Multimodal Face and Ear Images

The biometrics is an alternative for password-based recognition systems in recent days. But biometric recognition systems are venerable to spoof attacks. Hence, anti-spoofing techniques are attracting growing interest in this field. Liveness detection is the anti-spoofing technique, which has the aim to identify the dead and living modalities presented at the time of authentication. This paper presents an efficient and robust multimodal biometric system integrated with liveness detection technique. In the proposed system, face and fingerprint modalities are used at the enrolment time and only face modality is used to authenticate a person. Here, fingerprint is used to generate a fingerprint key using K-means clustering algorithm and then encrypt the face feature set with asymmetric encryption standard (AES) using fingerprint key. At the time of authentication, liveness of a person is checked using challenge--response method at the sensor level. If in any case, an intruder has passed liveness test, he/she may be detected at matching score level as template of system is secured with AES encryption. Multimodal biometric system integrated with liveness detection technique can increase the speed of the system. Matrix Laboratory (MATLAB) 2017b and multi-biometric integration (MUBI) tool are used for the experimental work, and results show that the proposed system gives better performance as compared to others.KeywordsFace recognitionEncrypted face templateLiveness detectionMultimodal biometric system

Modern face recognition systems leverage datasets containing images of hundreds of thousands of specific individuals' faces to train deep convolutional neural networks to learn an embedding space that maps an arbitrary individual's face to a vector representation of their identity. The performance of a face recognition system in face verification (1:1) and face identification (1:N) tasks is directly related to the ability of an embedding space to discriminate between identities. Recently, there has been significant public scrutiny into the source and privacy implications of large-scale face recognition training datasets such as MS-Celeb-1M and MegaFace, as many people are uncomfortable with their face being used to train dual-use technologies that can enable mass surveillance. However, the impact of an individual's inclusion in training data on a derived system's ability to recognize them has not previously been studied. In this work, we audit ArcFace, a state-of-the-art, open source face recognition system, in a large-scale face identification experiment with more than one million distractor images. We find a Rank-1 face identification accuracy of 79.71% for individuals present in the model's training data and an accuracy of 75.73% for those not present. This modest difference in accuracy demonstrates that face recognition systems using deep learning work better for individuals they are trained on, which has serious privacy implications when one considers all major open source face recognition training datasets do not obtain informed consent from individuals during their collection.

How well can human faces be identified by humans and by computers, using subjectively judged descriptions like long ears, wide-set eyes, etc.? Three classes of experiments are reported: 1) Gathering, analysis, and assessment of face-feature data for 255 faces. 2) Computer identification-studies. 3) Human identification-studies. A set of 22 features was evolved from an initially larger set to provide relevant, distinctive, relatively independent measures which can be judged reliably. Computer studies and a mathematical model established limits of performance of a person attempting to isolate a face from a population using feature descriptions. The model predicts that under certain conditions approximately 6 of an individual's features are required to isolate him from a population of 255. Human experiments under similar conditions showed unique identification occurred with an average of about 7 features. The model predicts that for a population of 4×106, only 14 feature-descriptions are required. These studies form a foundation for continuing research on real-time man-machine interaction for computer classification and identification of multidimensional vectors specified by noisy components.

A method is presented to infer the presence of a human face in an image through the identification of face-like textures. The selected textures are those of human hair and skin. The second-order statistics method is used for texture representation. This method employs a set of co-occurrence matrices, from which features can be calculated that can characterize a texture. The cascade-correlation neural network architecture is used for supervised classification of textures. The Kohonen self-organizing feature map shows the clustering of the different texture types. Classification performance is generally above 80%, which is sufficient to clearly outline a face in an image.< <ETX xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">&gt;</ETX>

When trying to design artificial systems which may supplement or replace some task of human vision, it is often instructive to examine the features of human vision itself. The article outlines some of the things known about human face perception and recognition which helps one to understand how the human brain achieves the miraculous feats of face perception that it does. It outlines some perceptual illusions involving faces, and demonstrates what these tells about how the brain processes faces. It then describes how the important information which forms the basis for internal representations of faces appears to be based upon rather precise patterns of light and dark. Faces appear quite different if these patterns are altered in some way. These observations are important when we turn to the question of how we match different images of faces for identity. It also describes some of the work on the identification of images of faces in security video images, and considers the requirements for image engineering to help in this difficult area.

Biometrics is an ever-emerging field which is now being utilized in many applications and systems. Face biometrics are quite prevalent and have already been deployed in many devices and applications. Iris biometrics though already studied and deployed in certain areas are considered very effective due to their lesser chances of changing with time and uniqueness. Face detection and iris detection for verification have been employed in various biometric applications already. Unimodal systems of face and iris are secure, but for enhanced levels of security, multimodal biometric systems are being created. Multimodal biometric systems are those systems which use the various biometric traits simultaneously, in order to authenticate a person’s identity. There is a direct enhancement in the verification accuracy due to such systems. The fusion of these two biometrics in consideration to make the process of unauthenticated or unauthorized intervention complex because it will be difficult and inefficient to get multiple biometric traits simultaneously in real time. In order to draw a contrast between a unimodal approach and multimodal approach for the biometrics, this paper presents a survey on using individual face classifiers or individual iris classifiers and compares it with the combined classifier results. After analysing the various techniques, it is concluded with the directions towards the future scope.

In the field of data security, biometric security is a significant emerging concern. The multimodal biometrics system with enhanced accuracy and detection rate for smart environments is still a significant challenge. The fusion of an electrocardiogram (ECG) signal with a fingerprint is an effective multimodal recognition system. In this work, unimodal and multimodal biometric systems using Convolutional Neural Network (CNN) are conducted and compared with traditional methods using different levels of fusion of fingerprint and ECG signal. This study is concerned with the evaluation of the effectiveness of proposed parallel and sequential multimodal biometric systems with various feature extraction and classification methods. Additionally, the performance of unimodal biometrics of ECG and fingerprint utilizing deep learning and traditional classification technique is examined. The suggested biometric systems were evaluated utilizing ECG (MIT-BIH) and fingerprint (FVC2004) databases. Additional tests are conducted to examine the suggested models with:1) virtual dataset without augmentation (ODB) and 2) virtual dataset with augmentation (VDB). The findings show that the optimum performance of the parallel multimodal achieved 0.96 Area Under the ROC Curve (AUC) and sequential multimodal achieved 0.99 AUC, in comparison to unimodal biometrics which achieved 0.87 and 0.99 AUCs, for the fingerprint and ECG biometrics, respectively. The overall performance of the proposed multimodal biometrics outperformed unimodal biometrics using CNN. Moreover, the performance of the suggested CNN model for ECG signal and sequential multimodal system based on neural network outperformed other systems. Lastly, the performance of the proposed systems is compared with previously existing works.

Unimodal biometric systems often face significant limitations due to sensitivity to noise intraclass variability and other factors. Multimodal biometric systems seek to alleviate some of these problems by providing multiple pieces of evidence of the same identity. However, the multimodal biometric system is limited to the time constraints due to its multiple processing stages. To overcome these limitations this paper presents a fast multimodal verification system by using the dynamic regions of the fingerprint image and enhanced iris segmentation method. The multimodal system is fused using rank level fusion at the verification stage. The performance of the biometric system shows improvement in the False Acceptance Rate (FAR) and Equal Error Rate (EER) curves. Furthermore, the time taken for the training and verification phase has a reduction of 15% when compared with the existing system tested on the slow processing mobile devices.

A biometric system is an amalgamation of algorithms that uses an individual’s unique biological data innate physical features (fingerprints, retina, face, ears, etc.) or behavioural features (signature, voice, keystroke dynamics, etc.) or both to authenticate the individual. This essentially makes a biometric system function as a pattern recognition system that helps an individual in any kind of adversarial environment. Hence, like any conventional authentication system, it is vulnerable to the attacks of malicious adversaries that try to manipulate the data and thus decline the security of the system by compromising its robustness. These possible attacks have not been considered while developing most biometric authentication systems. Our work describes multimodal biometrics, its feature extraction methods, research challenges and focuses on the major security issues in both unimodal and multimodal biometric systems. This paper summarizes some robust multimodal systems, their countermeasures and algorithms that have been devised to tackle these security challenges.KeywordsBiometricsMultimodal biometricsBiometric fusionSecurityAttacksCountermeasuresEncryption

Biometric systems such as fingerprint, iris, DNA became popular methods in user authentication. Compared to these biometric systems, keystroke biometric authentication systems have not gained so much attention because of lower accuracy compared to other biometric systems. A number of researches have been conducted on keystroke biometric using different generative and discriminative classifiers. As Hidden Markov Models have proven a great success in voice recognition, this study investigates Hidden Markov Models in keystroke dynamic. This paper proposes a novel user verification technique using 1-substate Hidden Markov Model through keystroke dynamic. To verify the effectiveness of the proposed system, extensive experiments have been conducted and 80% accuracy was achieved by the proposed system.

Ear Recognition has recently received significant attention in the literature. Even though current ear recognition systems have reached a certain level of maturity, their success is still limited. This paper presents an efficient complete ear-based biometric system that can process five frames/sec; Hence it can be used for surveillance applications. The ear detection is achieved using Haar features arranged in a cascaded Adaboost classifier. The feature extraction is based on dividing the ear image into several blocks from which Local Binary Pattern feature distributions are extracted. These feature distributions are then fused at the feature level to represent the original ear texture in the classification stage. The contribution of this paper is three fold: (i) Applying a new technique for ear feature extraction, and studying various optimization parameters for that technique; (ii) Presenting a practical ear recognition system and a detailed analysis about error propagation in that system; (iii) Studying the occlusion effect of several ear parts. Detailed experiments show that the proposed ear recognition system achieved better performance (94:34%) compared to other shape-based systems as Scale-invariant feature transform (67:92%). The proposed approach can also handle efficiently hair occlusion. Experimental results show that the proposed system can achieve about (78%) rank-1 identification, even in presence of 60% occlusion.

Telling one face from another: Electrocortical correlates of facial characteristics among individual female faces

Due to the enormous usage of the internet for transmission of data over a network, security and authenticity become major risks. Major challenges encountered in biometric system are the misuse of enrolled biometric templates stored in database server. To describe these issues various algorithms are implemented to deliver better protection to biometric traits such as physical (Face, fingerprint, Ear etc.) and behavioural (Gesture, Voice, tying etc.) by means of matching and verification process. In this work, biometric security system with fuzzy extractor and convolutional neural networks using face attribute is proposed which provides different choices for supporting cryptographic processes to the confidential data. The proposed system not only offers security but also enhances the system execution by discrepancy conservation of binary templates. Here Face Attribute Convolutional Neural Network (FACNN) is used to generate binary codes from nodal points which act as a key to encrypt and decrypt the entire data for further processing. Implementing Artificial Intelligence (AI) into the proposed system, automatically upgrades and replaces the previously stored biometric template after certain time period to reduce the risk of ageing difference while processing. Binary codes generated from face templates are used not only for cryptographic approach is also used for biometric process of enrolment and verification. Three main face data sets are taken into the evaluation to attain system performance by improving the efficiency of matching performance to verify authenticity. This system enhances the system performance by 8% matching and verification and minimizes the False Acceptance Rate (FAR), False Rejection Rate (FRR) and Equal Error Rate (EER) by 6 times and increases the data privacy through the biometric cryptosystem by 98.2% while compared to other work.

The interface of computer technologies and biology is having a huge impact on society. Human recognition technology research projects promises new life to many security consulting firms and personal identification system manufacturers. The physiological or biological features of an individual, called biometrics, are unique to each human and remain unaltered during a person's lifetime and give a promising solution for security management. This paper presents an intelligent multimodal biometric verification system for physical access control, based on fusion of iris, face and fingerprint patterns. Feature vectors are created independently for query images and are then compared with the enrolled templates of each biometric trait to compute the matching score. The final decision is made by fusion at this matching score level. The proposed system is designed to suit embedded solutions for high security access in pervasive environments using biometric features.

A multi-modal biometric authentication system based on iterative fusion operator trees is considered, where the per-user performance is also depending on the order of biometric feature evaluations. Therefore, each user would have its own optimal choice of order parameters of the biometric system. However, for a system approach there should be only one general parametrization of the whole biometric system for all users, and the necessary consequence are performance wins and losses for each user. The common approach for this purpose is to use the parameter setting that maximizes the average performance over users. However, this does not necessarily reflect a good balance of performance wins and losses. In order to gain more insight into the corresponding trade-offs, and to facilitate an effective selection of an order parameter, the concept of fairness relations and their maximality, which has become a recent topic of many investigations in communication network resource sharing tasks, is applied to this problem of multi-modal biometric systems. The central idea of the paper is that by applying fairness relation to select the parameters, we obtain a system where subjects have a more fair chance of being right authenticated. A comparative study of an experimental test data set shows that maxmin fairness gives solutions with major drawbacks for the majority of users, while proportional fairness comes close to the maximum average.