In recent years, software-based face presentation attack detection (PAD) methods have seen a great progress. However, most existing schemes are not able to generalize well in more realistic conditions. The objective of this competition is to evaluate and compare the generalization performances of mobile face PAD techniques under some real-world variations, including unseen input sensors, presentation attack instruments (PAI) and illumination conditions, on a larger scale OULU-NPU dataset using its standard evaluation protocols and metrics. Thirteen teams from academic and industrial institutions across the world participated in this competition. This time typical liveness detection based on physiological signs of life was totally discarded. Instead, every submitted system relies practically on some sort of feature representation extracted from the face and/or background regions using hand-crafted, learned or hybrid descriptors. Interesting results and findings are presented and discussed in this paper.

Face recognition is one of the foremost applications in computer vision, which often involves sensitive signals; privacy concerns have been raised lately and tackled by several recent privacy-preserving face recognition approaches. Those systems either take advantage of information derived from the database templates or require several interaction rounds between client and server, so they cannot address outsourced scenarios. We present a private face verification system that can be executed in the server without interaction, working with encrypted feature vectors for both the templates and the probe face. We achieve this by combining two significant contributions: 1) a novel feature model for Gabor coefficients' magnitude driving a Lloyd-Max quantizer, used for reducing plaintext cardinality with no impact on performance; 2) an extension of a quasi-fully homomorphic encryption able to compute, without interaction, the soft scores of an SVM operating on quantized and encrypted parameters, features and templates. We evaluate the private verification system in terms of time and communication complexity, and in verification accuracy in widely known face databases (XM2VTS, FERET, and LFW). These contributions open the door to completely private and noninteractive outsourcing of face verification.

This paper presents a smart photo sharing application for mo bile devices based on an efficient yet powerful face recognition engine. The system allows registering subjects either from live video (mobile camera) or from still images, linking the obtained biometric information to the already existing phone contacts. The photo sharing application allows the user to select images which will be automatically processed with the built-in face recognition module based on Local Binary Patterns: already registered contacts will be identified, so that the current picture can be sent to them in a smart, user-friendly manner. The application also allows to tag unknown subjects and store (or update) their biometric profiles. All the processing is done in the mobile device. The current application runs on Android O.S., having been tested in two mobile platforms: HTC Desire and Samsung Galaxy Tab. The computational burden for each of the processing modules is provided in both devices, demonstrating the feasibility of the proposed approach for mobile platforms.

Gabor filters have demonstrated their effectiveness in automatic face recognition, which can greatly benefit from an accurate statistical model for Gabor-based face representations. Previous approaches have modeled real and imaginary parts independently as Generalized Gaussians (GG). Since most Gabor-based face recognition systems discard coefficients' phase, we propose a novel statistical model for the magnitude of Gabor coefficients that accounts for the dependence between real and imaginary parts, assuming they are circularly symmetric and marginally GG distributed. The quality of the fit for our model is assessed using the Kullback-Leibler divergence, and optimal quantization of Gabor coefficients is shown as one of its applications.

In this paper we present a fully automatic system for face recognition across pose where no frontal view is needed in enrollment or test. The system uses three Active Appearance Models(AAMs): the first one is a generic multi resolution AAM, while the remaining ones are trained to cope with left/right variations (i.e. pose-dependent AAMs). During the fitting stage, pose is automatically estimated using eigenvector analysis, and a synthetic face is generated through texture warping. Results over CMU PIE Database show promising results compared to the performance achieved with manually land marked faces.

Generalized Gaussian (GG) densities have been recently proposed to model both real and imaginary parts of Gabor coefficients. However, when matching faces, most systems make use of magnitude information only, due to its smooth behavior with displacements. The first goal of this paper is to propose a novel statistical model for the magnitude of Gabor coefficients, supposed that both real and imaginary parts are GG distributed. The proposed model, namely the ß-Rayleigh distribution, Rß(¿), is a generalization of the standard Rayleigh, R(¿), density. The Kullback Leibler (KL) divergence is used to measure the fitting accuracy of the model, showing the benefits of Rß(¿) over standard R(¿). The second goal of the paper tackles the selection of distance measures for Gabor features comparison, a topic that has received little attention in the literature. Inspired by the proposed statistical model, different ¿ßnorms are tested on the XM2VTS database, showing interesting results that confirm that classical distances used in Gabor-based recognition systems do not provide the best performance.

In this paper we present a solution to obtain useful and reliable user logs in a Learning Management System (LMS).Current LMS logs are combined with biometric-based logs that show the student behaviour. Our system models the student behaviour, allowing to know exactly how much time the student spends in front of the computer examining the con-tents of the LMS. Besides, user verification and face tracking are also integrated, what guarantees that the student is the person actually interacting with the system. The presented multimodal solution for user tracking and user ver-ification combines face tracking, face verification, speakerverification and fingerprint verification. Face tracking and face verification are performed in a non-collaborative fashion. Fingerprint or speaker verification is performed on demand, with the aim of avoiding a negative influence of adverse environmental or behavioural human factors in the reliability of the user logs generated by the system. These circumstances can thwart the non collaborative face veri-fication performance involved in the tracking process. The presented solution solves the problem of user tracking and authentication even in adverse environments for face verification.

It has been shown in many different contexts that the Generalized Gaussian (GG) distribution represents a flexible and suitable tool for data modeling. Almost all the reported applications are focused on modeling points (fixed length vectors); a different but crucial scenario, where the employment of the GG has received little attention, is the modeling of sequential data, i.e. variable length vectors. This paper explores this last direction, describing a variant of the well known Hidden Markov Model (HMM) where the emission probability function of each state is represented by a GG. A training strategy based on the Expectation Maximization (EM) algorithm is presented. Different experiments using both synthetic and real data (EEG signal classification and face recognition) show the suitability of the proposed approach compared with the standard Gaussian HMM.

This paper tackles the problem of pose variations in a 2D face recognition scenario. Using a training set of sparse face meshes, we built a point distribution model and identified the parameters which are responsible for controlling the apparent changes in shape due to turning and nodding the head, namely the pose parameters. Given a test image and its associated mesh, the pose parameters are set to typical values of frontal faces, thus obtaining a virtual frontal mesh. Taking advantage of facial symmetry, we overcome problems due to self-occlusion and virtual frontal faces are synthesized via thin plate splines-based texture mapping. These corrected images are then fed into a recognition system that makes use of Gabor filtering for feature extraction. The CMU PIE database is used to assess the performance of the proposed method in a closed-set identification scenario where large pose variations are present, achieving state-of-the-art results.

Gabor filters are biologically motivated convolution kernels that have been widely used in the field of computer vision and, specially, in face recognition during the last decade. This paper proposes a statistical model of Gabor coefficients extracted from face images using generalized Gaussian distributions (GGD's). By measuring the Kullback-Leibler distance (KLD) between the pdf of the GGD and the relative frequency of the coefficients, we conclude that GGD's provide an accurate modeling. The underlying statistics allow us to reduce the required amount of data to be stored (i.e. data compression) via Lloyd-Max quantization. Verification experiments on the XM2VTS database show that performance does not drop when, instead of the original data, we use quantized coefficients.

This paper proposes novel ways to deal with pose variations in a 2-D face recognition scenario. Using a training set of sparse face meshes, we built a point distribution model and identified the parameters which are responsible for controlling the apparent changes in shape due to turning and nodding the head, namely the pose parameters. Based on them, we propose two approaches for pose correction: 1) a method in which the pose parameters from both meshes are set to typical values of frontal faces, and 2) a method in which one mesh adopts the pose parameters of the other one. Finally, we obtain pose corrected meshes and, taking advantage of facial symmetry, virtual views are synthesized via Thin Plate Splines-based warping. Given that the corrected images are not embedded into a constant reference frame, holistic methods are not suitable for feature extraction. Instead, the virtual faces are fed into a system that makes use of Gabor filtering for recognition. Unlike other approaches that warp faces onto a mean shape, we show that if only pose parameters are modified, client specific information remains in the warped image and discrimination between subjects is more reliable. Statistical analysis of the authentication results obtained on the XM2VTS database confirm the hypothesis. Also, the CMU PIE database is used to assess the performance of the proposed methods in an identification scenario where large pose variations are present, achieving state-of-the-art results and outperforming both research and commercial techniques.

This paper proposes, through the combination of concepts and tools from different fields within the computer vision community, an alternative path to the selection of key points in face images. The classical way of attempting to solve the face recognition problem using algorithms which encode local information is to localize a predefined set of points in the image, extract features from the regions surrounding those locations, and choose a measure of similarity (or distance) between correspondent features. Our approach, namely shape-driven Gabor jets, aims at selecting an own set of points and features for a given client. After applying a ridges and valleys detector to a face image, characteristic lines are extracted and a set of points is automatically sampled from these lines where Gabor features (jets) are calculated. So each face is depicted by R2points and their respective jets. Once two sets of points from face images have been extracted, a shape-matching algorithm is used to solve the correspondence problem (i.e., map each point from the first image to a point within the second image) so that the system is able to compare shape-matched jets. As a byproduct of the matching process, geometrical measures are computed and compiled into the final dissimilarity function. Experiments on the AR face database confirm good performance of the method against expression and, mainly, lighting changes. Moreover, results on the XM2VTS and BANCA databases show that our algorithm achieves better performance than implementations of the elastic bunch graph matching approach and other related techniques.

In this paper, we use a robust shape matching algorithm for establishing correspondences between points from face images, and Gabor vectors (jets) for feature computation. After extracting characteristic lines from faces, which is accomplished by using the ridges and valleys operator, shape is sampled through a set of points, where jets are calculated. So, each face is depicted by /spl Rfr//sup 2/ points, and their respective jets. Once two sets of points from face images have been extracted, the shape matching algorithm is used to solve the correspondence problem. A measure between the set of matched points gives an idea of the similarity of the faces. Also, comparison between shape-matched jets is possible, given that they are supposed to be located over the same face parts, leading to a system that is able to take shape and texture into account for face authentication.