Face recognition vendor test part 3:

Abstract

This is the third in a series of reports on ongoing face recognition vendor tests (FRVT) executed by the National Institute of Standards and Technology (NIST). The first two reports cover, respectively, the performance of one-to-one face recognition algorithms used for verification of asserted identities, and performance of one-to-many face recognition algorithms used for identification of individuals in photo data bases. This document extends those evaluations to document accuracy variations across demographic groups. MOTIVATION The recent expansion in the availability, capability, and use of face recognition has been accompanied by assertions that demographic dependencies could lead to accuracy variations and potential bias. A report from Georgetown University [14] work noted that prior studies [22], articulated sources of bias, described the potential impacts particularly in a policing context, and discussed policy and regulatory implications. Additionally, this work is motivated by studies of demographic effects in more recent face recognition [9, 16, 23] and gender estimation algorithms [5, 36]. AIMS AND SCOPE NIST has conducted tests to quantify demographic differences in contemporary face recognition algorithms. This report provides details about the recognition process, notes where demographic effects could occur, details specific performance metrics and analyses, gives empirical results, and recommends research into the mitigation of performance deficiencies. NIST intends this report to inform discussion and decisions about the accuracy, utility, and limitations of face recognition technologies. Its intended audience includes policy makers, face recognition algorithm developers, systems integrators, and managers of face recognition systems concerned with mitigation of risks implied by demographic differentials.

Is facial recognition too biased to be let loose?

“…But NIST also confirmed what Buolamwini and Gebru's gender-classification work suggested: most packages tended to be more accurate for white, male faces than for people of colour or for women 5 . In particular, faces classified in NIST's database as African American or Asian were 10-100 times more likely to be misidentified than those classified as white.…”

Convolutional neural net face recognition works in non-human-like ways

“…It is important to note that our data do not speak to the issue of whether CNNs show differential performance between races or sexes (see [ 17 ] for a recent survey). What we are reporting is that the networks are relatively blind to variations on these dimensions that humans regard as highly salient.…”

Opportunities and Challenges for Artificial Intelligence Applications in Infrastructure Management During the Anthropocene

“…The biases may be implicit or explicit, and can be the result of the individuals who wrote the algorithms or the data from which the algorithm was trained (Tomer, 2019). For example, facial recognition AI has been found to contain racial bias (Grother et al, 2019). It is unlikely that biases can fully be eliminated from Empathetic (and other) AI systems.…”