Why Temporal Persistence of Biometric Features, as Assessed by the Intraclass Correlation Coefficient, Is So Valuable for Classification Performance.

Lee Friedman,Hal S Stern,Larry R Price,Oleg V Komogortsev

doi:10.3390/s20164555

Lee Friedman, Hal S Stern + Show 2 more

Open Access

https://doi.org/10.3390/s20164555

Copy DOI

Abstract

It is generally accepted that relatively more permanent (i.e., more temporally persistent) traits are more valuable for biometric performance than less permanent traits. Although this finding is intuitive, there is no current work identifying exactly where in the biometric analysis temporal persistence makes a difference. In this paper, we answer this question. In a recent report, we introduced the intraclass correlation coefficient (ICC) as an index of temporal persistence for such features. Here, we present a novel approach using synthetic features to study which aspects of a biometric identification study are influenced by the temporal persistence of features. What we show is that using more temporally persistent features produces effects on the similarity score distributions that explain why this quality is so key to biometric performance. The results identified with the synthetic data are largely reinforced by an analysis of two datasets, one based on eye-movements and one based on gait. There was one difference between the synthetic and real data, related to the intercorrelation of features in real data. Removing these intercorrelations for real datasets with a decorrelation step produced results which were very similar to that obtained with synthetic features.

Highlights

In a recent report [1], we introduced the intraclass correlation coefficient (ICC) as an index of temporal persistence of single biometric features
The main findings of Experiment 1 are that using synthetic features with higher temporal persistence for biometric analysis produces lower equal error rate (EER) values, i.e., improved biometric performance
These improvements are due to an increased median and a decreased interquartile ranges (IQR) of the genuine similarity score distributions with no change in the impostor distributions

Summary

Introduction

In a recent report [1], we introduced the intraclass correlation coefficient (ICC) as an index of temporal persistence (e.g., stability, permanence) of single biometric features. (We understand that this limitation would mean that this analysis might not be directly applicable to several important biometric modalities, including fingerprints and iris scans. Other modalities such as face recognition can meet these criteria [2]. The analysis applies to brain structure [1] We think that this approach will be applicable to most physiological (EEG, ECG, etc.) and behavioral modalities.) The ICC can only be calculated if each subject is tested on 2 or more occasions.

Objectives

Results

Conclusion