Abstract:

Many large-scale biometric deployments (such as UIDAI, US-VISIT, or EU-VIS) mandate measuring and reporting of quality scores. If a low quality value is predictive of recognition failure (primarily a false negative, but possibly a false positive too), then a new sample can be collected while the subject is still present. Therefore, quality analysis is most helpful when the measures reflect the performance sensitivities of one or more target biometric matchers.  NIST addressed this problem in August 2004 when it issued the NIST Fingerprint Image Quality (NFIQ) algorithm. Since then NIST has been considering how quality measures should be evaluated, definition and development of quality measures for other biometrics, and the wider use of such measures. In addition NIST is active in the SC37 standardization activities on biometric quality and sample conformance.

This talk will dive into details about implementation plans, obstacles encountered, lessons learned, and intended impact of NIST biometric quality program.


Bio:

Elham Tabassi is a researcher at National Institute of Standards and Technology working on various biometric research projects including biometric sample quality, fusion, and performance assessment.  She is the principal architect of NIST Fingerprint Image Quality (NFIQ) which has become the defacto standard for measuring fingerprint image quality and is currently deployed in some of US Government and EU biometric applications.

She is the editor of several international biometric standards including biometric sample quality; data interchange formats for biometric data, and conformance to data interchange formats. She received ANSI’s 2012 Next Generation Award, the Department of Commerce Gold Medal in 2003 for her work on biometric system performance assessment, and two Bronze Medals in 2007 and 2010 for her works on biometric sample quality and development, evaluation, and standardization of iris image records in biometric systems respectively. Her research interests are in biometric sample quality, pattern recognition, data mining, and signal processing.  

Host:

Dr. Anil Jain