Skip to Content
No preview available

Actions

Download Analytics Citations

Export to: EndNote  |  Zotero  |  Mendeley

Collections

This file is not currently in any collections.

Understanding racial bias using facial phenotypes [other] Open Access

Recent works report disparate performance for intersectional racial groups across face recognition tasks: face verification and identification. However, the definition of racial groups has a significant impact on the underlying findings of such racial bias analysis. Previous studies define these groups based on either demographic information (e.g. African, Asian etc.) or skin colour in a shallow categorisation of lighter or darker skins. Albeit, both of group definitions have disadvantages for both bias investigation and the design of subsequent counter-bias solutions. Exposing protected attributes may impact on an individual's privacy or raise issues of consent by the use of such attributes. Furthermore, analysis based solely on skin tone may cause deficient interpretations of bias as racial characteristics have several traits that extend beyond skin tone alone. This study proposes a new evaluation strategy using facial phenotype attributes to assess racial bias in face recognition tasks. Using this methodology, we develop new categorical test cases and investigate the individual influence of those attributes on racial bias. Furthermore, we contribute corresponding attributes labels for face recognition tasks: RFW for face verification and VGGFace2 for face identification.

Descriptions

Resource type
Other
Contributors
Creator: Yucer, Seyma 1
Contact person: Yucer, Seyma 1
Editor: Breckon, Toby 1
1 Durham University,UK
Funder
Research methods
Other description
Deposit consists of metadata files of two datasets and one machine learning model trained parameters. Code published on GitHub:  https://github.com/seymayucer/FacialPhenotypes

Keyword
Facial phenotypes
Labels
Pre-trained models
Subject
Face perception--Racial bias
Location
Durham, England, United Kingdom
Language
English
Cited in
Identifier
ark:/32150/r2hm50tr746
doi:10.15128/r2hm50tr746
Rights
Creative Commons Attribution 4.0 International (CC BY)

Publisher
Durham University
Date Created

File Details

Depositor
S. Yucer Tektas
Date Uploaded
Date Modified
20 April 2021, 16:04:09
Audit Status
Audits have not yet been run on this file.
Characterization
File format: zip (ZIP Format)
Mime type: application/zip
File size: 242817654
Last modified: 2021:04:20 13:52:28+01:00
Filename: model_and_labels.zip
Original checksum: 2774ea5e7a10aab0b9db9eb6cdab8450
Activity of users you follow
User Activity Date
User N. Syrotiuk has updated Understanding racial bias using facial phenotypes [other] 3 months ago
User N. Syrotiuk has updated Understanding racial bias using facial phenotypes [other] 3 months ago
User N. Syrotiuk has updated Understanding Racial Bias using Facial Phenotypes [other] 3 months ago
User N. Syrotiuk has updated Understanding Racial Bias using Facial Phenotypes [other] 3 months ago
User N. Syrotiuk has updated Understanding Racial Bias using Facial Phenotypes [dataset] 3 months ago