The face and gesture recognition network (FG-NET) ageing database was released in 2004 in an attempt to support research activities aimed at understanding the changes in facial appearance caused by ageing. Since then the database was used for carrying out research in various disciplines including age estimation, age-invariant face recognition and age progression. On the basis of the analysis of published work where the FG-NET ageing database was used, conclusions related to the type of research carried out in relation to the impact of the dataset in shaping up the research topic of facial ageing are presented. This study also includes a review of key articles from different thematic areas, where the FG-NET ageing database was used and the presentation of benchmark results. The ultimate aims of this study are to present concrete facts related to research activities in facial ageing during the past decade, provide an indication of the main methodologies adopted, present a comprehensive list of benchmark results and most importantly provide roadmaps for future trends, requirements and research directions in facial ageing.

Finger image quality assessment is a crucial part of any system where a high biometric performance and user satisfaction is desired. Several algorithms measuring selected aspects of finger image quality have been proposed in the literature, yet only few of them have found their way into quality assessment algorithms used in practice. The authors provide comprehensive algorithm descriptions and make available implementations of adaptations of ten quality assessment algorithms from the literature which operates at the local or the global image level. They evaluate the performance on four datasets in terms of the capability in determining samples causing false non-matches and by their Spearman correlation with sample utility. The authors’ evaluation shows that both the capability in rejecting samples causing false non-matches and the correlation between features varies depending on the dataset.

The authors generated 316,250 entire distributions of IrisCode impostor scores, each distribution obtained by comparing one iris against hundreds of thousands of others in a database including persons spanning 152 nationalities. Altogether 100 billion iris comparisons were performed in this study. The purpose was to evaluate whether, in the tradition of Doddington's Zoo, some individuals are inherently more prone than most to generate iris false matches, while others are inherently less prone. With the standard score normalisation disabled, a detailed inter-quantile analysis showed that meaningful deviations from a universal impostors distribution occur only for individual distributions that are highly extreme in both their mean and their standard deviation, and which appear to make up <1% of the population. In general, when different persons are compared, the IrisCode produces relatively constant dissimilarity distances having an invariant narrow distribution, thanks to the large entropy which lies at the heart of this biometric modality. The authors discuss the implications of these findings and their caveats for various search strategies, including ‘1-to-first’ and ‘1-to-many’ iris matching.

To overcome the limitation of traditional illumination invariant methods for single sample face recognition, a modified version of gradientface named adaptively weighted orthogonal gradient binary pattern (AWOGBP), which is proved robust to illumination variation, is proposed in this study. First, the Tetrolet transform is performed on the images to obtain low frequency and high frequency components and the retina model processing is applied to low frequency component to make the image more robust to illumination, in the meantime, the authors multiply each element in high frequency components with a scale factor to accentuate details. Then the proposed AWOGBP is used to get the feature vectors of each direction and all the feature vectors are concatenated into the general feature vector for face recognition with the weights of the sub-graph based on their information entropy which is defined as the contribution to describe the whole face images. Finally the principle component analysis method is used to reduce dimensions and the nearest neighbour classifier is used for face image classification and recognition. Experimental results on CMU PIE and Extended Yale B face databases indicate that the proposed method is significantly better as compared with related state-of-the-art methods.

The vulnerability of biometric systems to external attacks using a physical artefact in order to impersonate the legitimate user has become a major concern over the last decade. Such a threat, commonly known as ‘spoofing’, poses a serious risk to the integrity of biometric systems. The usual low-complexity and low-cost characteristics of these attacks make them accessible to the general public, rendering each user a potential intruder. The present study addresses the spoofing issue analysing the feasibility to perform low-cost attacks with self-manufactured three-dimensional (3D) printed models to 2.5D and 3D face recognition systems. A new database with 2D, 2.5D and 3D real and fake data from 26 subjects was acquired for the experiments. Results showed the high vulnerability of the three tested systems, including a commercial solution, to the attacks.

Iris recognition systems typically enrol a person based on a single ‘best’ eye image. Research has shown that the probability of a false non-match result increases with increased difference in pupil dilation between the enrolment image and the probe image. Therefore, dilation-aware methods of enrolment should improve the accuracy of iris recognition. The authors examine a strategy to improve accuracy through a dilation-aware enrolment step that selects one or more enrolment images based on the observed distribution of dilation ratios for that eye. Additionally, they demonstrate that an image with median dilation is the optimal single eye image dilation-aware enrolment choice. Their results confirm that this dilation-aware enrolment strategy does improve matching accuracy compared with traditional single-image enrolment, and also compared with multi-image enrolment that does not take dilation into account.

From the intuition that natural face images lie on or near a low-dimensional submanifold, the authors propose a novel spectral graph based dimensionality reduction method, named orthogonal enhanced linear discriminant analysis (OELDA), for face recognition. OELDA is based on enhanced LDA (ELDA), which takes into account both the discriminative structure and geometrical structure of the face space, and generates non-orthogonal basis vectors. However, a significant fact is that eliminating the dependence of basis vectors can promote more effective recognition of unseen face images. For this purpose, the authors seek to improve the ELDA scheme by imposing orthogonal constraints on the basis vectors. Experimental results on real-world face datasets show that, benefitting from orthogonality, OELDA has more locality preserving power and discriminative power than LDA and ELDA, and achieves the highest recognition rates among compared methods.

The popularity of face recognition systems have increased due to their use in widespread applications. Driven by the enormous number of potential application domains, several algorithms have been proposed for face recognition. Face pose and image resolutions are among the two important factors that influence the performance of face recognition algorithms. In this study, the authors present a comparative study of three baseline face recognition algorithms to analyse the effects of two aforementioned factors. The algorithms studied include (a) the adaptive boosting (AdaBoost) with linear discriminant analysis as weak learner, (b) the principal component analysis (PCA)-based approach, and (c) the local binary pattern (LBP)-based approach. They perform an empirical study using the images with systematic pose variation and resolution from multi-pose, illumination, and expression database to explore the recognition accuracy. This evaluation is useful for practical applications because most engineers start development of a face recognition application using these baseline algorithms. Simulation results revealed that the PCA is more accurate in classifying the pose variation, whereas the AdaBoost is more robust in identifying low-resolution images. The LBP does not classify face images of size 20 × 20 pixels and below and has lower recognition accuracy than PCA and AdaBoost.

Verifying an identity claim by fingerprint recognition is a commonplace experience for millions of people in their daily life, for example, for unlocking a tablet computer or smartphone. The first processing step after fingerprint image acquisition is segmentation, that is, dividing a fingerprint image into a foreground region which contains the relevant features for the comparison algorithm, and a background region. The authors propose a novel segmentation method by global three-part decomposition (G3PD). On the basis of global variational analysis, the G3PD method decomposes a fingerprint image into cartoon, texture and noise parts. After decomposition, the foreground region is obtained from the non-zero coefficients in the texture image using morphological processing. The segmentation performance of the G3PD method is compared with five state-of-the-art methods on a benchmark which comprises manually marked ground truth segmentation for 10,560 images. Performance evaluations show that the G3PD method consistently outperforms existing methods in terms of segmentation accuracy.

In this study, the authors propose a novel fingerprint template protection scheme that is developed using Delaunay triangulation net constructed from the fingerprint minutiae. The authors propose two methods namely FS_INCIR and FS_AVGLO to construct a feature set from the Delaunay triangles. The feature set computed is quantised and mapped to a 3D array to produce fixed length 1D bit string. This bit string is applied with a DFT to generate a complex vector. Finally, the complex vector is multiplied by user's key to generate a cancellable template. The proposed computation of feature set maintained a good balance between security and performance. These methods are tested on FVC 2002 and FVC 2004 databases and the experimental results show satisfactory performance. Further, the authors analysed the four requirements namely diversity, revocability, irreversibility and accuracy for protecting biometric templates. Thus, the feasibility of proposed scheme is depicted.

The authors explore the possibilities of a dense model-free three-dimensional (3D) face reconstruction method, based on image sequences from a single camera, to improve the current state of forensic face comparison. They propose a new model-free 3D reconstruction method for faces, based on the Lambertian reflectance model to estimate the albedo and to refine the 3D shape of the face. This method avoids any form of bias towards face models and is therefore suitable in a forensic face comparison process. The proposed method can reconstruct frontal albedo images, from multiple non-frontal images. Also a dense 3D shape model of the face is reconstructed, which can be used to generate faces under pose. In the authors’ experiments, the proposed method is able to improve the face recognition scores in more than 90% of the cases. Using the likelihood ratio framework, they show for the same experiment that for data initially unsuitable for forensic use, the reconstructions become meaningful in a forensic context in more than 60% of the cases.

A combination of cryptographic and biometric systems, by performing specific binding technique on cryptographic key and biometric template, the fuzzy vault framework enhances the security level of current biometric cryptographic systems in terms of hiding secret key and protecting the template. Although the original scheme suggests the use of error-correction techniques (e.g. the Reed–Solomon code) to reconstruct the original polynomial, recent implementations do not share the same point of view. Instead, cyclic redundant code (CRC) is applied to identify the genuine polynomial from a set of candidates due to its simplicity. Within the scope of this study, the authors address a significant flaw of current CRC-based fuzzy vault schemes, which allows the potential of successful blend substitution attack. To overcome this problem, an integration of two novel modules into general fuzzy vault scheme, namely chaff point generator and verifier, is proposed. The new modules are designed to be integrated easily into the existing systems as well as simple to enhance. The proposed scheme can detect any modification in vault and, as a result, eliminate the blend substitution attack to improve general security. Moreover, the experimental results of this study with real-world datasets show an increase in genuine acceptance rates.