Patrick Grother is a scientist at the National Institute of Standards in Technology (NIST) responsible for biometric algorithm evaluation, and biometric performance testing standardization. He leads the Face Recognition Vendor Tests which constitute the world’s largest independent public tests of face recognition algorithms. These give quantitative support to developers, end-users and policy makers faced with algorithm selection, performance adequacy assessment, procurement specification. His current interests relate biometric failure analysis, image quality, demographic effects, and scalability.  Patrick co-chairs NIST’s biannual International Face Performance Conference (IFPC) on measurement, metrics and certification. He assists several US Government agencies in biometrics performance assessment and standardization. Since 2018, he has served as the chairman of the ISO/IEC/JTC 1 Subcommittee 37 on Biometrics where he has edited six performance testing and data interchange standards. He received the IEC 1906 Award in 2009, the ANSI Lohse IT Medal in 2013, and U.S. Department of Commerce Gold Medals in 2007, 2012 and 2021.

 

SESSION DESCRIPTION
Facial Recognition: Let’s Look at the Science

In a 2019 Business Insider article the headline read “Facial recognition is almost perfectly accurate.” Unlike many years before it seems the technology is starting to live up to the hype. A recent market research report estimated the global facial recognition market size is expected to grow from USD 3.8 billion in 2020 to USD 8.5 billion by 2025. Manufacturers are rushing to capitalize on the demand and are filling potential buyers’ email inboxes with marketing blurbs for cool-named products powered by lots of abbreviations.

To help you decide, the National Institute of Standards and Technology (NIST) Face Recognition Vendor Testing Program (FRVT) was established in 2000 to provide independent evaluations of both prototype and commercially available facial recognition algorithms. In this talk Patrick Grother, NIST Scientist, will explain how they evaluate algorithms. He will also give the findings or research they have conducted on the effectiveness of the technology when people wear face masks, demographics, and methods people use to trick algorithms. Patrick will end his talk with an update on recent developments in facial recognition.

 

LINKS OF INTEREST:
Patrick on 60 Minutes
Biometric update
Study with face masks