Towards Reliable Assessments of Demographic Disparities in Multi-Label Image Classifiers
Disaggregated performance metrics across demographic groups are a hallmark of fairness assessments in computer vision. These metrics successfully incentivized performance improvements on person-centric tasks such as face analysis and are used to understand risks of modern models. However, there is a...
Saved in:
Published in | arXiv.org |
---|---|
Main Authors | , , , , , , |
Format | Paper |
Language | English |
Published |
Ithaca
Cornell University Library, arXiv.org
16.02.2023
|
Subjects | |
Online Access | Get full text |
Cover
Loading…
Abstract | Disaggregated performance metrics across demographic groups are a hallmark of fairness assessments in computer vision. These metrics successfully incentivized performance improvements on person-centric tasks such as face analysis and are used to understand risks of modern models. However, there is a lack of discussion on the vulnerabilities of these measurements for more complex computer vision tasks. In this paper, we consider multi-label image classification and, specifically, object categorization tasks. First, we highlight design choices and trade-offs for measurement that involve more nuance than discussed in prior computer vision literature. These challenges are related to the necessary scale of data, definition of groups for images, choice of metric, and dataset imbalances. Next, through two case studies using modern vision models, we demonstrate that naive implementations of these assessments are brittle. We identify several design choices that look merely like implementation details but significantly impact the conclusions of assessments, both in terms of magnitude and direction (on which group the classifiers work best) of disparities. Based on ablation studies, we propose some recommendations to increase the reliability of these assessments. Finally, through a qualitative analysis we find that concepts with large disparities tend to have varying definitions and representations between groups, with inconsistencies across datasets and annotators. While this result suggests avenues for mitigation through more consistent data collection, it also highlights that ambiguous label definitions remain a challenge when performing model assessments. Vision models are expanding and becoming more ubiquitous; it is even more important that our disparity assessments accurately reflect the true performance of models. |
---|---|
AbstractList | Disaggregated performance metrics across demographic groups are a hallmark of fairness assessments in computer vision. These metrics successfully incentivized performance improvements on person-centric tasks such as face analysis and are used to understand risks of modern models. However, there is a lack of discussion on the vulnerabilities of these measurements for more complex computer vision tasks. In this paper, we consider multi-label image classification and, specifically, object categorization tasks. First, we highlight design choices and trade-offs for measurement that involve more nuance than discussed in prior computer vision literature. These challenges are related to the necessary scale of data, definition of groups for images, choice of metric, and dataset imbalances. Next, through two case studies using modern vision models, we demonstrate that naive implementations of these assessments are brittle. We identify several design choices that look merely like implementation details but significantly impact the conclusions of assessments, both in terms of magnitude and direction (on which group the classifiers work best) of disparities. Based on ablation studies, we propose some recommendations to increase the reliability of these assessments. Finally, through a qualitative analysis we find that concepts with large disparities tend to have varying definitions and representations between groups, with inconsistencies across datasets and annotators. While this result suggests avenues for mitigation through more consistent data collection, it also highlights that ambiguous label definitions remain a challenge when performing model assessments. Vision models are expanding and becoming more ubiquitous; it is even more important that our disparity assessments accurately reflect the true performance of models. |
Author | Usunier, Nicolas Ventura, Denisse Chern, Bobbie Ross, Candace Hall, Melissa Kulkarni, Harshad Gustafson, Laura |
Author_xml | – sequence: 1 givenname: Melissa surname: Hall fullname: Hall, Melissa – sequence: 2 givenname: Bobbie surname: Chern fullname: Chern, Bobbie – sequence: 3 givenname: Laura surname: Gustafson fullname: Gustafson, Laura – sequence: 4 givenname: Denisse surname: Ventura fullname: Ventura, Denisse – sequence: 5 givenname: Harshad surname: Kulkarni fullname: Kulkarni, Harshad – sequence: 6 givenname: Candace surname: Ross fullname: Ross, Candace – sequence: 7 givenname: Nicolas surname: Usunier fullname: Usunier, Nicolas |
BookMark | eNqNyr0KwjAUQOEgCv6-wwXnQk1aq6O0ioIu0sGtpHqrV9Kk5rb4-jr4AE5n-M5Y9K2z2BMjqdQiWEVSDsWM-RmGoVwmMo7VSFxy99b-xnBGQ7o0CBtmZK7Rtgyuggxrd_e6edAVMuJGe2oJGcjCqTMtBUddooFDre8IqdHMVBF6nopBpQ3j7NeJmO-2eboPGu9eHXJbPF3n7ZcKmSSrhYrWy7X67_oAHEtDyw |
ContentType | Paper |
Copyright | 2023. This work is published under http://arxiv.org/licenses/nonexclusive-distrib/1.0/ (the “License”). Notwithstanding the ProQuest Terms and Conditions, you may use this content in accordance with the terms of the License. |
Copyright_xml | – notice: 2023. This work is published under http://arxiv.org/licenses/nonexclusive-distrib/1.0/ (the “License”). Notwithstanding the ProQuest Terms and Conditions, you may use this content in accordance with the terms of the License. |
DBID | 8FE 8FG ABJCF ABUWG AFKRA AZQEC BENPR BGLVJ CCPQU DWQXO HCIFZ L6V M7S PIMPY PQEST PQQKQ PQUKI PRINS PTHSS |
DatabaseName | ProQuest SciTech Collection ProQuest Technology Collection Materials Science & Engineering Collection ProQuest Central (Alumni) ProQuest Central ProQuest Central Essentials ProQuest Central Technology Collection ProQuest One Community College ProQuest Central Korea SciTech Premium Collection ProQuest Engineering Collection Engineering Database Publicly Available Content Database ProQuest One Academic Eastern Edition (DO NOT USE) ProQuest One Academic ProQuest One Academic UKI Edition ProQuest Central China Engineering Collection |
DatabaseTitle | Publicly Available Content Database Engineering Database Technology Collection ProQuest Central Essentials ProQuest One Academic Eastern Edition ProQuest Central (Alumni Edition) SciTech Premium Collection ProQuest One Community College ProQuest Technology Collection ProQuest SciTech Collection ProQuest Central China ProQuest Central ProQuest Engineering Collection ProQuest One Academic UKI Edition ProQuest Central Korea Materials Science & Engineering Collection ProQuest One Academic Engineering Collection |
DatabaseTitleList | Publicly Available Content Database |
Database_xml | – sequence: 1 dbid: 8FG name: ProQuest Technology Collection url: https://search.proquest.com/technologycollection1 sourceTypes: Aggregation Database |
DeliveryMethod | fulltext_linktorsrc |
Discipline | Physics |
EISSN | 2331-8422 |
Genre | Working Paper/Pre-Print |
GroupedDBID | 8FE 8FG ABJCF ABUWG AFKRA ALMA_UNASSIGNED_HOLDINGS AZQEC BENPR BGLVJ CCPQU DWQXO FRJ HCIFZ L6V M7S M~E PIMPY PQEST PQQKQ PQUKI PRINS PTHSS |
ID | FETCH-proquest_journals_27781349693 |
IEDL.DBID | BENPR |
IngestDate | Thu Oct 10 16:20:06 EDT 2024 |
IsOpenAccess | true |
IsPeerReviewed | false |
IsScholarly | false |
Language | English |
LinkModel | DirectLink |
MergedId | FETCHMERGED-proquest_journals_27781349693 |
OpenAccessLink | https://www.proquest.com/docview/2778134969?pq-origsite=%requestingapplication% |
PQID | 2778134969 |
PQPubID | 2050157 |
ParticipantIDs | proquest_journals_2778134969 |
PublicationCentury | 2000 |
PublicationDate | 20230216 |
PublicationDateYYYYMMDD | 2023-02-16 |
PublicationDate_xml | – month: 02 year: 2023 text: 20230216 day: 16 |
PublicationDecade | 2020 |
PublicationPlace | Ithaca |
PublicationPlace_xml | – name: Ithaca |
PublicationTitle | arXiv.org |
PublicationYear | 2023 |
Publisher | Cornell University Library, arXiv.org |
Publisher_xml | – name: Cornell University Library, arXiv.org |
SSID | ssj0002672553 |
Score | 3.4578073 |
SecondaryResourceType | preprint |
Snippet | Disaggregated performance metrics across demographic groups are a hallmark of fairness assessments in computer vision. These metrics successfully incentivized... |
SourceID | proquest |
SourceType | Aggregation Database |
SubjectTerms | Ablation Assessments Classifiers Computer vision Data collection Datasets Demographics Image classification Performance measurement Qualitative analysis Reliability analysis Task complexity |
Title | Towards Reliable Assessments of Demographic Disparities in Multi-Label Image Classifiers |
URI | https://www.proquest.com/docview/2778134969 |
hasFullText | 1 |
inHoldings | 1 |
isFullTextHit | |
isPrint | |
link | http://utb.summon.serialssolutions.com/2.0.0/link/0/eLvHCXMwfV1LSwMxEB5sF8GbT7TWEtBrcJ_ZzUnU7lrFliIV9laSTQIL9mG3Xv3tZtZtPQg9hkBIQvgmM_PNfAA3ATNamUTQSEhFw4S7VCQRp5FSxshACK7QURyO2OA9fMmjvAm4VQ2tcoOJNVCrRYEx8ls_jhNspcf43fKTomoUZlcbCY0WOL4XYprWeUhH47dtlMVnsf0zB_-AtrYe2SE4Y7HUqyPY0_Nj2K9Jl0V1Avmk5qxWBHnBWMJE7rd9MiuyMKSvZ78dpcuC9MsKBQOx_ykp56QunKWvQuoP8jyzoEBqecvSoLT1KVxn6eRxQDfbmTZPppr-HTA4g7b1_fU5EC5d5QWMKxdFomNXeJG07q91EqRhIS8uoLtrpc7u6Us4QPV0JCF7rAvt9epLX1kbu5Y9aCXZU6-5Tjsafqc_4eCIkg |
link.rule.ids | 786,790,12792,21416,33406,33777,43633,43838 |
linkProvider | ProQuest |
linkToHtml | http://utb.summon.serialssolutions.com/2.0.0/link/0/eLvHCXMwfV1NSwMxEB20RfTmJ35UDeg1uG12s5uTiHXd6rZ4qNBbSTYJLPTLbv3_ZuK2HoSeAyEJw5vM5OU9gHvGrdE2kTSSStMwEQGVSSRopLW1ikkpNBaK_QHPPsO3UTSqG25VTatcY6IHaj0vsEf-0InjBKX0uHhcfFF0jcLX1dpCYxeaIeMM4zxJXzc9lg6P3Y2Z_YNZnzvSQ2h-yIVZHsGOmR3DnqdcFtUJjIaesVoRZAXjBybytFHJrMjckq6Z_upJlwXplhXaBaL6KSlnxH-bpblUZkJ6UwcJxJtblhaNrU_hLn0ZPmd0vZxxHTDV-G977AwarvI350CECnSbcaEDtIiOA9mOlCt-XYmgLA9FcQGtbTNdbh--hf1s2M_HeW_wfgUH6KOOdOQ2b0Fjtfw21y7brtSNP9IfsfeIAg |
openUrl | ctx_ver=Z39.88-2004&ctx_enc=info%3Aofi%2Fenc%3AUTF-8&rfr_id=info%3Asid%2Fsummon.serialssolutions.com&rft_val_fmt=info%3Aofi%2Ffmt%3Akev%3Amtx%3Ajournal&rft.genre=article&rft.atitle=Towards+Reliable+Assessments+of+Demographic+Disparities+in+Multi-Label+Image+Classifiers&rft.jtitle=arXiv.org&rft.au=Hall%2C+Melissa&rft.au=Chern%2C+Bobbie&rft.au=Gustafson%2C+Laura&rft.au=Ventura%2C+Denisse&rft.date=2023-02-16&rft.pub=Cornell+University+Library%2C+arXiv.org&rft.eissn=2331-8422 |