Lessons From Collecting a Million Biometric Samples

• NCJ Number: 252378
• Journal: Image and Vision Computing Volume: 58 Dated: February 2017 Pages: 96-107
• Author(s): P. Jonathon Phillips; Paatrick J. Flynn; Kevin W. Bowyer
• Date Published: February 2017
• Length: 12 pages
• Grant Number(s): 2015-IJ-CX-K014
• Agencies: NIJ-Sponsored
• Publication Type: Research (Applied/Empirical)
• Award: Design and Implementation of Forensic Facial Identification Experts Test

Annotation

This review of a 10-year biometric sampling effort summarizes the design and execution of data collections, identifies key challenges, and discusses some lessons learned.

Abstract

Over the past decade, independent evaluations have become common in many areas of experimental computer science, including face and gesture recognition. A key attribute of many successful independent evaluations is a curated data set. Desired aspects associated with these data sets include appropriateness to the experimental design, a corpus size large enough to allow statistically rigorous characterization of results, and the availability of comprehensive meta-data that enable inferences to be made on various data set attributes. (Publisher abstract modified)