The National Institute of Standards and Technology (NIST) has announced the release of three new biometric databases aimed at helping accelerate the development of more secure and accurate biometric identification systems.
The databases — knows as SD 300, SD 301, and SD 302 — contain different types of biometric data collected at different times including fingerprints, facial photographs and iris scans.
“The data will help anyone who is interested in testing the error rates of biometric identification systems,” said NIST Computer Scientist Greg Fiumara.
NIST has said that the datasets have been stripped of any sensitive information and that all individuals involved consented to having their data included.
The first of the three databases, SD 300, contains a collection of hardcopy fingerprints taken from 900 old ink cards, and is meant to help manufacturers of modern fingerprint scanners determine if their systems can properly read older ink records.
SD 301 is the first multimodal dataset that NIST has ever released, containing both fingerprint and facial biometric markers that are linked together and can be used as one set of identifying data.
The third database, SD 302, contains scrubbed fingerprint data collected by a mixture of eight devices — some commercially available, others just prototypes — from an Intelligence Advanced Research Procter Activity funded competition, aimed at improving fingerprint capture technology, that NIST helped to carry out.
The new databases have user guides to help researchers integrate them into their own work, and NIST says it plans to offer more datasets in the future to further the development of biometric identification.
“[The datasets] represent the first in what is intended to be an expanding collection of biometric resources,” officials from NIST said in a statement.
December 19, 2019 — by Tony Bitzionis