A keyless multimodal-based user authentication scheme using generative adversarial networks

Biometrics provide enhanced security and convenience compared to conventional methods of individual authentication. A more robust and effective method of individual authentication has emerged due to recent advancements in multimodal biometrics. Unimodal systems offer lower security and lack the robustness found in multimodal biometric systems. The research paper introduces a novel approach, employing multiple biometric modalities, including face, fingerprint, and iris, to authenticate users in a multimodal biometric system. The paper proposes the ”Secure Sense” framework, which combines multiple biometric modalities to improve person verification accuracy. The proposed system utilizes both web-based and real-time datasets. For the web-based dataset, we employed the Chicago Face dataset for facial data, the MMU1 dataset for iris data, and the SOCOfing dataset for fingerprint data. In real-time data collection, facial data is captured using a Zebronics Zeb-Gem webcam, fingerprint data is obtained using the Mantra MFS scanner, and iris data is collected using the Mantra MIS scanner. In the envisioned system, we introduce an innovative approach that employs a decision-level fusion technique across three unique biometric modalities, resulting in an impressive accuracy rate of approximately 93% across all modalities.

Biometrics identification using multiple modalities has attracted the attention of many researchers as it produces more robust and trustworthy results than single modality biometrics. In this paper, we present a novel multimodal recognition system that trains a deep learning network to automatically learn features after extracting multiple biometric modalities from a single data source, i.e., facial video clips. Utilizing different modalities, i.e., left ear, left profile face, frontal face, right profile face, and right ear, present in the facial video clips, we train supervised denoising auto-encoders to automatically extract robust and non-redundant features. The automatically learned features are then used to train modality specific sparse classifiers to perform the multimodal recognition. Moreover, the proposed technique has proven robust when some of the above modalities were missing during the testing. The proposed system has three main components that are responsible for detection, which consists of modality specific detectors to automatically detect images of different modalities present in facial video clips; feature selection, which uses supervised denoising sparse auto-encoders network to capture discriminative representations that are robust to the illumination and pose variations; and classification, which consists of a set of modality specific sparse representation classifiers for unimodal recognition, followed by score level fusion of the recognition results of the available modalities. Experiments conducted on the constrained facial video dataset (WVU) and the unconstrained facial video dataset (HONDA/UCSD), resulted in a 99.17% and 97.14% Rank-1 recognition rates, respectively. The multimodal recognition accuracy demonstrates the superiority and robustness of the proposed approach irrespective of the illumination, nonplanar movement, and pose variations present in the video clips even in the situation of missing modalities.

A cancelable biometric identification scheme based on bloom filter and format-preserving encryption

Smartphone market leaders jostle for biometric technology firsts

Cross-modal hashing maps heterogeneous multimedia data into a common Hamming space to realize fast and flexible cross-modal retrieval. Supervised cross-modal hashing methods have achieved considerable progress by incorporating semantic side information. However, they heavily rely on large-scale labeled cross-modal training data which are hard to obtain, since multiple modalities are involved. They also ignore the rich information contained in the large amount of unlabeled data across different modalities, which can help to model the correlations between different modalities. To address these problems, in this paper, we propose a novel semi-supervised cross-modal hashing approach by generative adversarial network (SCH-GAN). The main contributions can be summarized as follows: 1) we propose a novel generative adversarial network for cross-modal hashing, in which the generative model tries to select margin examples of one modality from unlabeled data when given a query of another modality (e.g., giving a text query to retrieve images and vice versa). The discriminative model tries to distinguish the selected examples and true positive examples of the query. These two models play a minimax game so that the generative model can promote the hashing performance of the discriminative model and 2) we propose a reinforcement learning-based algorithm to drive the training of proposed SCH-GAN. The generative model takes the correlation score predicted by discriminative model as a reward, and tries to select the examples close to the margin to promote a discriminative model. Extensive experiments verify the effectiveness of our proposed approach, compared with nine state-of-the-art methods on three widely used datasets.

Biometric recognition is more and more employed in authentication and access control of various applications. Biometric data are strongly linked with the user and do not allow revocability nor diversity, without an adapted post-processing. Cancelable biometrics, including the very popular algorithm BioHashing, is used to cope with the underlying privacy and security issues. The principle is to transform a biometric template in a BioCode, in order to enhance user privacy and application security. These schemes are used for template protection of several biometric modalities, as fingerprints or face and the robustness is generally related to the hardness to recover the original biometric template by an impostor. In this paper, we propose to use genetic algorithms to approximate the original biometric feature and spoof the authentication system. We show through experimental results on fingerprints the efficiency of the proposed attack on the BioHashing algorithm, by approximating the original FingerCode, given the seed and the corresponding BioCode.

"This paper explores the application of neural networks in multimodal biometric authentication systems, emphasizing the integration of multiple biometric modalities to enhance security, accuracy, and robustness against various attacks. Traditional authentication methods, such as passwords and single-modal biometrics, often suffer from vulnerabilities, including spoofing, environmental factors, and data breaches. To address these challenges, multimodal authentication systems combine several biometric traits, such as facial recognition, fingerprint scanning, voice recognition, and keystroke dynamics, to achieve higher reliability and resistance to security threats. The study provides an overview of public datasets used for training neural network-based biometric authentication models, including VoxCeleb, RAVDESS, MOBIO, and SDUMLA-HMT. These datasets contain diverse biometric information necessary for developing robust multimodal authentication systems. The paper evaluates the effectiveness of existing approaches using key performance metrics such as accuracy, false acceptance rate (FAR), false rejection rate (FRR), and area under the curve (AUC). Additionally, specialized metrics are considered, including failure to enroll rate (FTE), failure to acquire rate (FTA), and template stability (TS), which are crucial for real-world applications. The role of neural networks in multimodal biometric authentication is analyzed by examining state-of-the-art architectures, including convolutional neural networks (CNNs) and deep learning-based feature fusion methods. Various fusion levels—feature-level, score-level, and decision-level—are discussed to determine the optimal integration strategy for improving authentication performance. The results indicate that multimodal systems significantly outperform unimodal authentication methods by reducing vulnerability to spoofing and environmental variations. Experimental findings suggest that integrating multiple biometric traits enhances the system’s adaptability to dynamic conditions, reducing both false acceptance and false rejection rates. Despite these advantages, several challenges remain, including computational complexity, data privacy concerns, and the need for real-time processing capabilities. Future research should focus on optimizing multimodal fusion techniques, improving generalization across different datasets, and enhancing the security of biometric templates against adversarial attacks. Additionally, developing lightweight neural network architectures suitable for mobile and embedded systems is essential for the practical deployment of multimodal authentication technologies"

Template-Protection-Techniken stellen eine wichtige Ergänzung zu biometrischen Systemen, um das gewünschte Sicherheitsniveau zu erreichen und die Anforderungen an den Schutz der Privatsphäre zu erfüllen. Diese Techniken ermöglichen Erneuerung und Widerruf der Referenzendaten in biometrischen Systemen. In dieser Arbeit wird ein systematisches Evaluierungsframework entwickelt, das sich an die wesentlichen Kriterien und Anforderungen an biometrische Template-Protection-Verfahren festhält. Seine Anwendbarkeit wird durch die Analyse der verschiedenen Algorithmen demonstriert. Die in dieser Arbeit präsentierte Evaluierung ist grundlegend für eine vollständige Analyse. Darüber hinaus ermöglicht es einen Nachweis der Sicherheit und Einhaltung der Privatsphäre. Es schafft eine Basis für Zertifizierung und Benchmark der Template-Protection-Verfahren.

In this digital world, numerous credit card-based transactions take place all over the world. Concomitantly, gaps in process flows and technology result in many fraudulent transactions. Owing to the spurt in the number of reported fraudulent transactions, customers and credit card service providers incur significant financial and reputation losses respectively. Therefore, building a powerful fraud detection system is paramount. It is noteworthy that fraud detection datasets, by nature, are highly unbalanced. Consequently, almost all of the supervised classifiers, when built on the unbalanced datasets, yield high false negative rates. But, the extant oversampling methods while reducing the false negatives, increase the false positives. In this paper, we propose a novel data oversampling method using Generative Adversarial Network (GAN). We use GAN and its variant to generate synthetic data of fraudulent transactions. To evaluate the effectiveness of the proposed method, we employ machine learning classifiers on the data balanced by GAN. Our proposed GAN-based oversampling method simultaneously achieved high precision, F1-score and dramatic reduction in the count of false positives compared to the state-of-the-art synthetic data generation based oversampling methods such as Synthetic Minority Oversampling Technique (SMOTE), Adaptive Synthetic Sampling (ADASYN) and random oversampling. Moreover, an ablation study involving the oversampling based on the ensemble of SMOTE and GAN/WGAN generated datasets indicated that it is outperformed by the proposed methods in terms of F1 score and false positive count.

MotivationThe application of machine learning approaches in phylogenetics has been impeded by the vast model space associated with inference. Supervised machine learning approaches require data from across this space to train models. Because of this, previous approaches have typically been limited to inferring relationships among unrooted quartets of taxa, where there are only three possible topologies. Here, we explore the potential of generative adversarial networks (GANs) to address this limitation. GANs consist of a generator and a discriminator: at each step, the generator aims to create data that is similar to real data, while the discriminator attempts to distinguish generated and real data. By using an evolutionary model as the generator, we use GANs to make evolutionary inferences. Since a new model can be considered at each iteration, heuristic searches of complex model spaces are possible. Thus, GANs offer a potential solution to the challenges of applying machine learning in phylogenetics.ResultsWe developed phyloGAN, a GAN that infers phylogenetic relationships among species. phyloGAN takes as input a concatenated alignment, or a set of gene alignments, and infers a phylogenetic tree either considering or ignoring gene tree heterogeneity. We explored the performance of phyloGAN for up to 15 taxa in the concatenation case and 6 taxa when considering gene tree heterogeneity. Error rates are relatively low in these simple cases. However, run times are slow and performance metrics suggest issues during training. Future work should explore novel architectures that may result in more stable and efficient GANs for phylogenetics.Availability and implementationphyloGAN is available on github: https://github.com/meganlsmith/phyloGAN/.

The need for fool proof authentication procedures away from traditional authentication mechanisms like passwords, security PINS has led to the advent of biometric authentication in information systems. Biometric data extracted from physiological features of a person including but not limited to fingerprints, palm prints, face or retina for purpose of verification & identification is saved as biometric templates. The inception of biometrics in access control systems has not been without its own hitches & like other systems it has its fair share of challenges. Biometric fingerprints being the most mature of all biometric spheres are the most widely adopted biometric authentication systems. Biometric systems effectiveness lies on how secure they are at preventing inadvertent disclosure of biometric templates in an information system’s archive. This however has not been the case as biometric templates have been fraudulently accessed to gain unauthorized access in identification and verification systems. In order to achieve strong and secure biometric systems, biometric systems developers need to build biometric systems that properly secure biometric templates. Several biometric template protection schemes and approaches have been proposed and used to safeguard stored biometric templates. Despite there being various biometric template protection schemes and approaches in existence, none of them has provided the most authentic, reliable, efficient and deterrent means to totally secure biometric fingerprint templates. This research sought to establish status of the current biometric template protection techniques and methods by conducting a survey and analyzing data gathered from a sample of seventy-eight (78) respondents. We will report these results and give our conclusion based on findings of the survey in this paper.

Depression disorder has become one of the major psychological diseases endangering human health. Researcher in the affective computing community is supporting the development of reliable depression severity estimation system, from multiple modalities (speech, face, text), to assist doctors in their diagnosis. However, the limited amount of annotated data has become the main bottleneck restricting the study on depression screening, especially when deep learning models are used. To alleviate this issue, in this work we propose to use Deep Convolutional Generative Adversarial Network (DCGAN) for features augmentation to improve depression severity estimation from speech. To the best of our knowledge, this approach is the first attempt to apply the Generative Adversarial Network for depression severity estimation from speech. Besides, to measure the quality of the augmented features, we propose three different measurement criteria, characterizing the spatial, frequency and representation learning of the augmented features. Finally, the augmented features are used to train depression estimation models. Experiments are carried out on speech signals from the Audio Visual Emotion Challenge (AVEC2016) depression dataset, and the relationship between the model performance and data size is explored. Our experimental results show that: 1) The combination of the three proposed evaluation criteria can effectively and comprehensively evaluate the quality of the augmented features. 2) When increasing the size of the augmented data, the performance of depression severity estimation gradually improves and the model converges to a certain stable state. 3) The proposed DCGAN based data augmentation approach effectively improves the performance of depression severity estimation, with the root mean square error (RMSE) reduced to 5.520 and mean absolute error (MAE) reduced to 4.634, which is better than most of the state of the art results on AVEC 2016.

Integrated biometrics template protection technique based on fingerprint and palmprint feature-level fusion

The security of information and personal privacy are the growing concerns in today’s human life worldwide. The storage of biometric data in the database has raised the prospect of compromising the database leading to grave risks and misuse of the person’s privacy such as growth in terrorism and identity fraud. When a person’s biometric data stored is revealed, their security and privacy are being compromised. This research described a detailed evaluation on several outbreaks and threats associated with the biometric technology. It analyzed the user’s fear and intimidations to the biometric technology alongside the protection steps for securing the biometric data template in the database. It is known that, when somebody’s biometric data template is compromised from the database that consequently might indicate proof of identity robbery of that person. Mixed method to compute and articulate the results as well as a new tactic of encryption-decryption algorithm with a design pattern of Model View Template (MVT) are used for securing the biometric data template in the database. The model managed information logically, the view indicated the visualization of the data, and the template directed the data migration into pattern object. Factors influencing fear of biometric technology such as an exposer of personal information, improper data transfer, and data misuse are found. Strong knowledge of the ideal technology like the private skills of the biometric technology, data secrecy and perceived helpfulness are established. The fears and attacks along the technology like a counterfeit of documents and brute-force attack are known. The designed algorithm based on the cryptographic module of the Fernet keys instance are utilized. The Fernet keys are combined to generate a multiFernet key, integrated with biometric data to produce two encrypted files (byte and text file). These files are incorporated with Twilio message and firmly stored in the database. The storage database has security measures that guard against an impostor’s attack. The database system can block the attacker from unauthorized access. Thus, significantly increased individual data privacy and integrity.

The fuzzy vault is an innovative way to share secret keys, combining traditional cryptography with biometrics and biometric template protection. This method forms the basis for the reliable operation of unmanned aerial vehicles (UAVs) through anonymizing drone operators and safely using their data and onboard information. However, due to the inherent instability of biometrics, traditional fuzzy vault schemes face challenges, such as reduced recognition rates with increased chaff points, impractical runtimes due to high-order polynomial reconstruction, and susceptibility to correlation attacks. This paper proposes an efficient fuzzy vault scheme to address these challenges. We generate two secret keys based on biometrics: the first key is produced from the operator’s unique features like the face and iris, using a confidence interval; the second key, used to construct a polynomial, is based on what the operator remembers. These dual-key fuzzy vaults enable the stable generation of genuine points during encoding, easy extraction during decoding, and effective operator authentication while maintaining anonymity. Our experimental results demonstrate improved security and secret acquisition accuracy using the AR face database. These results are achieved regardless of increased false vaults, enabling real-time polynomial reconstruction and resilience against correlation attacks. Importantly, our enhanced fuzzy vault scheme allows the application of this secure, real-time authentication process, safeguarding the anonymity of drone operators.