Book Chapter10.1017/CBO9780511921056.017
Multibiometrics for Human Identification: Predicting Performance in Large-Scale Identification Systems by Score Resampling
Sergey Tulyakov,Venu Govindaraju +1 more
4
TL;DR: This paper identifies two effects responsible for the prediction errors in previously proposed predictions: the binomial approximation effect and the score mixing effect and proposes a score resampling method for prediction, which is not susceptible to the bino mial approximation effect.
read more
Abstract: In this paper we investigate the problem of predicting the closed set identification performance of biometric matchers in large-scale applications giv en their corresponding performances in small-scale applications. We identify two maj r effects responsible for the prediction errors in previously proposed m thods: the binomial approximation effect and the score mixing effect. We propose t use a score resampling method for prediction, which is not susceptible to the bino mial approximation effect. We also reduce score mixing effect by using sco re selection based on identification trial statistics. The experiments on NIST biometric sco re dataset show the accuracy of our proposed prediction method.
read more
Chat with Paper
AI Agents for this Paper
Find similar papers on Google Scholar, PubMed and Arxiv
Write a critical review of this paper
Analyze citations of this paper to find unaddressed research gaps
Citations
On the Relation Between ROC and CMC
Raymond Veldhuis,Kiran B. Raja +1 more
TL;DR: It is proved that the probabilistic CMC plotted as a function of fractional rank, i.e., linearly compressed to a domain ranging from 0 to 1, will converge to the average Probabilistic ROC when the gallery size increases.
2
Matching Score Fusion Methods
Sergey Tulyakov,Venu Govindaraju +1 more
TL;DR: This chapter describes the complexity types of combination methods and characterize some of the existing fusion methods using these types, and provides suggestions on how more powerful higher complexity combinations can be constructed.
References
Parameter and quantile estimation for the generalized pareto distribution
TL;DR: In this paper, the authors show that unless the sample size is 500 or more, estimators derived by either the method of moments or probability-weighted moments are more reliable.
1.3K
Extreme Value Dependence in Financial Markets: Diagnostics, Models, and Financial Implications
TL;DR: In this article, the authors present a general framework for identifying and modeling the joint-tail distribution based on multivariate extreme value theories, arguing that the multivariate approach is the most efficient and effective way to study extreme events such as systemic risk and crisis.
697
The relation between the ROC curve and the CMC
R.M. Bolle,Jonathan H. Connell,Sharath Pankanti,Nalini K. Ratha,Andrew W. Senior +4 more
- 17 Oct 2005
TL;DR: The cumulative match curve (CMC) is used as a measure of 1: m identification system performance and is related to the FAR and FRR of a 1:1 matcher, i.e., the matcher that is used to rank the candidates by sorting the scores.
211
Error analysis of pattern recognition systems: the subsets bootstrap
TL;DR: It is argued that biometric match score accuracy is best expressed in terms of a curve, the Receiver Operating Characteristic curve, and confidence intervals, or margins of error, should be provided for this curve for determining whether accuracy differences between systems are really statistically significant.
99
Models of large population recognition performance
Patrick J. Grother,P J. Phillips +1 more
- 27 Jun 2004
TL;DR: New binomial models of open- and closed-set identification recognition performance are presented, giving formulae for identification and false match rates as functions of database size, match rank and operating threshold.
81