2002 Projects

Study of Liveness Detection for Biometric Devices
Stephanie Schuckers, Lawrence Hornak and Timothy Norman Orthopedics (WVU)

Multimodal Biometric System
Anil K. Jain (MSU)

Estimation Study
Michael Schuckers (WVU) and James Wayman (San Jose State University)

Template Aging Study
Michael Schuckers (WVU) and James Wayman (San Jose State University)

Issues in Large Scale Biometric Authentication Infrastructure
Bojan Cukic (WVU)

Summaries:

Study of Liveness Detection for Biometric Devices

Stephanie Schuckers, Lawrence Hornak and Timothy Norman Orthopedics (WVU)

This research investigates determining liveness directly from COTs biometric sensor signals. Previous work performed liveness detection through quantification of the temporal finger perspiration pattern. In this work we (1) Select and secure a broad range of biometric fingerprint systems for liveness testing with IAB input, (2) Undertake liveness/spoof test, explore available physiological info, evaluate algorithmic approaches, and (3) Undertake further study of the perspiration algorithm on fingerprint devices.

Multimodal Biometric System

Anil K. Jain (MSU)

The objective of this project is to design a robust multimodal biometric system that uses a combination of features from fingerprint, hand geometry and face. Data corresponding to three biometric indicators‐ fingerprint, hand geometry and face ‐ will be collected from a number of subjects. Images pertaining to each modality will be processed to extract a feature vector. The techniques developed in this project will be used to optimally combine the information acquired from these three modalities in order to verify a subject’s claimed identity. The False Reject Rate (FRR) and False Accept Rate (FAR) will be used to evaluate the performance of the system.

Estimation Study

Michael Schuckers (WVU) and James Wayman (San Jose State University)

The goal of this study is to develop methodologies to estimate the variability in false match and false non‐match rates for biometric testing. Several methodologies are under development by the participants of this project to develop techniques for estimating false match and false non‐match rates. This study continues and combines the work that has been done independently. Specifically, we convert the work that has been done into an appropriate methodology for sample size calculation.

Template Aging Study

Michael Schuckers (WVU) and James Wayman (San Jose State University)

The goal of this study is to follow a group of individuals repeatedly over time to determine how large an effect there is due to the aging of templates. This test will be done for a variety of modalities based on IAB input. A general statistical framework to begin to model and understand template aging is sought.

Issues in Large Scale Biometric Authentication Infrastructure

Bojan Cukic (WVU)

Biometric technologies can improve the overall assurance of an information system by incorporating the uniqueness of personal biometric signatures into the security and safety management. The downside of using biometric signature is the problem of scalability. Digital biometric identifiers present at (transferred to) multiple sites can be intercepted, pooled, analyzed and, generally, misused. We will investigate system design principles that minimize these downsides of biometric technology‐based authentication protocols.