Equal Error Rate Research Articles

GNSS spoofing detection for UAVs using Doppler frequency and Carrier-to-Noise Density Ratio

Unmanned Aerial Vehicles (UAVs) are typical real-time embedded systems, which require precise locations for completing flight missions. The Global Navigation Satellite System (GNSS) plays a crucial role in navigation and positioning for UAVs. However, GNSS spoofing attacks pose an increasing threat to GNSS-dependent UAVs. Existing spoofing detection methods primarily rely on simulated data, perception data from multiple UAVs, or various control parameters. This paper proposes SigFeaDet, a signal feature-based GNSS spoofing detection approach for UAVs utilizing machine learning techniques. The core concept revolves around identifying anomalies in signal features arising from differences between authentic and spoofing signals. Key signal features, including Carrier-to-Noise Density Ratio (CN0) and Doppler frequency crucial for GNSS positioning, are employed to discern spoofing signals. Various machine learning algorithms are leveraged to train on GNSS signal data, determining the most effective classifier. TEXBAT GNSS dataset is processed to extract spoofing signal data, and flight experiments are conducted to gather GNSS data, augmenting the authentic GNSS signal dataset. The detection accuracy exceeds 95%. Equal Error Rate (EER) is approximately 5%. We evaluate various impact factors on SigFeaDet to show its robustness, including differences in velocities, altitudes, and experimental locations (10 kilometers apart), and the accuracy consistently surpasses 99%.

Separable contextual graph neural networks to identify tailgating-oriented traffic congestion

Identifying largescale traffic congestion with thousands of robots is a complicated task involving structured-data graphical modeling, and it has recently gained more research attention. This study aims to model traffic metrics that exhibit negative correlations in specific spatiotemporal contexts of automated material handling systems in semiconductor manufacturing. Numerous studies have focused on detecting vehicle congestion using multivariate time-series data. However, they faced challenges regarding the understanding of the complex graph structural joint spatiotemporal dynamics of the traffic rails. Existing methods fail to distinguish between congestion caused by high traffic volumes and nondisruptive delivery robot operations, such as parking and hoisting. This study introduces a novel approach to detect tailgating-oriented congestion using separable contextual graph neural networks (SC-GNNs). The proposed method employs deviation-based anomaly detection with attentive GNNs to capture spatial correlations and temporal dependencies. SC-GNNs were proposed for dual metrics with distinct characteristics, and dynamic thresholds were applied for each GNN to ensure reliable and interpretable anomaly detection. To demonstrate the outperformance of the proposed method, we conducted simulation experiments via a high-fidelity simulator similar to actual semiconductor fabrication. The experimental results showed that our method achieved both a 36% reduction in the false positive rate and a 32% decrease in the equal error rate compared to the best performance of existing state-of-the-art anomaly detection models. Overall, the proposed approach significantly improved the accuracy and interpretability of anomaly detection in multivariate time series, particularly effective in the context of identifying anomalies related to traffic congestion tailgating.

Effect of identical twins on deep speaker embeddings based forensic voice comparison

Deep learning has gained widespread adoption in forensic voice comparison in recent years. It is mainly used to learn speaker representations, known as embedding features or vectors. In this work, the effect of identical twins on two state-of-the-art deep speaker embedding methods was investigated with special focus on metrics of forensic voice comparison. The speaker verification performance has been assessed using the likelihood-ratio framework by likelihood ratio cost and equal error rate. The AVTD twin speech dataset was applied. The results show a significant reduction in speaker verification performance when twin samples are present. Neither the adaptation of LR score calculation to twin samples, nor fine-tuning the pre-trained speaker embedding models seemed to be able to leverage this limitation. It was found that the recognition of same or different speakers was possible even in the case of identical twins but the performance dropped greatly. The lowest EER of the best performing model was 3.4% in the case of non-twin; at the same time, EER was 25.3% when twins were present. This doesn’t mean that the presented methods are useless in case of identical twins, but it must be taken into consideration that in case of a higher likelihood-ratio score (which indicates same speakers on the tested samples), the possibility of twins must also be considered in a real casework.

Online Signature Biometrics for Mobile Devices.

This paper addresses issues concerning biometric authentication based on handwritten signatures. Our research aimed to check whether a handwritten signature acquired with a mobile device can effectively verify a user's identity. We present a novel online signature verification method using coordinates of points and pressure values at each point collected with a mobile device. Convolutional neural networks are used for signature verification. In this paper, three neural network models are investigated, i.e., two self-made light SigNet and SigNetExt models and the VGG-16 model commonly used in image processing. The convolutional neural networks aim to determine whether the acquired signature sample matches the class declared by the signer. Thus, the scenario of closed set verification is performed. The effectiveness of our method was tested on signatures acquired with mobile phones. We used the subset of the multimodal database, MobiBits, that was captured using a custom-made application and consists of samples acquired from 53 people of diverse ages. The experimental results on accurate data demonstrate that developed architectures of deep neural networks can be successfully used for online handwritten signature verification. We achieved an equal error rate (EER) of 0.63% for random forgeries and 6.66% for skilled forgeries.

Enhancing security in brain–computer interface applications with deep learning: Electroencephalogram-based user identification

Electroencephalogram (EEG) signals have gained widespread use in medical applications, and the utilization of EEG signals as a biometric feature for user identification systems in brain–computer interface systems has recently garnered significant attention. This paper presents a deep learning framework that uses a deep residual neural network (ResNet) for the identification of distinct individuals based on their EEG signals. The proposed framework utilizes continuous wavelet transform to convert one-dimensional EEG signals into two-dimensional spatial images. By incorporating both the frequency and time characteristics of the EEG signals, this model effectively considers and analyses the intricate details present in the data, leading to improved performance. The ResNet model considers the interdependencies between different time instances of the EEG signals. These unique features, collectively, provide sufficient discriminative information to accurately identify individuals. The proposed method achieved an exceptional classification accuracy of 99.73% and an equal error rate of 0.0041 for 64 channels within 109 individuals. The results illustrate the superiority of the proposed method over existing approaches.

Fixed-text keystroke dynamics authentication data set—collection and analysis

AbstractKeystroke dynamics authentication is a method of authenticating a user and could be an alternative or addition to one-time codes, with minimal user inconvenience. In this study, a new data set was collected for 6 unique passwords, adding to the limited available data sets for keystroke dynamics available for researchers. Data was collected by emulating legitimate users familiar with the passwords and a wider range of attackers with limited login attempts. The data set is analyzed with the use of various methods, and the effects of password length and complexity are investigated. Two algorithms were employed, one achieving an average equal error rate varying between 10.2 and 18.1% depending on the password, and the other method achieving an average true accept rate of 98% and true reject rate of 90.4% by comparing across multiple individuals in the data set. These results provide a benchmark for further studies on this data set.

An Efficient, Fast and Accurate Online Signature Verification Using Blended Feature Vector and Deep Learning

Online signature verification is one of the biometric authentication techniques. Online signature verification reduces the human error that may occur due to physical signature verification. Many researchers have provided a method of online signature verification still there is a scope for improvement. In this paper, a novel blended feature vector (BFV) is formed by combining two feature vectors. The first of these feature vectors is formed using a raw online signature database and the other feature vector is formed of the features extracted from the grayscale images obtained from the signature database. The algorithm makes use of the probability distribution of the features to provide a feature vector that results in an enhanced verification system. Two online signature databases, namely SVC2004 and ATVS-SSig are used. BFV is used to train the proposed mixed sequence deep neural network (MS-DNN). The use of a blended feature vector in the training of MS-DNN, instead of a whole online signature database, improves the training time of deep neural network to a great extent. This reduction in training time might be useful for new user registration. The performance parameters considered are validation efficiency and equal error rate. Results are compared with the existing state of technology and the comparison shows that the use of a blended feature vector as a classification vector improves the validation efficiency (99.5%) and also there is an improvement in the verification success rate in terms of equal error rate (EER 1.5%) as compared to the existing research.

Enhancing Speaker Recognition Models with Noise-Resilient Feature Optimization Strategies

This paper delves into an in-depth exploration of speaker recognition methodologies, with a primary focus on three pivotal approaches: feature-level fusion, dimension reduction employing principal component analysis (PCA) and independent component analysis (ICA), and feature optimization through a genetic algorithm (GA) and the marine predator algorithm (MPA). This study conducts comprehensive experiments across diverse speech datasets characterized by varying noise levels and speaker counts. Impressively, the research yields exceptional results across different datasets and classifiers. For instance, on the TIMIT babble noise dataset (120 speakers), feature fusion achieves a remarkable speaker identification accuracy of 92.7%, while various feature optimization techniques combined with K nearest neighbor (KNN) and linear discriminant (LD) classifiers result in a speaker verification equal error rate (SV EER) of 0.7%. Notably, this study achieves a speaker identification accuracy of 93.5% and SV EER of 0.13% on the TIMIT babble noise dataset (630 speakers) using a KNN classifier with feature optimization. On the TIMIT white noise dataset (120 and 630 speakers), speaker identification accuracies of 93.3% and 83.5%, along with SV EER values of 0.58% and 0.13%, respectively, were attained utilizing PCA dimension reduction and feature optimization techniques (PCA-MPA) with KNN classifiers. Furthermore, on the voxceleb1 dataset, PCA-MPA feature optimization with KNN classifiers achieves a speaker identification accuracy of 95.2% and an SV EER of 1.8%. These findings underscore the significant enhancement in computational speed and speaker recognition performance facilitated by feature optimization strategies.

Securing Internet-of-Medical-Things networks using cancellable ECG recognition

Reinforcement of the Internet of Medical Things (IoMT) network security has become extremely significant as these networks enable both patients and healthcare providers to communicate with each other by exchanging medical signals, data, and vital reports in a safe way. To ensure the safe transmission of sensitive information, robust and secure access mechanisms are paramount. Vulnerabilities in these networks, particularly at the access points, could expose patients to significant risks. Among the possible security measures, biometric authentication is becoming a more feasible choice, with a focus on leveraging regularly-monitored biomedical signals like Electrocardiogram (ECG) signals due to their unique characteristics. A notable challenge within all biometric authentication systems is the risk of losing original biometric traits, if hackers successfully compromise the biometric template storage space. Current research endorses replacement of the original biometrics used in access control with cancellable templates. These are produced using encryption or non-invertible transformation, which improves security by enabling the biometric templates to be changed in case an unwanted access is detected. This study presents a comprehensive framework for ECG-based recognition with cancellable templates. This framework may be used for accessing IoMT networks. An innovative methodology is introduced through non-invertible modification of ECG signals using blind signal separation and lightweight encryption. The basic idea here depends on the assumption that if the ECG signal and an auxiliary audio signal for the same person are subjected to a separation algorithm, the algorithm will yield two uncorrelated components through the minimization of a correlation cost function. Hence, the obtained outputs from the separation algorithm will be distorted versions of the ECG as well as the audio signals. The distorted versions of the ECG signals can be treated with a lightweight encryption stage and used as cancellable templates. Security enhancement is achieved through the utilization of the lightweight encryption stage based on a user-specific pattern and XOR operation, thereby reducing the processing burden associated with conventional encryption methods. The proposed framework efficacy is demonstrated through its application on the ECG-ID and MIT-BIH datasets, yielding promising results. The experimental evaluation reveals an Equal Error Rate (EER) of 0.134 on the ECG-ID dataset and 0.4 on the MIT-BIH dataset, alongside an exceptionally large Area under the Receiver Operating Characteristic curve (AROC) of 99.96% for both datasets. These results underscore the framework potential in securing IoMT networks through cancellable biometrics, offering a hybrid security model that combines the strengths of non-invertible transformations and lightweight encryption.

A standard ranking algorithm for robust iris template protection

In iris biometric recognition systems, protecting the storage and transmission of iris templates is crucial, and template protection techniques are pivotal for ensuring their security. A prevalent approach involves using indexing methods as an effective algorithm for iris template protection, leveraging the index or rank of the extracted iris code to generate a secure iris template. Meantime, many privacy threats to biometric data have emerged, necessitating heightened protection measures. Specifically, protecting the privacy of iris data is imperative within the context of iris template protection during recognition processes. As stipulated by the international standard ISO/IEC 30136, effective iris template protection must concurrently meet the criteria of irreversibility, revocability, and unlinkability. Nevertheless, existing indexing methods on iris template protection faced the formidable challenge of simultaneously fulfilling these three privacy requirements while maintaining the efficacy of iris recognition. This paper introduces a standard ranking (standardR) algorithm, named standardR, designed to enhance the security of iris templates by transforming each iris template into an irreversible representation. The experimental results on the benchmarked Casia-Iris-interval dataset, along with two additional iris datasets MMU1 and UBRIS 1, demonstrate the efficacy of the proposed algorithm. The proposed standardR algorithm achieves an equal error rate (EER) of 0.1695% and an area under the curve of 0.93011% with the Casia-Iris-Interval dataset. Furthermore, the algorithm maintains efficient recognition with a reduced iris code length of 1280 bits, a time complexity of O(n log n), and satisfies the biometric template protection (BTP) requirements in irreversibility, unlinkability, and renewability.

T-EER: Parameter-Free Tandem Evaluation of Countermeasures and Biometric Comparators.

Presentation attack (spoofing) detection (PAD) typically operates alongside biometric verification to improve reliablity in the face of spoofing attacks. Even though the two sub-systems operate in tandem to solve the single task of reliable biometric verification, they address different detection tasks and are hence typically evaluated separately. Evidence shows that this approach is suboptimal. We introduce a new metric for the joint evaluation of PAD solutions operating in situ with biometric verification. In contrast to the tandem detection cost function proposed recently, the new tandem equal error rate (t-EER) is parameter free. The combination of two classifiers nonetheless leads to a set of operating points at which false alarm and miss rates are equal and also dependent upon the prevalence of attacks. We therefore introduce the concurrent t-EER, a unique operating point which is invariable to the prevalence of attacks. Using both modality (and even application) agnostic simulated scores, as well as real scores for a voice biometrics application, we demonstrate application of the t-EER to a wide range of biometric system evaluations under attack. The proposed approach is a strong candidate metric for the tandem evaluation of PAD systems and biometric comparators.

Continuous Authentication in the Digital Age: An Analysis of Reinforcement Learning and Behavioral Biometrics

This research article delves into the development of a reinforcement learning (RL)-based continuous authentication system utilizing behavioral biometrics for user identification on computing devices. Keystroke dynamics are employed to capture unique behavioral biometric signatures, while a reward-driven RL model is deployed to authenticate users throughout their sessions. The proposed system augments conventional authentication mechanisms, fortifying them with an additional layer of security to create a robust continuous authentication framework compatible with static authentication systems. The methodology entails training an RL model to discern atypical user typing patterns and identify potentially suspicious activities. Each user’s historical data are utilized to train an agent, which undergoes preprocessing to generate episodes for learning purposes. The environment involves the retrieval of observations, which are intentionally perturbed to facilitate learning of nonlinear behaviors. The observation vector encompasses both ongoing and summarized features. A binary and minimalist reward function is employed, with principal component analysis (PCA) utilized for encoding ongoing features, and the double deep Q-network (DDQN) algorithm implemented through a fully connected neural network serving as the policy net. Evaluation results showcase training accuracy and equal error rate (EER) ranging from 94.7% to 100% and 0 to 0.0126, respectively, while test accuracy and EER fall within the range of approximately 81.06% to 93.5% and 0.0323 to 0.11, respectively, for all users as encoder features increase in number. These outcomes are achieved through RL’s iterative refinement of rewards via trial and error, leading to enhanced accuracy over time as more data are processed and incorporated into the system.

Efficient Deepfake Audio Detection Using Spectro-Temporal Analysis and Deep Learning

With the advancement of deepfake technology, particularly in the audio domain, there is an imperative need for robust detection mechanisms to maintain digital security and integrity. Our research presents an "Enhanced Deepfake Audio Detection with Spectro-Temporal Deep Learning Approach," aimed at addressing the escalating challenge of distinguishing genuine audio from sophisticated deepfake manipulations. Our methodology leverages the ADD2022 dataset, which encompasses a wide range of audio clips, to train and evaluate a novel deep learning model. The core of our approach consists of a meticulous data preprocessing phase, where audio samples are resampled, normalized, and subjected to silence removal to ensure uniformity and enhance model input quality. Our deep learning model architecture innovatively combines Convolutional Neural Networks (CNNs) for extracting spectral features and Recurrent Neural Networks (RNNs) or Long Short-Term Memory (LSTM) networks for analyzing temporal dynamics. This hybrid model is adept at identifying the nuanced anomalies characteristic of deepfake audios. The model undergoes rigorous training and optimization, with performance evaluated against key metrics such as accuracy, precision, F1, recall score, and the Equal Error Rate (EER). Our findings exhibit the Enhanced Deepfake Audio Detection model's superior capability to discern deepfake audio, highlighting its potential as a pivotal tool in the fight against digital audio manipulation. This research contributes significantly to the field of digital forensics, offering a scalable and effective solution for ensuring the authenticity of audio content in an era dominated by deepfake technology.

The Proposal of Countermeasures for DeepFake Voices on Social Media Considering Waveform and Text Embedding

In recent times, advancements in text-to-speech technologies have yielded more natural-sounding voices. However, this has also made it easier to generate malicious fake voices and disseminate false narratives. ASVspoof stands out as a prominent benchmark in the ongoing effort to automatically detect fake voices, thereby playing a crucial role in countering illicit access to biometric systems. Consequently, there is a growing need to broaden our perspectives, particularly when it comes to detecting fake voices on social media platforms. Moreover, existing detection models commonly face challenges related to their generalization performance. This study sheds light on specific instances involving the latest speech generation models. Furthermore, we introduce a novel framework designed to address the nuances of detecting fake voices in the context of social media. This framework considers not only the voice waveform but also the speech content. Our experiments have demonstrated that the proposed framework considerably enhances classification performance, as evidenced by the reduction in equal error rate. This underscores the importance of considering the waveform and the content of the voice when tasked with identifying fake voices and disseminating false claims.

DeepDet: YAMNet with BottleNeck Attention Module (BAM) for TTS synthesis detection

Spoofed speeches are becoming a big threat to society due to advancements in artificial intelligence techniques. Therefore, there must be an automated spoofing detector that can be integrated into automatic speaker verification (ASV) systems. In this study, we recommend a novel and robust model, named DeepDet, based on deep-layered architecture, to categorize speech into two classes: spoofed and bonafide. DeepDet is an improved model based on Yet Another Mobile Network (YAMNet) employing a customized MobileNet combined with a bottleneck attention module (BAM). First, we convert audio into mel-spectrograms that consist of time–frequency representations on mel-scale. Second, we trained our deep layered model using the extracted mel-spectrograms on a Logical Access (LA) set, including synthesized speeches and voice conversions of the ASVspoof-2019 dataset. In the end, we classified the audios, utilizing our trained binary classifier. More precisely, we utilized the power of layered architecture and guided attention that can discern the spoofed speech from bonafide samples. Our proposed improved model employs depth-wise linearly separate convolutions, which makes our model lighter weight than existing techniques. Furthermore, we implemented extensive experiments to assess the performance of the suggested model using the ASVspoof 2019 corpus. We attained an equal error rate (EER) of 0.042% on Logical Access (LA), whereas 0.43% on Physical Access (PA) attacks. Therefore, the performance of the proposed model is significant on the ASVspoof 2019 dataset and indicates the effectiveness of the DeepDet over existing spoofing detectors. Additionally, our proposed model is robust enough that can identify the unseen spoofed audios and classifies the several attacks accurately.

Fingerprint Reconstruction: Approaches to Improve Fingerprint Images

Fingerprint reconstruction methods have been initially proposed to spoof the fingerprint identification systems, wherein the fingerprints are generated from the fingerprint features stored in the database for template matching/identification purpose. The reconstructed fingerprints attempt to validate in the absence of the user/person. The poor fingerprint Images with scratches on fingerprint image or latent fingerprints or overlapping fingerprints shall also be reconstructed for personality identification. In this paper we discuss the two fingerprint reconstruction methods, one which uses minutiae features for reconstruction and the other one uses deep learning methods to reconstruct the fingerprint images. The poor fingerprint image which fails to validate the identity due to various reasons like poor skin condition/large cuts on the fingers/wet fingers/poor scanning of images shall be reconstructed for increasing the matching accuracy. The requirement of performance measure parameters used for evaluation of these systems are equal error rate, false acceptance rate, false rejection rate and average matching score. The deep learning methods are more suitable for reconstructing the fingerprint images that appear damaged due to poor skin condition/large cuts on the fingers/wet fingers/poor scanning of images. In terms of matching score comparison, the deep learning methods have matching scores in between 23-94% whereas for minutiae-based techniques the matching score is between 82 and 99.99%. The other performance parameter is the equal error rate (ERR) required to meet has to be closer to 0. The matching score is computed with the assumptions of false acceptance rate (FAR) ranging from 1% to 0%.

Whisper40: A Multi-Person Chinese Whisper Speaker Recognition Dataset Containing Same-Text Neutral Speech

Whisper speaker recognition (WSR) has received extensive attention from researchers in recent years, and it plays an important role in medical, judicial, and other fields. Among them, the establishment of a whisper dataset is very important for the study of WSR. However, the existing whisper dataset suffers from the problems of a small number of speakers, short speech duration, and lack of neutral speech with the same-text as the whispered speech in the same dataset. To address this issue, we present Whisper40, a multi-person Chinese WSR dataset containing same-text neutral speech spanning around 655.90 min sourced from volunteers. In addition, we use the current state-of-the-art speaker recognition model to build a WSR baseline system and combine the idea of transfer learning for pre-training the speaker recognition model using neutral speech datasets and transfer the empirical knowledge of specific network layers to the WSR system. The Whisper40 and CHAINs datasets are then used to fine-tune the model with transferred specific layers. The experimental results show that the Whisper40 dataset is practical, and the time delay neural network (TDNN) model performs well in both the same/cross-scene experiments. The equal error rate (EER) of Chinese WSR after transfer learning is reduced by 27.62% in comparison.

Enhancing Recognition in Multimodal Biometric Systems: Score Normalization and Fusion of Online Signatures and Fingerprints

Multimodal biometrics employs multiple modalities within a single system to address the limitations of unimodal systems, such as incomplete data acquisition or deliberate fraud, while enhancing recognition accuracy. This study explores score normalization and its impact on system performance. To fuse scores effectively, prior normalization is necessary, followed by a weighted sum fusion technique that aligns impostor and genuine scores within a common range. Experiments conducted on three biometric databases demonstrate the promising efficacy of the proposed approach, particularly when combined with Empirical Modal Decomposition (EMD). The fusion system exhibits strong performance, with the best outcome achieved by merging the online signature and fingerprint modalities, resulting in a normalized Min-Max score-based Equal Error Rate (EER) of 1.69%.

THE NO TRAIN NO GAIN SYSTEM FOR O-COCOSDA AND VLSP 2022 - A-MSV SHARED TASK: ASIAN MULTILINGUAL SPEAKER VERIFICATION

This paper proposes a semi-supervised multilingual speaker verification (MSV) system submitted for the 2 tasks, MSV for the Asian language inside the training set (T01) and outside the training set (T02) in O-COCOSDA and VLSP challenge 2022.To solve the problem, our strategy is training a baseline acoustic model with given labeled data (MSV CommonVoice) andfine-tuning the trained acoustic model with both given labeled data and given unlabeled data (MSV Youtube). To achieve the fine-tuning step, the unlabeled data is converted to labeled data by pseudo labeling technique using the clustering method with the embedding vectors extracted from the trained acoustic model. Besides, we also apply test-time augmentation, back-end scoring, and score normalization with the AS-Norm technique to improve the result. When evaluated on the VLSP 2022 challenge's given test set, our best system with baseline ECAPA-TDNN achieves an equal error rate (EER) of 2.296% in T01 and 3.3296% in T02, which ranks second rank in both two tasks.

Multimodal biometric user authentication using improved decentralized fuzzy vault scheme based on Blockchain network

Biometric security has emerged as a major concern in the field of Biometric Systems. The existing centralized Biometric Cryptosystems approach still suffers from privacy issues and network-based attacks. In particular, the Biometric Cryptosystems are susceptible to Attack via Record Multiplicity, brute force and collusion attacks. To address these issues, this paper proposes a multimodal biometric user authentication scheme using a decentralized fuzzy vault based on Blockchain technology. In this scheme, the facial biometric images and dorsal hand images together form the multimodal biometric. The proposed security framework is implemented in two phases. In the first phase, multimodal biometric user authentication using an improved fuzzy vault scheme is implemented. In the second phase, a privacy preservation technique using a decentralized fuzzy vault biometric system based on the Blockchain network is implemented. The performance of the proposed framework is verified on the biometric datasets consisting of both genuine and imposter users, and a comparative analysis with the existing state-of-the-art approaches is also presented. The Detection Error Trade-off (DET) curve and the Receiver Operating Characteristic (ROC) curve, obtained for verifying the efficacy of the proposed framework, are also presented. The proposed framework yielded an accuracy of 99.8% and an Equal Error Rate of 0.08. Furthermore, to validate the biometric template protection, the security analysis of the proposed scheme based on irreversibility, unlinkability, and revocability criteria is described. The proposed framework achieves significant improvement in terms of a low False Acceptance Rate (0.08), and a high True Acceptance Rate (0.96).