English (United Kingdom)

https://curated-unify.zendy.io/wp-json/zendy-region/v1/featured_content/oa?rat=en

https://curated-unify.zendy.io/wp-json/zendy-region/v1/highlighted_journal/

Global: Zendy Plus annual

Presents the access of premium content as premium feature

Premium Content

Presents the keyphrase highlighting as premium feature

Keyphrase Highlighting

Presents the summarisation as premium feature

Summarisation

Insights

Presents the pdf analysis as premium feature

PDF Analysis

Presents the zaia usage as premium feature

ZAIA

Global: Zendy Plus monthly

Global: Zendy Tools annual

Global: Zendy Tools monthly

Global: Zendy Free Plan

Human face features can be used to determine individual identity as well asdemographic information like gender and race. However, the extent to whichblack-box commercial face recognition algorithms (CFRAs) use gender and racefeatures to determine identity is poorly understood despite increasingdeployments by government and industry. In this study, we quantified the degreeto which gender and race features influenced face recognition similarity scoresbetween different people, i.e. non-mated scores. We ran this study using fivedifferent CFRAs and a sample of 333 diverse test subjects. As a control, wecompared the behavior of these non-mated distributions to a commercial irisrecognition algorithm (CIRA). Confirming prior work, all CFRAs produced highersimilarity scores for people of the same gender and race, an effect known as"broad homogeneity". No such effect was observed for the CIRA. Next, we appliedprincipal components analysis (PCA) to similarity score matrices. We show thatsome principal components (PCs) of CFRAs cluster people by gender and race, butthe majority do not. Demographic clustering in the PCs accounted for only 10 %of the total CFRA score variance. No clustering was observed for the CIRA. Thisdemonstrates that, although CFRAs use some gender and race features toestablish identity, most features utilized by current CFRAs are unrelated togender and race, similar to the iris texture patterns utilized by the CIRA.Finally, reconstruction of similarity score matrices using only PCs that showedno demographic clustering reduced broad homogeneity effects, but also decreasedthe separation between mated and non-mated scores. This suggests it's possiblefor CFRAs to operate on features unrelated to gender and race, albeit withsomewhat lower recognition accuracy, but that this is not the currentcommercial practice.

Quantifying the Extent to Which Race and Gender Features Determine Identity in Commercial Face Recognition Algorithms