under real unconstrained working conditions. Motivated by such need the MobBIO database was created using an Asus EeePad Transformer tablet, with mobile biometric systems in mind. The proposed database is composed by three modalities: iris, face and voice.

The rising challenges in the field of iris recognition, concerning the development of accurate recognition
algorithms using images acquired under an unconstrained set of conditions, is leading to the a renewed interest
in the area. Although several works already report excellent recognition rates, these values are obtained by
acquiring images in very controlled environments. The use of such systems in daily security activities, such
as airport security and bank account management, is therefore hindered by the inherent unconstrained nature
under which images are to be acquired. The proposed work focused on mutual context information from iris
centre and iris limbic contour to perform robust and accurate iris segmentation in noisy images. A random
subset of the UBIRIS.v2 database was tested with a promising E1 classification rate of 0.0109.

The known attacks to Bifid cipher are based on the fact that the period of that cipher
is constant. The Medusa cipher is based on the Bifid sharing some of its characteristics.
However, the Medusa was built so that the length of the blocks of the message is variable,
and dependent of the original message, with the purpose of avoiding the aplication of the
mentioned attacks.
An implementation of the Medusa has been developed in C++. The method of construction
of the key requires dealing with large integers, bigger than what is allowed by actual CPUs.
Therefore, in order to implement the Medusa cipher, was necessary to use a way to represent
large (non negatives) integers, and to perform operations on them, as addition, subtraction,
multiplication and division of integers. With this long precision arithmetic it is possible to
obtain the key of the Medusa from an integer value obtained by the Diffie-Hellman protocol,
and to cipher and decipher messages.

Biometrics represents a return to a traditional way of identifying someone relying on what that person is instead of what that person knows or owns. Even though the significant amount of research that has been done in this field, there is still much to do as new emerging scenarios of application appear everyday. Biometric recognition systems are no longer restricted to forensic investigation or control management of employees. They have been gaining a visibility and applicability in daily use devices which reinforces their usability in all aspects of our day to day life. With this spread of biometric applications, nowadays commonly found in our laptops, our smart phones, some bank management services and airport custom services, a necessity for improved security also is rising. The importance of protecting our identity and our data has become crucial as our devices are filled with sensible information of many kinds. Therefore the presentation attack or liveness detection methods as countermeasures against spoofing attacks are more important than ever. New methods should be developed which address the new acquisition scenarios and which deal with the increased noise in the biometric data collected. Its of utmost importance to develop robust liveness detection methods. In particular, we worked on iris and fingerprint. These two biometric traits are very often chosen against others due to its characteristics. Among the objectives of this thesis were the purpose of making contributions in iris and fingerprint liveness detection proposing novel approaches whether from the imaging scenarios perspective, in the case of iris, or from the classification approach, in the case of fingerprint. Contributions were made regarding both traits, that exceeded the state-of-the art and resulted in both conferences and journal publications. Not only the spoofing attacks concern the biometric researchers but also the ability of the methods to deal with the noisy data. Therefore, the development of robust methods that overcome the compromised quality of data is a necessity of biometric research of nowadays. Therefore, another objective was to contribute to the fingerprint recognition problem developing robust methods to minutiae extraction. The work developed resulted in a proposed method for fingerprint orientation map estimation and a fingerprint image enhancement that over performed existing ones. This work aimed and succeeded to propose robust and realistic methods in both the iris and fingerprint liveness detection problem as well as in some steps of fingerprint recognition. It has to be noted that the focus of attention of this work was the quality of data and not the computational efficiency, therefore this one should have to be addressed if an application of the proposed methods to a real-world scenario was aimed. Another objective was to create new databases and promote common platforms of evaluation of methods such as biometric competitions. Therefore, along the work developed, two biometric databases were constructed and two biometric competitions were organized. Both databases had a strong impact in the research community and they continue to be disseminated. Publications using these benchmark datasets are numerous and continue to appear regularly.