Multibiometric Score Normalization
Arun Ross (WVU) and Anil K. Jain (MSU)
Solidifying CITeR’s Liveness Core Competency
Stephanie Schuckers (Clarkson & WVU), Lawrence Hornak (WVU) and T. Norman (Orthopedics WVU)
Multimodal Biometric Systems: Phase II
Anil K. Jain (MSU)
Biometrics Business Case Study
Virginia Kleist and Richard Riley, Dept of Management, Man. Inf. Systems (WVU)
Geometric Coding of Biometric Images
Xin Li (WVU)
A Study of Various Methodologies for Error Estimation in Biometric Systems
Michael Schuckers (St. Lawrence University)
Socio‐Legal Assessment Study
Lisa Nelson (Pitt)
Statistical Basis of Multimodal Systems
Bojan Cukic and Harshinder Singh (WVU)
Summaries:
Multibiometric Score Normalization
Arun Ross (WVU) and Anil K. Jain (MSU)
The efforts will: (i) systematically study of the role of score normalization in multimodal matching performance, (ii) developing robust and efficient score normalization techniques and (iii) explore automatic template selection and update using clustering principles.
Solidifying CITeR’s Liveness Core Competency
Stephanie Schuckers (Clarkson & WVU), Lawrence Hornak (WVU) and T. Norman (Orthopedics WVU)
A unique specialized capability and skill set for performing spoofing, cadaver, and liveness testing and detection research has been developed. This project will build this capability; establish a broad‐based fingerprint liveness testing resource for members and the biometrics community, and advance liveness performance and research in select biometrics.
Multimodal Biometric Systems: Phase II
Anil K. Jain (MSU)
Three different issues in designing a multimodal biometric system are investigated: (i) combine face and iris biometrics to reduce failure to enroll rate and decrease FRR, (ii) combine multiple face recognition approaches to improve the face recognition performance, and (iii) operate a multibiometric system in an identification (cascade) mode.
Biometrics Business Case Study
Virginia Kleist and Richard Riley, Dept of Management, Man. Inf. Systems (WVU)
This proposed research applies prior work in information technology to the specific problem of measuring the cost benefit payoffs from specific biometric technologies. We will link three business issues (measuring the impact of information technology, the performance aspects of expenditures on technology and the specific return on IT investment at the strategic level) to the business case for biometrics investment.
Geometric Coding of Biometric Images
Xin Li (WVU)
The objective of this project is to develop new geometric coding and processing algorithms for biometric images (e.g. fingerprint, face, and iris), which will improve the matching performance of existing verification/recognition systems.
A Study of Various Methodologies for Error Estimation in Biometric Systems
Michael Schuckers (St. Lawrence University)
Several methods for estimating false accept and false reject rates have been developed and recognized for their potential. Doddington’s Rule of 30, Schuckers’ Beta‐binomial approach and the subset bootstrap create inferential intervals that allow the consumer to assess plausible values for the error rate of interest. The goal of this proposal is to compare the quality of estimation for each of these methods against identical data.
Socio‐Legal Assessment Study
Lisa Nelson (Pitt)
The proposed study is designed to set the groundwork for the generation of social scientific data on perceptions of biometrics and privacy in divergent settings as well as on the policy impact of biometric privacy legislation on consumer confidence and on the implementation of biometric technology.
Statistical Basis of Multimodal Systems
Bojan Cukic and Harshinder Singh (WVU)
This work seeks to provide a clear analytical justification for multi‐biometrics. Recently, Jain proposed the methodology to select weighting system that increases the probability of correct identification. We seek to generalize and extend this result based on the distributions describing genuine and imposter population.