Biometric Identification Approach Research Articles

Deep learning-based photoplethysmography biometric authentication for continuous user verification

Biometric authentication methods have gained prominence as secure and convenient alternatives to traditional passwords and PINs. In this paper, we propose a novel approach for biometric authentication using photoplethysmography (PPG) signals and deep learning techniques. PPG is a non-invasive method that measures variations in blood volume within microvascular tissue beds, and it is typically used for monitoring heart rate and oxygen saturation. Our research leverages the unique characteristics of PPG signals to develop a robust and continuous user verification system. The primary goal of our study is to explore the feasibility and effectiveness of PPG-based biometric authentication, enabling a seamless and secure means of confirming the identity of individuals. We use a diverse dataset of PPG signals from various individuals, ensuring that it encompasses differences in skin tone, age, and other variables that can influence PPG signal characteristics. The collected data undergoes careful preprocessing, including noise removal, baseline correction, and heartbeat segmentation. For the core of our authentication system, we design and train a multiscale feature fusion deep learning (MFFD) model. This model, utilizing a Convolutional Neural Network (CNN) architecture, takes as input the relevant features extracted from PPG signals and learns to differentiate between individuals based on their unique PPG patterns. In this study, the input is constructed by gradually incorporating various features, beginning with a single PPG signal. In this study, the CNN model was trained independently, followed by the implementation of score fusion techniques. Our evaluation demonstrates the effectiveness of the PPG-based biometric authentication system, achieving high accuracy while addressing key security concerns. We consider false acceptance rate (FAR) and false rejection rate (FRR) to assess the system's performance. The model achieves the Accuracy of 99.5 % on BIDMC, 98.6 % on MIMIC, 99.2 % on CapnoBase dataset.

Integration of Face and Gait Recognition via Transfer Learning: A Multiscale Biometric Identification Approach

Integration of Face and Gait Recognition via Transfer Learning: A Multiscale Biometric Identification Approach

Integration of orthogonal Fourier profilometry with blood flow mapping for 3D fingertip biometric authentication

We present herein the combination of Fourier transform profilometry based orthogonal projection with a laser speckle imaging method to provide three-dimensional (3D) fingertip blood flow imaging and shape reconstruction for biometric authentication purposes. In the proposed approach, orthogonal sinusoidal grating patterns with a fixed spatial frequency emanating from a digital projector first illuminate the surface of a person's finger. Then, the projector is blocked and a laser source illuminates the finger. The projector and laser work alternately such that only one source illuminates the finger at a given time allowing no illumination crosstalk between the two devices. The deformed grating patterns and the diffuse backscattered laser light, respectively, are recorded on the same CCD camera. In the computer, the 3D shape of the finger is reconstructed simultaneously with a blood flow map of the finger. Finally, the images are merged yielding a dynamic flow topography visualization of the person's finger. This dual-modality strategy utilizes the strengths of each imaging method in a complementary way, thus enabling simple 3D shape recovery and characterization of a specific user, and increasing the specificity of authentication security. The method is experimentally validated on both human and fake fingers through VanderLugt optical correlation framework. Overall, the results demonstrate the utility of our proposed approach for biometric authentication. We envision that our optoelectronic setup could be potentially integrated into a wide variety security platform.

Multimodal Approach for Enhancing Biometric Authentication.

Unimodal biometric systems rely on a single source or unique individual biological trait for measurement and examination. Fingerprint-based biometric systems are the most common, but they are vulnerable to presentation attacks or spoofing when a fake fingerprint is presented to the sensor. To address this issue, we propose an enhanced biometric system based on a multimodal approach using two types of biological traits. We propose to combine fingerprint and Electrocardiogram (ECG) signals to mitigate spoofing attacks. Specifically, we design a multimodal deep learning architecture that accepts fingerprints and ECG as inputs and fuses the feature vectors using stacking and channel-wise approaches. The feature extraction backbone of the architecture is based on data-efficient transformers. The experimental results demonstrate the promising capabilities of the proposed approach in enhancing the robustness of the system to presentation attacks.

Transfer Learning Approaches for Gait-Based Wireless Person Identification

The human gait is a biometric feature that is unique to any individual. Since the human body affects the propagation of wireless signals, wireless networks can be used for identification of persons walking nearby. This biometric identification approach has been object of research in several studies described in the literature. In this work, we present a gait-based wireless user identification system that outperforms these existing solutions in terms of cost, effort and accuracy. Firstly, using a mesh network of low-cost wireless sensor nodes, the identification error rate could be reduced by more than 80in comparison to the best of prior studies. For a set of six persons to be identified, the identification accuracy could thus be improved to more than 99. Secondly, using the technique of Transfer Learning, the capabilities of the proposed system can be easily transferred to identify unknown persons and locations with only a few minutes of training effort. Training the system on the identification of six unknown persons, it has been found that the training effort can be reduced by about 96without a significant loss of accuracy.

Computational Intelligence Approach for Biometric Gait Identification

Gait recognition has gained significant attention in recent years due to its potential applications in various fields, including surveillance, security, and healthcare. Biometric gait identification, which involves recognizing individuals based on their walking patterns, is a challenging task due to the inherent variations in gait caused by factors such as clothing, footwear, and walking speed. In this paper, we propose a computational intelligence approach for biometric gait identification. Specifically, we integrate an intelligent convolutional model to identify human gaits based on the inertial sensory data captured from the body movement during the human walk. Extensive experiments on two datasets demonstrated that the efficiency of the proposed approach outperforms the existing methods. Our approach has the potential to be used in real-world applications such as surveillance systems and healthcare monitoring, where accurate and efficient identification of individuals based on their gait is crucial.

Ultrasonic Guided Wave Inversion Based on Deep Learning Restoration for Fingerprint Recognition.

As an established biometric authentication approach, fingerprint scanning has received considerable attention due to its high accuracy and reliability. In this article, the fingerprint reconstruction at any position is achieved in large physical domains, which monitors wavefield variations of plate-like structures within arrays through the ultrasonic guided wave. Accurate reconstruction and quantitative characterization of fingerprints are obtained using fast inversion tomography (FIT) based on the deep learning convolutional neural network (DLCNN). Parametric optimization is conducted to reveal submillimeter fingerprint minutiae, and a specific DLCNN model is proposed for the artifact removal in FIT reconstructions. The results prove that the FIT based on DLCNN restoration can significantly improve the imaging quality in terms of increased resolution, reduced reconstruction errors, and higher fingerprint matching confidence. The reconstruction also allows an exponential improvement in computational efficiency as a result of much-reduced sensor numbers. Several factors affecting the performance of the proposed reconstruction method are discussed at the end.

Multi-Layer Digital Validation of Candidate Service Appointment with Digital Signature and Bio-Metric Authentication Approach

Proposed work promotes a unique data security protocol for validating candidate’s service appointment. Process initiated with concealment of private share within the first segment of each region of the e-letter at commission’s server. This is governed by hash operations determining circular orientation of private share fragments and their hosted matrix intervals. Signed e-letter downloaded at the posted place is validated through same hash operations and public share. Candidate’s on spot taken fingerprint are concealed in two segments for each region of the eletter adopting similar hiding strategies. The copyright signature of posting place is similarly shielded on fourth segment of each region using hash operations. The certified e-letter is thoroughly validated at commission’s server and signatures stored justify authenticity of appointment and proper candidature at the posting place. The superior test results from wider angles establishes the efficacy of the proposed protocol over the existing approaches.

Imperceptible black-box waveform-level adversarial attack towards automatic speaker recognition

Automatic speaker recognition is an important biometric authentication approach with emerging applications. However, recent research has shown its vulnerability on adversarial attacks. In this paper, we propose a new type of adversarial examples by generating imperceptible adversarial samples for targeted attacks on black-box systems of automatic speaker recognition. Waveform samples are created directly by solving an optimization problem with waveform inputs and outputs, which is more realistic in real-life scenario. Inspired by auditory masking, a regularization term adapting to the energy of speech waveform is proposed for generating imperceptible adversarial perturbations. The optimization problems are subsequently solved by differential evolution algorithm in a black-box manner which does not require any knowledge on the inner configuration of the recognition systems. Experiments conducted on commonly used data sets, LibriSpeech and VoxCeleb, show that the proposed methods have successfully performed targeted attacks on state-of-the-art speaker recognition systems while being imperceptible to human listeners. Given the high SNR and PESQ scores of the yielded adversarial samples, the proposed methods deteriorate less on the quality of the original signals than several recently proposed methods, which justifies the imperceptibility of adversarial samples.

Novel Robust Photoplethysmogram-Based Authentication

As the tremendous advances in the biosensing technology, the biometric authentication approach has been adopted to replace the conventional password-based systems. Due to their harmless nature, photoplethysmogram (PPG) related applications have been quite intriguing in recent years. PPG can measure volumetric blood-flow changes in the peripheral circulation. Because PPG signals are very easy to acquire, compared to other biometrics (fingerprints, iris/retina, etc.), especially by low-cost wearable electronic devices, they are widely adopted to measure heart rates. In this paper, we devise a novel robust PPG based authentication system, which is capable of continuously authenticating the user instead of one-time authentication as carried out by the conventional techniques. This new system consists of preprocessing and filtering, motion artifact (MA) removal, template and feature extraction, and training. The associated multiwavelet-based feature extraction mechanism facilitates more reliable features than the conventional scalar-wavelet schemes so that it allows the learning model to better distinguish users. The associated autoencoder perfectly transforms the input PPG signal into the latent space. Finally, any distance measure can be adopted to classify or authenticate users. In order to build a sufficiently large dataset, we have combined three public datasets and our own dataset collected by ourselves locally together to create a new dataset consisting of 120 subjects totally. We have already built the hardware prototype successfully for real-time training data collection and real-time authentication. The equal-error-rate at the authentication stage reaches 5.5% and the identification accuracy reaches 98% at the identification stage.

A generic framework for deep incremental cancelable template generation

In a post-COVID-19 world, extensive study of deep learning-based biometric authentication techniques prompts the need to secure them. Further, the biometric data is assumed to be largely immutable; thus, if it is compromised, it is lost forever. Hence, reliable and secure biometric authentication is of utmost importance. In this paper, we address the security and privacy concerns of biometric templates generated via deep networks. We propose a cancelable biometric authentication approach. The framework consists of a lightweight Convolutional Neural Network (CNN) with a few-shot enrollment for generating biometric templates. Further, to enhance biometric templates’ discriminative power and to provide revocability, biometric templates are projected onto a random subspace (based on the user-specific key). Later projected biometric templates are mapped onto robust n-bit unique codes (using a KNN classifier) and protected via. SHA-3 hash digest. Moreover, a real-world biometric authentication system is always dynamic (users keep on changing). Thus we have also integrated phase-wise incremental learning within a deep learning-based cancelable biometric authentication framework. This is the first work in which deep cancelable templates are generated incrementally to the best of our knowledge. We analyze the proposed scheme for its performance and privacy preservation on three benchmarks constrained iris data-sets and over one unconstrained iris data-set along with one publicly available knuckle data-set. Furthermore, it has been demonstrated that the proposed cancelable incremental framework strictly follows the four fundamental properties of cancelability viz. non-invertibility, unlinkability, revocability, and usability.

Voice In Ear

With the rapid growth of wearable computing and increasing demand for mobile authentication scenarios, voiceprint-based authentication has become one of the prevalent technologies and has already presented tremendous potentials to the public. However, it is vulnerable to voice spoofing attacks (e.g., replay attacks and synthetic voice attacks). To address this threat, we propose a new biometric authentication approach, named EarPrint, which aims to extend voiceprint and build a hidden and secure user authentication scheme on earphones. EarPrint builds on the speaking-induced body sound transmission from the throat to the ear canal, i.e., different users will have different body sound conduction patterns on both sides of ears. As the first exploratory study, extensive experiments on 23 subjects show the EarPrint is robust against ambient noises and body motions. EarPrint achieves an Equal Error Rate (EER) of 3.64% with 75 seconds enrollment data. We also evaluate the resilience of EarPrint against replay attacks. A major contribution of EarPrint is that it leverages two-level uniqueness, including the body sound conduction from the throat to the ear canal and the body asymmetry between the left and the right ears, taking advantage of earphones' paring form-factor. Compared with other mobile and wearable biometric modalities, EarPrint is a low-cost, accurate, and secure authentication solution for earphone users.

An optimized rubber sheet model for normalization phase of IRIS recognition

Iris recognition is a promising biometric authentication approach and it is a very active topic in both research and realistic applications because the pattern of the human iris differs from person to person, even between twins. In this paper, an optimized iris normalization method for the conversion of segmented image into normalized form has been proposed. The existing methods are converting the Cartesian coordinates of the segmented image into polar coordinates. To get more accuracy, the proposed method is using an optimized rubber sheet model which converts the polar coordinates into spherical coordinates followed by localized histogram equalization. The experimental result shows the proposed method scores an encouraging performance with respect to accuracy.

An optimized rubber sheet model for normalization phase of iris recognition

Iris recognition is a promising biometric authentication approach and it is a very active topic in both research and realistic applications because the pattern of the human iris differs from person to person, even between twins. In this paper, an optimized iris normalization method for the conversion of segmented image into normalized form has been proposed. The existing methods are converting the Cartesian coordinates of the segmented image into polar coordinates. To get more accuracy, the proposed method is using an optimized rubber sheet model which converts the polar coordinates into spherical coordinates followed by localized histogram equalization. The experimental result shows the proposed method scores an encouraging performance with respect to accuracy.

P‐130: A Full Screen Biometric Identification Approach for OLED Displays by Using Near‐Infrared OLED

In this paper, we propose a full screen biometric identification approach for organic light emitting diode (OLED) displays using a near infra‐red(NIR) OLED coupled to a large area thin film transistor (TFT) pixel sensing circuit. Preliminary estimates for photons propagating from the NIR OLED source and reflecting back to the subpixel are presented. Although the signal landing on the subpixel is small, it may be amplified by the large area active pixel sensor circuits integrated in the TFT layer. The proposed approach has the key advantage of compatibility with existing large area fabrication processes. Although there are still some open technical issues such as NIR OLED lifetime, responsivity and sensor characteristics, the proposed approach demonstrates the potential to implement a full screen biometric identification within a large area display.

Securing electronic healthcare records: A mobile-based biometric authentication approach

In the Healthcare 4.0 era, recent breakthroughs in emerging and disruptive technologies, such as Artificial Intelligence (AI), machine learning and robotics, have significantly changed the way in which healthcare services are delivered to patients. Advancements in patient monitoring systems and wearable devices have created more powerful and flexible real-time remote patient observation. Healthcare provision has experienced remarkable improvements facilitated by cloud computing and wireless technologies, while wireless sensor networks now play an important role in mobile technologies for e-healthcare. In this paper, we propose a mobile-based healthcare system for the Healthcare 4.0 era. The proposed approach enables patients to self-authenticate, using their own mobile and wearable devices, establishing a session key between owned devices. After mutual authentication, the cloud server verifies each user. The security of the approach is described through the AVISPA tool. We analyze the embedded security and functionality which improves the healthcare system and calculate the relative communication and computation costs. Results show that the proposed approach provides greater security compared to other state-of-the-art schemes.

Investigation of Problems faced during capturing of Gait Signals

To solve the issue of close human contact in biometric authentication system, relatively a new technique gait recognition is used. The human gait is a common feature for identifying the walking manner of the person during walking and it is viewed as significant indicator for gait function of individual for experimental and research setting. Gait symmetry is usually considered as function of locomotion between the changes of the body and its activities. An exclusive advantage of gait as a biometric is its latent for detection at a distance or at low resolution or when other biometrics might not be perceivable. In this paper we investigate the problem of people recognition by their gait. The coordination and cyclic nature of the body motion makes gait as unique characteristics of each individual, thus a good biometric identification approach. The purpose of this paper is to describe some of the problems that make it difficult to apply gait as a biometric identification and use recent literature to show suggestions being made to solve some of these technical issues.

Investigation of problems faced during capturing of gait signals

To solve the issue of close human contact in biometric authentication system, relatively a new technique gait recognition is used. The human gait is a common feature for identifying the walking manner of the person during walking and it is viewed as significant indicator for gait function of individual for experimental and research setting. Gait symmetry is usually considered as function of locomotion between the changes of the body and its activities. An exclusive advantage of gait as a biometric is its latent for detection at a distance or at low resolution or when other biometrics might not be perceivable. In this paper we investigate the problem of people recognition by their gait. The coordination and cyclic nature of the body motion makes gait as unique characteristics of each individual, thus a good biometric identification approach. The purpose of this paper is to describe some of the problems that make it difficult to apply gait as a biometric identification and use recent literature to show suggestions being made to solve some of these technical issues.

Arabian horse identification based on whale optimised multi-class support vector machine

In this study, a biometric identification approach for Arabian horse identification is proposed based on the optimised Multi-Class Support Vector Machine (MCSVM). The identification approach is performed in three phases: feature extraction, classification, and optimisation. The feature extraction phase uses Histogram of Oriented Gradient (HOG) to extract features vectors from muzzle print images of the Arabian horses and then stored in the database with its labels. The second phase is the classification phase which uses MCSVM for training and testing classification. Finally in the optimised MCSVM phase, three different swarms, Particle Swarm Optimisation (PSO), Grey Wolf Algorithm (GWA) and Whale Optimisation (WO), are used to optimise MCSVM parameters to enhance the identification accuracy of the Arabian horse. The results obtained show that the MCSVM achieves accuracy of 93.2% and increases to 97.4% with WO algorithm which achieves the best accuracy compared to PSO and GWA.

Dimension Reduction of Hand and Face Feature Level Fusion in Multimodal Biometric Authentication

The proposed work is a multimodal biometric authentication approach with image texture feature dimension reduction of trained feature vector which leads reduction in memory size and in turn reduces the computational time. In this paper hand and face features are used for person identification. The texture features of hand image are extracted using Haar and several Daubechie’s of 2D-DWT followed by 2D- edge detector gives better identification with reduction in feature vector and face features are extracted by neighborhood common characterization with block based segmentation approach to estimate the disparity in face. The neighborhood common characterization based structure recognition with a person representative per sample is more effective. The neighborhood features are constructed by extracting the similar blocks in the image, the intra pixel disparity feature is obtained by exploiting external common images to estimate the feasible facial disparities. Neighborhood common characterization reduces the overall residual of the given features over the local feature, common disparity dictionary, and shape based residual of a block. Neighbourhood common characterization representation, of face recognize with one representative per person more effectively. The system uses either of the biometric traits for person identification with 99.98% of authentication rate.