Markov chain model for multimodal biometric rank fusion

Multimodal biometric aims at increasing reliability of biometric systems through utilizing more than one biometric in decision-making process. An effective fusion scheme is necessary for combining information from various sources. Such information can be integrated at several distinct levels, such as sensor level, feature level, match score level, rank level, and decision level. In this paper, we present a multimodal biometric system utilizing face, iris, and ear biometric features through rank level fusion method using novel Markov chain approach. We first apply fisherimage technique to face and ear image databases for recognition and Hough transform and Hamming distance techniques for iris image recognition. The main contribution is in introducing Markov chain approach for biometric rank aggregation. One of the distinctive features of this method is that it satisfies the Condorcet criterion, which is essential in any fair rank aggregation system. The experimentation shows superiority of the proposed approach to other recently introduced biometric rank aggregation methods. The developed system can be effectively used by security and intelligence services for controlling access to prohibited areas and protecting important national or public information.     

Symmetric ear and profile face fusion for identical twins and non-twins recognition

Humans have bilateral body symmetry such that the left and right sides are mirror images of each other. This study tries to measure the performance on human recognition where the stored templates in the database are acquired from one side of a biometric trait such as left profile face, while the tested samples correspond to the other side of the same trait after applying a horizontal flip. Two different biometric traits are used in this study, namely profile face and ear biometrics. The experiments are conducted using the feature extraction methods namely Principal Component Analysis, Scale-Invariant Feature Transform, Local Binary Patterns, Local Phase Quantization and Binarized Statistical Image Features. Several experiments are performed on identical twins and non-twins individuals using ND-Twins-2009-2010 and UBEAR databases. Furthermore, the symmetry of profile face and ear is used to propose a hybrid approach of human recognition system that involves feature-level and score-level fusion of both traits. The proposed method is superior to all the unimodal and multimodal biometric methods that are implemented in this study for human recognition in the case of symmetry.     

Complex common vector for multimodal biometric recognition

A novel multimodal biometric recognition algorithm based on a complex common vector (CCV) is proposed. The CCV generalises the common vector method for the complex field to perform feature fusion and classification. Theoretical analysis proves that the CCV could produce a unique common vector for every fusion feature in a given class. The iris and the face are used as two distinct biometric modals to test the algorithm. Experimental results show that the proposed algorithm achieves much better performance than other conventional multimodal biometric algorithms.     

基于心电与光电容积脉搏波特征层融合的身份识别方法

针对单模态的心电信号(ECG)或光电容积脉搏波信号(PPG)识别技术中存在的精度不高,未考虑类内相关性等问题,该文提出基于判别相关分析法(DCA)对ECG与PPG组合特征矩阵进行特征层融合以及对K-最近邻(KNN)和支持向量机(SVM)分类器在决策层融合的识别方法。实验结果表明,使用融合特征(ECG-PPG)与融合分类器(KNN-SVM)的方法对23名受试者进行分类识别的准确率可以达到98.2%,识别精度在常规环境下优于单模态识别。为多模生物特征身份识别提供了一种有效模型。     

基于自适应并行结构的多模态生物特征识别

传统多模态生物特征识别方法当出现生物特征缺失时,识别性能会明显下降。针对此问题,提出一种融合人脸、虹膜和掌纹的自适应并行结构多模态生物识别方法。该方法在设计融合策略时,考虑到所有可能的输入缺失,构造并行结构的融合函数集,在实际应用时根据输入状态自适应的选择融合策略进行识别。实验仿真结果表明该方法既可提高识别可靠性又可实现当有生物特征缺失时的性能稳定。     

基于音视频的情感识别方法研究

近年来,情感计算逐渐成为人机交互发展突破的关键,而情感识别作为情感计算的重要部分,也受到了广泛的关注。本文实现了基于ResNet18的面部表情识别系统和基于HGFM架构的语音情感识别模型,通过调整参数,训练出了性能较好的模型。在此基础上,通过特征级融合和决策级融合这两种多模态融合策略,实现了包含视频和音频信号的多模态情感识别系统,展现了多模态情感识别系统性能的优越性。两种不同融合策略下的音视频情感识别模型相比视频模态和音频模态,在准确率上都有一定的提升,验证了多模态模型往往比最优的单模态模型的识别性能更好的结论。本文所实现的模型取得了较好的情感识别性能,融合后的音视频双模态模型的准确率达到了76.84%,与现有最优模型相比提升了3.50%,在与现有的音视频情感识别模型的比较中具有性能上的优势。      

An adaptive multimodal biometric management algorithm

This paper presents an evolutionary approach to the sensor management of a biometric security system that improves robustness. Multiple biometrics are fused at the decision level to support a system that can meet more challenging and varying accuracy requirements as well as address user needs such as ease of use and universality better than a single biometric system or static multimodal biometric system. The decision fusion rules are adapted to meet the varying system needs by particle swarm optimization, which is an evolutionary algorithm. This paper focuses on the details of this new sensor management algorithm and demonstrates its effectiveness. The evolutionary nature of adaptive, multimodal biometric management (AMBM) allows it to react in pseudoreal time to changing security needs as well as user needs. Error weights are modified to reflect the security and user needs of the system. The AMBM algorithm selects the fusion rule and sensor operating points to optimize system performance in terms of accuracy.     

移动设备生物特征识别多模态融合标准研究及展望

针对用于移动设备的生物特征识别多模态融合技术框架不统一以及标准缺失的问题,提出了多模态融合的分类、层级以及标准统一技术框架。首先分析国内外与移动设备生物特征识别相关的标准化现状;其次研究移动设备生物特征识别标准的本地识别以及远程识别应用模式,分析提出多特征、多算法、多实例、多传感器4种多模态融合分类方法,研究并提出样本级融合、特征级融合、分数级融合和决策级融合4种多模态融合的层级,并且提出用于移动设备的生物特征识别多模态融合标准技术框架;最后对移动设备生物特征识别多模态融合技术应用进行展望。     

Biometric bits extraction through phase quantization based on?feature level fusion

Biometric bits extraction has emerged as an essential technique for the study of biometric template protection as well as biometric cryptosystems. In this paper, we present a non-invertible but revocable bits extraction technique by means of quantizing the facial data from two feature extractors in the phase domain, which we coin as aligned feature-level fusion phase quantization (AFPQ). In this technique, we utilize helper data to achieve the revocability requirement of bits extraction. The feature averaging and remainder normalization technique are integrated with the helper data to reduce feature variance within the same individual and increase the distinctiveness of bit strings of different individuals to achieve good recognition performance. A scenario in which the system is compromised by an adversary is also considered. As a generic technique, AFPQ can be easily extended to multiple different biometric modalities.     

基于总错误率和特征关联的自适应融合多模态生物特征识别

针对单生物特征识别准确率和鲁棒性差的问题, 提出了一种基于总错误率(TER)和特征关联自适应融合多模态生物特 征识别方法。首先将TER作为判别特征引入到多模态识别,以代替传统的匹配分 数;其次在不确定度量理论的基 础上,考虑人脸特征和语音特征之间的时空关联性,提出了一种基于特征关联的多特征 自适应融合策略,利用特征关联 系数自适应调节不同识别特征对识别结果的贡献。仿真实验表明,与几种代表性的融合算法 相比,本文所 提出的融合模式可以有效提高多生物特征识别系统的准确性和鲁棒性。     

手指多模态Gabor编码特征局部融合方法研究

鲁棒性特征提取一直是生物特征识别领域研究的一个重要问题。由于手指姿态易变,这个问题在手指多模态生物特征描述方面显得更为突出。为了较为稳定地刻画手指多模态生物特征信息,本文提出了一种新的基于Gabor特征编码的局部灰度特征提取方法。首先,对手指的三个模态指纹、指静脉和指节纹图像分别进行Gabor滤波,刻画它们的纹理方向特性,并分别获取Gabor方向特征编码。然后,分别对特征编码图像进行局部灰度特征分析,并以局部串联的方式对手指多模态生物特征进行融合。实验表明,该方法在自制的手指姿态多变的数据库中具有良好的识别性能。      

Wireless Network Security Based on the Image Address Masking (IAM) Mechanism

A multimodal biometric system is applied to recognize individuals as authentication, identification and verification for claimed identity. Multimodal biometrics increases the security level accuracy, spoof of attacks, noise in collected data, intra-class variations, inter-class variations, non universality etc. In this paper a multi modal biometric algorithm is designed by integrating iris, palm print, face and signature based on encoded discrete wavelet transform for image analysis and authentication. Multi level wavelet based fusion approach is applied, integrated and encoded into single composite image for matching decision. It reduces the memory size, increases the recognition accuracy and ERR using multimodal biometric approach when compared to individual biometric traits. The complexity of fusion and the reconstruction algorithm is suitable for many real time applications.     

Factorial HMM and Parallel HMM for Gait Recognition

Information fusion offers a promising solution to the development of a high-performance classification system. In this paper, the problem of multiple gait features fusion is explored with the framework of the factorial hidden Markov model (FHMM). The FHMM has a multiple-layer structure and provides an alternative to combine several gait features without concatenating them into a single augmented feature. Besides, the feature concatenation is used to directly concatenate the features and the parallel HMM (PHMM) is introduced as a decision-level fusion scheme, which employs traditional fusion rules to combine the recognition results at decision level. To evaluate the recognition performances, McNemar's test is employed to compare the FHMM feature-level fusion scheme with the feature concatenation and the PHMM decision-level fusion scheme. Statistical numerical experiments are carried out on the Carnegie Mellon University motion of body and the Institute of Automation of the Chinese Academy of Sciences gait databases. The experimental results demonstrate that the FHMM feature-level fusion scheme and the PHMM decision-level fusion scheme outperform feature concatenation. The FHMM feature-level fusion scheme tends to perform better than the PHMM decision-level fusion scheme when only a few gait cycles are available for recognition.     

多通道信息融合的改进乘积规则

用传统的多通道信息融合乘积计算出的融合分值的动态范围比较小,融合效果不理想,本文在假定所有通道都能同时正常工作的情况下,提出多通道信息融合的改进乘积规则,基本思想是对原乘积规则中的概率密度和融合分值平滑,使得融合分值的动态范围增大,文中的两个实验表明,改进的乘积规则的识别率有明显提高,最后文中给出了在已知每个单一通道系统识别率的条件下,同理想融合方法集成出的系统识别率的上限计算公式。     

Novel mixture model–based approaches for person verification using multimodal biometrics

With growing concerns about security, the world over, biometric-based person verification is gaining more and more attention. Recently, multimodal biometric has attracted increasing focus among researchers as this overwhelms many limitations of unimodal biometric systems and hence more reliable. In this paper, we propose four different feature extraction techniques namely Principle Component Analysis Mixture Model (PCA MM), Singular Value Decomposition Mixture Model (SVD MM), Independent Component Analysis I Mixture Model (ICA I MM), and Independent Component Analysis II Mixture Model (ICA II MM) to design a multimodal biometric system at feature level. The proposed methods begin with modeling the multimodal biometrics data using Gaussian Mixture Model followed by a subspace methods like PCA, SVD, ICA I, and ICA II. Extensive experiments are carried out to observe the verification performance of the proposed methods at feature and match score level on large dataset of 150 users. We compare the results of the combined biometric with the results of individual biometric and also results of the proposed schemes against conventional (without mixture model) subspace approaches. The experimental results demonstrate the effectiveness of the proposed methods in designing a robust multimodal biometric system for accurate person verification.     

Enhanced Biometric Score Fusion Scheme Based on the AdaBoost Algorithm

Information fusion in biometric systems, either multimodal or intramodal fusion, usually provides an improvement in recognition performance. This paper presents an improved score-level fusion scheme called boosted score fusion. The proposed framework is a two-stage design where an existing fusion algorithm is adopted at the first stage. At the second stage, the weights obtained by the AdaBoost algorithm are utilized to boost the performance of the previously fused results. The experimental results demonstrate that the performance of several score-level fusion methods can be improved by using the presented method.     

Multimodal of Ear and Face Biometric Recognition Using Adaptive Approach Runge–Kutta Threshold segmentation and Classifier with Score Level Fusion

Wireless Personal Communications - A multimodal biometric system uses more than one biometric methods of an individual to mitigate some of the drawbacks of a unimodal biometric system and improve...     

A multimodal hand-based verification system with an aliveness-detection module

This paper presents a multimodal biometrie verification system based on the following hand features: palmprint, four digitprints and four fingerprints. The features are obtained using the Karhunen-Loève transform based approach, and information fusion at the matching-score level was applied. We experimented with different resolutions of the regions of interest, different numbers of features and several normalization and fusion techniques at the matching-score level. To increase the reliability of the system to spoof attacks we included an aliveness-detection module based on thermal images of the hand dor sa. The verification performance when using a system configuration with optimum parameters, i.e., resolution, number of features, normalization and fusion technique, showed an equal error rate (EER) of 0.0020%, which makes the system appropriate for the implementation of high-security biometric systems.