Decision-level fusion in fingerprint verification

Prabhakar, Salil; Jain, Anil K.

doi:10.1016/s0031-3203(01)00103-0

Cited by 321 publications

(46 citation statements)

References 24 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Our approach consists of detecting discrete components in the genuine and impostor matching score distributions, and then modeling the observed distribution of matching scores as a mixture of discrete and continuous components. Hence, this approach generalizes the work of [11,12].…”

Section: Estimation Of Marginal Distributionssupporting

confidence: 56%

“…Thus, for a set of genuine and impostor matching scores, it is important to be able to estimate f gen (x) and f imp (x) reliably and accurately. Previous studies by Griffin [11] and Prabhakar et al [12] assume that the distribution function F has a continuous density with no discrete components. In reality, most matching algorithms apply thresholds at various stages in the matching process.…”

Section: Estimation Of Marginal Distributionsmentioning

confidence: 99%

See 1 more Smart Citation

A Principled Approach to Score Level Fusion in Multimodal Biometric Systems

Dass

Nandakumar

Jain

2005

Lecture Notes in Computer Science

Self Cite

136

View full text Add to dashboard Cite

Abstract. A multimodal biometric system integrates information from multiple biometric sources to compensate for the limitations in performance of each individual biometric system. We propose an optimal framework for combining the matching scores from multiple modalities using the likelihood ratio statistic computed using the generalized densities estimated from the genuine and impostor matching scores. The motivation for using generalized densities is that some parts of the score distributions can be discrete in nature; thus, estimating the distribution using continuous densities may be inappropriate. We present two approaches for combining evidence based on generalized densities: (i) the product rule, which assumes independence between the individual modalities, and (ii) copula models, which consider the dependence between the matching scores of multiple modalities. Experiments on the MSU and NIST multimodal databases show that both fusion rules achieve consistently high performance without adjusting for optimal weights for fusion and score normalization on a case-by-case basis.

show abstract

Section: Estimation Of Marginal Distributionssupporting

confidence: 56%

Section: Estimation Of Marginal Distributionsmentioning

confidence: 99%

A Principled Approach to Score Level Fusion in Multimodal Biometric Systems

Dass

Nandakumar

Jain

2005

Lecture Notes in Computer Science

Self Cite

136

View full text Add to dashboard Cite

show abstract

“…Recently, Lu Xiaoguang et. al., combined 3 independent classifiers (PCA, ICA, LDA) in their method, and experimental results show that the combination classifier is better than any of the individual classifiers [1][2][3]. Aiming at the problem of face recognition, this paper presents a method combined with support vector machine and the distance metric, constituting two class classifier combination method.…”

Section: Face Recognition Methods Based On Multiple Classifier Combinamentioning

confidence: 99%

Face Recognition Based on the Combination Method of Multiple Classifier

Yang¹,

Hao²

2016

IJSIP

View full text Add to dashboard Cite

show abstract

“…The steps include segmentation, 2,32 enhancement, 10,28,33 and matching, 8,24 etc. Second, researchers propose to combine the evidence obtained from multiple sources including multiple biometric traits, 1,9 multiple sensors, 18 multiple representations and matchers, 13,20,21 multiple¯ngers, 21 and multiple impressions of a same¯nger. 14,23,26 Third, new features are explored for matching beyond the most commonly used minutiae feature.…”

Section: Introductionmentioning

confidence: 99%

A Novel Method Using Videos for Fingerprint Verification

Wei

Yin

2012

Int. J. Patt. Recogn. Artif. Intell.

View full text Add to dashboard Cite

Traditional¯ngerprint veri¯cations use single image for matching. However, the veri¯cation accuracy cannot meet the need of some application domains. In this paper, we propose to use videos for¯ngerprint veri¯cation. To take full use of the information contained in¯ngerprint videos, we present a novel method to use the dynamic as well as the static information in ngerprint videos. After preprocessing and aligning processes, the Inclusion Ratio of two matching¯ngerprint videos is calculated and used to represent the similarity between these two videos. Experimental results show that video-based method can access better accuracy than the method based on single¯ngerprint.

show abstract

Decision-level fusion in fingerprint verification

Cited by 321 publications

References 24 publications

A Principled Approach to Score Level Fusion in Multimodal Biometric Systems

A Principled Approach to Score Level Fusion in Multimodal Biometric Systems

Face Recognition Based on the Combination Method of Multiple Classifier

A Novel Method Using Videos for Fingerprint Verification

Contact Info

Product

Resources

About