Face Biometric Systems Research Articles

Antispoofing in face biometrics: a comprehensive study on software-based techniques

The vulnerability of the face recognition system to spoofing attacks has piqued the biometric community's interest, motivating them to develop anti-spoofing techniques to secure it. Photo, video, or mask attacks can compromise face biometric systems (types of presentation attacks). Spoofing attacks are detected using liveness detection techniques, which determine whether the facial image presented at a biometric system is a live face or a fake version of it. We discuss the classification of face anti-spoofing techniques in this paper. Anti-spoofing techniques are divided into two categories: hardware and software methods. Hardware-based techniques are summarized briefly. A comprehensive study on software-based countermeasures for presentation attacks is discussed, which are further divided into static and dynamic methods. We cited a few publicly available presentation attack datasets and calculated a few metrics to demonstrate the value of anti-spoofing techniques.

Antispoofing in face biometrics: a comprehensive study on software-based techniques

The vulnerability of the face recognition system to spoofing attacks has piqued the biometric community's interest, motivating them to develop antispoofing techniques to secure it. Photo, video, or mask attacks can compromise face biometric systems (types of presentation attacks). Spoofing attacks are detected using liveness detection techniques, which determine whether the facial image presented at a biometric system is a live face or a fake version of it. We discuss the classification of face anti-spoofing techniques in this paper. Anti-spoofing techniques are divided into two categories: hardware and software methods. Hardware-based techniques are summarized briefly. A comprehensive study on software-based countermeasures for presentation attacks is discussed, which are further divided into static and dynamic methods. We cited a few publicly available presentation attack datasets and calculated a few metrics to demonstrate the value of anti-spoofing techniques.

IoT-Based Biometric Recognition Systems in Education for Identity Verification Services: Quality Assessment Approach

Traditional identity verification of students based on the human proctoring approach can cause a scam identity verification and ineffective processing time, particularly among vast groups of students. Most student identification cards are outdated personal information. Several biometric recognition approaches have been proposed and can be adopted to strengthen students’ identity verification. Most educational adoption technology struggles with evaluation and validation techniques to ensure that biometric recognition systems are unquestionably suitable for utilization and implementation to fulfill student identity verification. This study presents the internet of things to develop flexible biometric recognition systems and an approach to assess the quality of biometric systems for educational use by investigating the effectiveness of identity verification of various biometric recognition technologies compared to the traditional verification method. The unimodal, multimodal, and semi-multimodal biometric technologies were tested using the developed internet of things-base biometric recognition systems examined by applying the proposed quality metrics of scoring factors based on accuracy, error rate, processing time, and cost. Hundreds of undergraduate exam takers were a sample group. Key findings indicate that the designed and presented systems suitably attain identity verification of exam students using a unimodal biometric. A unimodal face biometric system promises excellent support. A unimodal fingerprint biometric system assures the second excellent aid for student identity verification. However, multimodal and semi-multimodal biometric systems provide better accuracy with few handling times and higher costs. This study contributes significantly to the knowledge of utilizing biometric recognition for identity verification in smart educational use.

DOG-ADTCP: A new feature descriptor for protection of face identification system

Nowadays the face biometric system is gaining popularity for authentication in access control applications. At the same time, the threats to authentication systems are also increased in terms of spoofing or presentation attacks where an intruder attempts to spoof the face recognition system by using genuine user photos or video to gain access. In order to effectively secure the secrecy of a genuine user, there is an urgent need for building a face authentication system with anti-spoofing countermeasures. In this paper, we introduce a novel face anti-spoofing approach, which is mainly based on contrast and texture features of both real and spoof photos. A modified Difference of Gaussian (DOG) filtering method and Angle Directional Ternary Co-relation pattern (ADTCP) are used. The publicly available NUAA, MSU-MFSD and REPLAY-ATTACK photo-impostor datasets are tested on the approach, which includes images with different illumination and area of the face. The accuracy of the proposed approach is evaluated using different metrics. The results show that our proposed method is superior to other state-of-the-art (SOTA) practices when tested on three photo-impostor datasets.

Self-supervised 2D face presentation attack detection via temporal sequence sampling

Conventional 2D face biometric systems are vulnerable to presentation attacks performed with different face artefacts, e.g., printouts, video-replays and wearable 3D masks. The research focus in face presentation attack detection (PAD) has been recently shifting towards end-to-end learning of deep representations directly from annotated data rather than designing hand-crafted (low-level) features. However, even the state-of-the-art deep learning based face PAD models have shown unsatisfying generalization performance when facing unknown attacks or acquisition conditions due to lack of representative training and tuning data available in the existing public benchmarks. To alleviate this issue, we propose a video pre-processing technique called Temporal Sequence Sampling (TSS) for 2D face PAD by removing the estimated inter-frame 2D affine motion in the view and encoding the appearance and dynamics of the resulting smoothed video sequence into a single RGB image. Furthermore, we leverage the features of a Convolutional Neural Network (CNN) by introducing a self-supervised representation learning scheme, where the labels are automatically generated by the TSS method as the stabilized frames accumulated over video clips of different temporal lengths provide the supervision. The learnt feature representations are then fine-tuned for the downstream task using labelled face PAD data. Our extensive experiments on four public benchmarks, namely Replay-Attack, MSU-MFSD, CASIA-FASD and OULU-NPU, demonstrate that the proposed framework provides promising generalization capability and encourage further study in this domain.

Beating the bias in AI-based biometrics

OneSpan research scientist Ismini Psychoula outlines the many types of bias that can exist in face and other biometric systems – and how to counter them.

Face spoofing detection using surface and sub-surface reflections analysis

Reflection based analysis has been used in previous research for various objectives. Materials classification is one of them. Basically, each material consists of two types of reflections: surface and sub-surface. To separate these two reflections, polarized light could be applied. Previously, multi-reflections characteristics were analyzed using polarized light to classify objects such as between metals and non-metals. However, no trial has been done using the same method to distinguish real and fake faces that could be used to combat spoofing attempts in face biometric system. Since human skin is multi layers structure, it also produces multi reflections. In this paper, driven by the theory, surface and sub-surface reflections of both genuine human face and paper face mask were statistically examined. In addition, iPad displayed face images were also used as spoofing attempts. Images of genuine and spoofing faces were captured using polarized light under two different polarization angles: 0 and 90 degrees. Each angle captured images with surface and sub-surface reflections, accordingly. Those reflections were analyzed based on the mean, standard deviation, skewness and kurtosis. Modality distribution of each image was also studied using another method called the bimodality coefficient (BC). From the results, it is not possible to distinguish between genuine face and printed photos because of the multi reflections’ similarities. However, iPad displayed face images have been successfully identified as spoofing trials.

An optimal feature enriched region of interest (ROI) extraction for periocular biometric system.

With the onset of COVID-19 pandemic, wearing of face mask became essential and the face occlusion created by the masks deteriorated the performance of the face biometric systems. In this situation, the use of periocular region (region around the eye) as a biometric trait for authentication is gaining attention since it is the most visible region when masks are used. One important issue in periocular biometrics is the identification of an optimal size periocular ROI which contains enough features for authentication. The state of the art ROI extraction algorithms use fixed size rectangular ROI calculated based on some reference points like center of the iris or centre of the eye without considering the shape of the periocular region of an individual. This paper proposes a novel approach to extract optimum size periocular ROIs of two different shapes (polygon and rectangular) by using five reference points (inner and outer canthus points, two end points and the midpoint of eyebrow) in order to accommodate the complete shape of the periocular region of an individual. The performance analysis on UBIPr database using CNN models validated the fact that both the proposed ROIs contain enough information to identify a person wearing face mask.

Performance Evaluation of Face Recognition system by Concatenation of Spatial and Transformation Domain Features

Face biometric system is one of the successful applications of image processing. Person recognition using face is the challenging task since it involves identifying the 3D object from 2D object. The feature extraction plays a very important role in face recognition. Extraction of features both in spatial as well as frequency domain has more advantages than the features obtained from single domain alone. The proposed work achieves spatial domain feature extraction using Asymmetric Region Local Binary Pattern (ARLBP) and frequency domain feature extraction using Fast Discrete Curvelet Transform (FDCT). The obtained features are fused by concatenation and compared with trained set of features using different distance metrics and Support Vector Machine (SVM) classifier. The experiment is conducted for different face databases. It is shown that the proposed work yields 95.48% accuracy for FERET, 92.18% for L-space k, 76.55% for JAFFE and 81.44% for NIR database using SVM classifier. The results show that the proposed system provides better recognition rate for SVM classifier when compare to the other distance matrices. Further, the work is also compared with existing work for performance evaluation.

Matrix-Regularized One-Class Multiple Kernel Learning for Unseen Face Presentation Attack Detection

The functionality of face biometric systems is severely challenged by presentation attacks (PA’s), and especially those attacks that have not been available during the training phase of a PA detection (PAD) subsystem. Among other alternatives, the one-class classification (OCC) paradigm is an applicable strategy that has been observed to provide good generalisation against unseen attacks. Following an OCC approach for the unseen face PAD from RGB images, this work advocates a matrix-regularised multiple kernel learning algorithm to make use of several sources of information each constituting a different view of the face PAD problem. In particular, drawing on the one-class null Fisher classification principle, we characterise different deep CNN representations as kernels and propose a multiple kernel learning (MKL) algorithm subject to an ( $r,p$ )-norm ( $1\leq r,p$ ) matrix regularisation constraint. The propose MKL algorithm is formulated as a saddle point Lagrangian optimisation task for which we present an effective optimisation algorithm with guaranteed convergence. An evaluation of the proposed one-class MKL algorithm on both general object images in an OCC setting as well as on different face PAD datasets in an unseen zero-shot attack detection setting illustrates the merits of the proposed method compared to other one-class multiple kernel and deep end-to-end CNN-based methods.

The Effects of Facial Expressions on Face Biometric System’s Reliability

The human mood has a temporary effect on the face shape due to the movement of its muscles. Happiness, sadness, fear, anger, and other emotional conditions may affect the face biometric system’s reliability. Most of the current studies on facial expressions are concerned about the accuracy of classifying the subjects based on their expressions. This study investigated the effect of facial expressions on the reliability of a face biometric system to find out which facial expression puts the biometric system at greater risk. Moreover, it identified a set of facial features that have the lowest facial deformation caused by facial expressions to be generalized during the recognition process, regardless of which facial expression is presented. In order to achieve the goal of this study, an analysis of 22 facial features between the normal face and six universal facial expressions is obtained. The results show that the face biometric systems are affected by facial expressions where the disgust expression achieved the most dissimilar score, while the sad expression achieved the lowest dissimilar score. Additionally, the study identified the five and top ten facial features that have the lowest facial deformations on the face shape in all facial expressions. Besides that, the relativity score showed less variances between the sample using the top facial features. The obtained results of this study minimized the false rejection rate in the face biometric system and subsequently the ability to raise the system’s acceptance threshold to maximize the intrusion detection rate without affecting the user convenience.

A FACE SPOOFING DETECTION SYSTEM USING SURF ANALYSIS BASED ON GENETIC ALGORITHM AND ARTIFICIAL INTELLIGENCE TECHNIQUE.

Modern face biometric systems are susceptible to spoofing attacks and a secure face spoof detection system demands the capability to recognize whether a face is from a real person or a spoofed image that is created by an unauthenticated person. Inspired by the feature selection algorithm, characterization of printing artifacts, and differences in light reflection, we proposed to approach the problem of spoofing detection from a pattern analysis point of view. Indeed, face prints often contain printing quality faults that can be well detected using pattern features, the Speech up Robust Feature (SURF) descriptor. Hence, introduces a novel approach based on face pattern image analysis to find out if there is live in front of a camera or a printed face. The proposed approach analyzes the pattern and quality of the facial images using the SURF descriptor as a feature extraction algorithm. Compared to a lot of previous works, our proposed face spoofing detection approach is robust, computationally fast, and does not require user-cooperation. In addition, the feature optimization technique is used for the selection of a unique feature set from the ROI of face images. Convolutional Neural Network (CNN) classifier is used for the training of the proposed spoof detection system. It is seen that the designed hybrid system face spoof detection achieves high performance than the existing system and execution time is also well. The proposed method is assessed using the MATLAB simulator in computer vision and image processing toolbox. The experimental analysis on a publicly accessible database presented brilliant results compared to existing works by using the concept of feature optimization and artificial intelligence technique.

Penerapan Estimasi Posisi dan Tracking Wajah Pada Sistem Presensi Mahasiswa

The current presence system can be done with a computerized system, one of which is the face biometric system. This study focuses on the application of position estimation and tracking based on clustering on people's faces to determine the position in three dimensions. Position estimation can be obtained by making a kernel that is ready to be used to predict three-dimensional coordinates of faces based on two-dimensional coordinates of two images. Position estimation can be done by utilizing the Machine Learning algorithm family. In this study, Least Absolute Shrinkage and Selection Operators (LASSO) is used to perform the position estimation. Meanwhile, clustering in this study uses the K-Means algorithm. Based on the test results, the kernel error obtained in estimating the face location is 9.23 cm. The tracking accuracy of an object based on clustering is 100%.

Verifying the Effectiveness of New Face Spoofing DB with Capture Angle and Distance

Face recognition is a representative biometric that can be easily used; however, spoofing attacks threaten the security of face biometric systems by generating fake faces. Thus, it is not advisable to only consider sophisticated spoofing cases, such as three-dimensional masks, because they require additional equipment, thereby increasing the implementation cost. To prevent easy face spoofing attacks through print and display, the two-dimensional (2D) image analysis method using existing face recognition systems is reasonable. Therefore, we proposed a new database called the “pattern recognition-face spoofing advancement database” that can be used to prevent such attacks based on 2D image analysis. To the best of our knowledge, this is the first face spoofing database that considers the changes in both the angle and distance. Therefore, it can be used to train various positional relationships between a face and camera. We conducted various experiments to verify the efficiency of this database. The spoofing detection accuracy of our database using ResNet-18 was found to be 96.75%. The experimental results for various scenarios demonstrated that the spoof detection performances were better for images with pinch angle, near distance images, and replay attacks than those for front images, far distance images, and print attacks, respectively. In the cross-database verification result, the performance when tested with other databases (DBs) after training with our DB was better than the opposite. The results of cross-device verification in terms of camera type showed negligible difference; thus, it was concluded that the type of image sensor does not affect the detection accuracy. Consequently, it was confirmed that the proposed DB that considers various distances, capture angles, lighting conditions, and backgrounds can be used as a training DB to detect spoofing attacks in general face recognition systems.

Anti-spoofing in face recognition-based biometric authentication using Image Quality Assessment

Despite the rapid growth of face recognition-based biometrics for both authentication and identification, the security of face biometric systems against presentation attacks (also called spoofing attacks) remains a great concern. Indeed, Face recognition-based authentication techniques can be easily spoofed using various types of attacks such photographs, videos or forged 3D masks. This work proposes a fast and non-intrusive anti-spoofing solution based on Image Quality Assessment (IQA) and motion cues to distinguish between genuine and fake face-appearances. Quality measures are computed following a novel approach which enables us to highlight these liveness-related motion cues, thus outlining the distinction between real faces and spoofing attacks. Moreover, our method is well suited for real-time mobile applications as it takes into consideration both reliable robustness and low complexity of employed algorithms. Our approach is extensively evaluated on three public databases that include different types of presentation attacks. The obtained results proved to outperform state-of-the-art approaches.

Bio-Inspired Presentation Attack Detection for Face Biometrics.

Today, face biometric systems are becoming widely accepted as a standard method for identity authentication in many security settings. For example, their deployment in automated border control gates plays a crucial role in accurate document authentication and reduced traveler flow rates in congested border zones. The proliferation of such systems is further spurred by the advent of portable devices. On the one hand, modern smartphone and tablet cameras have in-built user authentication applications while on the other hand, their displays are being consistently exploited for face spoofing. Similar to biometric systems of other physiological biometric identifiers, face biometric systems have their own unique set of potential vulnerabilities. In this work, these vulnerabilities (presentation attacks) are being explored via a biologically-inspired presentation attack detection model which is termed “BIOPAD.” Our model employs Gabor features in a feedforward hierarchical structure of layers that progressively process and train from visual information of people's faces, along with their presentation attacks, in the visible and near-infrared spectral regions. BIOPAD's performance is directly compared with other popular biologically-inspired layered models such as the “Hierarchical Model And X” (HMAX) that applies similar handcrafted features, and Convolutional Neural Networks (CNN) that discover low-level features through stochastic descent training. BIOPAD shows superior performance to both HMAX and CNN in all of the three presentation attack databases examined and these results were consistent in two different classifiers (Support Vector Machine and k-nearest neighbor). In certain cases, our findings have shown that BIOPAD can produce authentication rates with 99% accuracy. Finally, we further introduce a new presentation attack database with visible and near-infrared information for direct comparisons. Overall, BIOPAD's operation, which is to fuse information from different spectral bands at both feature and score levels for the purpose of face presentation attack detection, has never been attempted before with a biologically-inspired algorithm. Obtained detection rates are promising and confirm that near-infrared visual information significantly assists in overcoming presentation attacks.

Demographic Effects in Facial Recognition and Their Dependence on Image Acquisition: An Evaluation of Eleven Commercial Systems

We examined the effect of demographic factors on the performance of 11 commercial face biometric systems tested as part of the 2018 U.S. Department of Homeland Security, Science and Technology Directorate biometric technology rally. Each system that participated in this evaluation was tasked with acquiring face images from a diverse population of 363 subjects in a controlled environment. Biometric performance was assessed using a systematic, repeatable test process measuring both efficiency (transaction times), and accuracy (mated similarity scores) using a leading commercial algorithm. Prior works have documented differences in biometric algorithm performance across demographic categories and proposed that skin phenotypes offer a superior explanation for these differences. To test this concept, we developed an automatic method for measuring relative facial skin reflectance using subjects’ enrollment images and quantified the effect of this metric and other demographic covariates on performance using linear modeling. Both the efficiency and accuracy of the tested acquisition systems were significantly affected by multiple demographic covariates, including skin reflectance, gender, age, eyewear, and height. Skin reflectance had the strongest net linear effect on performance. Linear modeling showed that lower (darker) skin reflectance was associated with lower efficiency (higher transaction times) and accuracy (lower mated similarity scores). Skin reflectance was also a statistically better predictor of these effects than self-identified race labels. Unlike other covariates, the degree to which skin reflectance altered accuracy varied between systems. We show that the size of this skin reflectance effect was inversely related to the overall accuracy of the system such that the effect was almost negligible for the system with the highest overall accuracy. These results suggest that, in evaluations of biometric accuracy, the magnitude of measured demographic effects depends on image acquisition.

REAL TIME FACE LIVENESS DETECTION WITH IMAGE QUALITY AND TEXTURE PARAMETER

Face Biometric systems are vulnerable to spoofing attacks. The activeness of face recognition has raised concerns about face spoof attacks, where a printed photo or video of an authorized person face could be used to gain access to facilities or service. Hence, we present a novel approach based on analyzing facial image quality and face image texture for detecting whether there is a live person or a fake person. The proposed approach analyzes the spoof face detection based on video and printout image. Two different feature is extracted (chromatic moment and Local Binary Pattern (LBP)) from the face image. Extracted feature is fed into Classifier, SVM classifier trained for different face spoof attack (i.e. printed photo and played video) is used to distinguish between real and fake face.

A survey of liveness detection methods for face biometric systems

PurposeBiometric systems are widely used for face recognition. They have rapidly developed in recent years. Compared with other approaches, such as fingerprint recognition, handwriting verification and retinal and iris scanning, face recognition is more straightforward, user friendly and extensively used. The aforementioned approaches, including face recognition, are vulnerable to malicious attacks by impostors; in such cases, face liveness detection comes in handy to ensure both accuracy and robustness. Liveness is an important feature that reflects physiological signs and differentiates artificial from real biometric traits. This paper aims to provide a simple path for the future development of more robust and accurate liveness detection approaches.Design/methodology/approachThis paper discusses about introduction to the face biometric system, liveness detection in face recognition system and comparisons between the different discussed works of existing measures.Originality/valueThis paper presents an overview, comparison and discussion of proposed face liveness detection methods to provide a reference for the future development of more robust and accurate liveness detection approaches.

Face Anti-Spoofing using Speeded-Up Robust Features and Fisher Vector Encoding

The vulnerabilities of face biometric authentication systems to spoofing attacks have received a significant attention during the recent years. Some of the proposed countermeasures have achieved impressive results when evaluated on intratests, i.e., the system is trained and tested on the same database. Unfortunately, most of these techniques fail to generalize well to unseen attacks, e.g., when the system is trained on one database and then evaluated on another database. This is a major concern in biometric antispoofing research that is mostly overlooked. In this letter, we propose a novel solution based on describing the facial appearance by applying Fisher vector encoding on speeded-up robust features extracted from different color spaces. The evaluation of our countermeasure on three challenging benchmark face-spoofing databases, namely the CASIA face antispoofing database, the replay-attack database, and MSU mobile face spoof database, showed excellent and stable performance across all the three datasets. Most importantly, in interdatabase tests, our proposed approach outperforms the state of the art and yields very promising generalization capabilities, even when only limited training data are used.