Biometric match score fusion using RVM: A case study in multi-unit iris recognition
16 Jun 2012-pp 65-70
TL;DR: Experimental results on the CASIA-Iris-V4 Thousand database show that RVM provides better accuracy compared to single unit iris recognition and existing fusion algorithms.
Abstract: This paper presents a novel fusion approach to combine scores from different biometric classifiers using Relevance Vector Machine. RVM uses a combination of kernel functions on training data for classification and compared to SVM, it requires significantly reduced number of relevance vectors. The proposed RVM based fusion algorithm is evaluated using a case study on multi-unit iris recognition. Experimental results on the CASIA-Iris-V4 Thousand database show that RVM provides better accuracy compared to single unit iris recognition and existing fusion algorithms. With respect to SVM fusion, it is observed that, the accuracy of RVM and SVM are comparable, however, the time for RVM fusion is significantly reduced.
Citations
More filters
[...]
TL;DR: A path forward is proposed to advance the research on ocular recognition by improving the sensing technology, heterogeneous recognition for addressing interoperability, utilizing advanced machine learning algorithms for better representation and classification, and developing algorithms for ocular Recognition at a distance.
Abstract: Display Omitted A literature review of ocular modalities such as iris and periocular is presented.Information fusion approaches that combine ocular modalities with other modalities are reviewed.Future research directions are presented on sensing technologies, algorithms, and fusion approaches. Biometrics, an integral component of Identity Science, is widely used in several large-scale-county-wide projects to provide a meaningful way of recognizing individuals. Among existing modalities, ocular biometric traits such as iris, periocular, retina, and eye movement have received significant attention in the recent past. Iris recognition is used in Unique Identification Authority of India's Aadhaar Program and the United Arab Emirate's border security programs, whereas the periocular recognition is used to augment the performance of face or iris when only ocular region is present in the image. This paper reviews the research progression in these modalities. The paper discusses existing algorithms and the limitations of each of the biometric traits and information fusion approaches which combine ocular modalities with other modalities. We also propose a path forward to advance the research on ocular recognition by (i) improving the sensing technology, (ii) heterogeneous recognition for addressing interoperability, (iii) utilizing advanced machine learning algorithms for better representation and classification, (iv) developing algorithms for ocular recognition at a distance, (v) using multimodal ocular biometrics for recognition, and (vi) encouraging benchmarking standards and open-source software development.
138 citations
Cites background or methods from "Biometric match score fusion using ..."
...[218] explore the application of Relevance Vector Machines (RVM) to perform score-level fusion from different classifiers....
[...]
...[218] CASIA v4 Relevance Vector Machines (RVM) perform score-level fusion for multiple irises....
[...]
TL;DR: The proposed iGRVM which incorporates incremental and granular learning in RVM can be a good alternative for biometric score classification with faster testing time.
Abstract: This paper focuses on extending the capabilities of relevance vector machine which is a probabilistic, sparse, and linearly parameterized classifier. It has been shown that both relevance vector machine and support vector machine have similar generalization performance but RVM requires significantly fewer relevance vectors. However, RVM has certain limitations which limits its applications in several pattern recognition problems including biometrics such as (1) slow training process, (2) difficult to train with large training samples, and (3) may not be suitable to handle large class imbalance. To address these limitations, we propose iGRVM which incorporates incremental and granular learning in RVM. The proposed classifier is evaluated in context to multimodal biometrics score classification using the NIST BSSR1, CASIA-Iris-Distance V4, and Biosecure DS2 databases. The experimental analysis illustrates that the proposed classifier can be a good alternative for biometric score classification with faster testing time. HighlightsThe proposed iGRVM incorporates incremental and granular learning in RVM.Experiments are performed on NIST BSSR1, CASIA-Iris-Distance V4, and Biosecure DS2 databases.Results illustrate that iGRVM can be a good alternative for biometric score classification.
26 citations
01 Dec 2013
TL;DR: This work proposes a reliable two-stage multi-instance finger vein recognition system based on minutiae matching method by integrating a unified minutia alignment and pruning approach using Genetic algorithm and the k-modified Hausdorff distance measurement.
Abstract: Among the various multi-modal biometric approaches, multi-instance biometric appears to be understudied despite it inherits the merits of multimodal biometrics system. Multi-instance biometrics is useful when the signal quality is too low for robust verification. As compared to other multi-modal approach, multi-instance fusion reduces the need of multiple acquisitions using different sensors and thus lessen both transaction time and sensor cost. In this work, we propose a reliable two-stage multi-instance finger vein recognition system based on minutiae matching method by integrating a unified minutia alignment and pruning approach using Genetic algorithm and the k-modified Hausdorff distance (k-MHD) measurement. The proposed method is evaluated by using the SDUMLA-HMT Finger Vein database. Experiments show the proposed method is able to attain promising recognition rate compared to its single biometrics counterpart. The best result is achieved by applying the k-nearest neighbor measurement alongside, where the recognition rate can be up to 99.7% when MHD is used for matching.
20 citations
Cites result from "Biometric match score fusion using ..."
...As compared to other multi-modal approach, it reduces the need of multiple acquisitions using different sensors and thus lessen both transaction time and sensor cost....
[...]
24 Nov 2016
TL;DR: In this article, the authors present an overview of the multi-biometric score-level fusion problem, along with the proposed solution in the literature, and a discussion is made to provide a clearer view of future developments especially under the identification scenario where many related applications are rapidly growing.
Abstract: Multi-biometrics is the use of multiple biometric recognition sources to provide a more dependable verification or identification decision. Fusion of multi-biometric sources can be performed on different levels, such as the data, feature, or score level. This work presents an overview of the multi-biometric score-level fusion problem, along with the proposed solution in the literature. A discussion is made to provide a comparison between multi-biometric fusion in both scenarios. This discussion aims at providing a clearer view of future developments especially under the identification scenario where many related applications are rapidly growing such as forensics and ubiquitous surveillance.
16 citations
01 Jan 2020
TL;DR: In this paper, a multimodal biometric system uses more than one biometric trait or modality for recognition of an individual, which fuses different types of input at different levels: Score level, Feature level and Decision level.
Abstract: Human identification systems based on biometrics are used in many applications to increase the security level. There are different biometric traits which are used in various applications. Monomodal biometric systems face many challenges such as error rates, using only single biometric for human recognition. Today, to increase the security of the authentication system, various multimodal biometric systems are proposed. A multimodal biometric system uses more than one biometric trait or modality for recognition of an individual. Multimodal biometric systems fuses different types of input at different level: Score level, Feature level and Decision level to get the better performance of the system.
16 citations
References
More filters
TL;DR: A multimodal biometric approach for identity verification using two competent traits, iris and retina, which diminishes the drawback of single biometric system and improves the performance of an authentication system.
Abstract: The recognition accuracy of a single biometric authentication system is often much reduced due to the environment, user mode and physiological defects. So a multimodal biometric approach for identity verification using two competent traits, iris and retina is proposed. This multimodal approach diminishes the drawback of single biometric system and improves the performance of an authentication system. Iris and Retina biometric recognition offers a highly reliable solution to person authentication. Iris recognition system is composed of segmentation, normalization, feature encoding and matching. Instead of using the entire iris code, only the bits that are consistent in the iris code called the Best bits are considered in the feature matching process. This reduces the computational time and storage requirements of iris code. To enhance the performance of recognition, the iris recognition process is applied to left and right irises separately and the corresponding distance scores are generated for each iris of a person. These scores are combined using the weighted sum fusion rule which further increases the recognition rate. In order to provide liveness verification for our authentication system, we also employed retinal blood vessel pattern recognition. This ensures the presence of only alive persons eliminating the possible spoofing attack. It is composed of segmentation, enhancement, feature encoding and matching. The scores from iris and retina recognition are then combined using weighted sum fusion rule, which further increases the recognition rate. To validate our approach, experiments were conducted on the iris and retina images obtained from CASIA and VARIA datasets respectively. A multimodal biometric database was constructed and then 1500 inter, intra image comparisons were made using the two datasets. The experimental results reveal that our multimodal biometric authentication system is much more reliable and precise than the single biometric approaches.
26 citations
"Biometric match score fusion using ..." refers background in this paper
...It is our assertion that more research is required to utilize the full potential of RVM in biometrics....
[...]
TL;DR: The proposed iris recognition system, which can select the good quality data between left and right eye images of same person, is composed of four steps and Support Vector Machines and Euclidian distance are used as classification methods.
Abstract: In this paper, we propose the iris recognition system, which can select the good quality data between left and right eye images of same person. Although iris recognition system has achieved good performance, but it is affected by the quality of input images. So, eye image check algorithm, which can select the good quality image is very important. The proposed system is composed of four steps. At the first step, both eye images are captured at the same time. At the second step, the eye image check algorithm picks out noisy and counterfeit data between both eye images and offer a good qualified image to the next step. At the third step, Daubechies' Wavelet is used as a feature extraction method. Finally, Support Vector Machines(SVM) and Euclidian distance are used as classification methods. Experiment results involve 1694 eye images of 111 different people and the best accuracy rate of 99.1%.
17 citations
"Biometric match score fusion using ..." refers methods in this paper
...The proposed RVM fusion algorithm is evaluated in context to multiunit iris recognition....
[...]
04 Feb 2009
TL;DR: Experimental results on multi-instance and multi-unit iris verification show that the proposed fusion framework with PCR rule yields the best verification accuracy even when individual biometric classifiers provide highly conflicting match scores.
Abstract: This paper presents a framework for multi-biometric match score fusion when non-ideal conditions cause conflict in the results of different classifiers. The proposed framework uses belief function theory to effectively fuse the match scores and density estimation technique to compute the belief assignments. Fusion is performed using belief models such as Transferable Belief Model (TBM) and Proportional Conflict Redistribution (PCR) Rule followed by the likelihood ratio based decision making. Experimental results on multi-instance and multi-unit iris verification show that the proposed fusion framework with PCR rule yields the best verification accuracy even when individual biometric classifiers provide highly conflicting match scores.
11 citations
"Biometric match score fusion using ..." refers background in this paper
...It is our assertion that more research is required to utilize the full potential of RVM in biometrics....
[...]
27 Aug 2007
TL;DR: This paper investigates the feasibility of fusing both irises for personal authentication and the performance of some very simple fusion strategies and shows that the difference between the left and the right irises of the same persons is close to the Difference between the irises captured from different persons.
Abstract: Traditional personal authentication methods have many instinctive defects. Biometrics is an effective technology to overcome these defects. Among the available biometric approaches, iris recognition is one of the most accurate techniques. Combining the left and the right irises of same persons can improve the authentication accuracy and reduce the spoof attack risks. Furthermore, the fusion need not add any other hardware to the existing iris recognition systems. This paper investigates the feasibility of fusing both irises for personal authentication and the performance of some very simple fusion strategies. The experimental results show that the difference between the left and the right irises of the same persons is close to the difference between the irises captured from different persons. And combining the information of both irises can dramatically improve the authentication accuracy even when the quality of the iris images are not good enough. The results also show that the Minimum and the Product strategies can obtain the perfect performance, i.e. both FARs and FRRs of these two strategies can be reduce to 0%.
10 citations
"Biometric match score fusion using ..." refers background in this paper
...It is our assertion that more research is required to utilize the full potential of RVM in biometrics....
[...]