<?xml version="1.0" encoding="UTF-8"?>
<modsCollection xmlns="http://www.loc.gov/mods/v3">
<mods ID="ruggeri-nozza-2023-multi">
    <titleInfo>
        <title>A Multi-dimensional study on Bias in Vision-Language models</title>
    </titleInfo>
    <name type="personal">
        <namePart type="given">Gabriele</namePart>
        <namePart type="family">Ruggeri</namePart>
        <role>
            <roleTerm authority="marcrelator" type="text">author</roleTerm>
        </role>
    </name>
    <name type="personal">
        <namePart type="given">Debora</namePart>
        <namePart type="family">Nozza</namePart>
        <role>
            <roleTerm authority="marcrelator" type="text">author</roleTerm>
        </role>
    </name>
    <originInfo>
        <dateIssued>2023-07</dateIssued>
    </originInfo>
    <typeOfResource>text</typeOfResource>
    <relatedItem type="host">
        <titleInfo>
            <title>Findings of the Association for Computational Linguistics: ACL 2023</title>
        </titleInfo>
        <name type="personal">
            <namePart type="given">Anna</namePart>
            <namePart type="family">Rogers</namePart>
            <role>
                <roleTerm authority="marcrelator" type="text">editor</roleTerm>
            </role>
        </name>
        <name type="personal">
            <namePart type="given">Jordan</namePart>
            <namePart type="family">Boyd-Graber</namePart>
            <role>
                <roleTerm authority="marcrelator" type="text">editor</roleTerm>
            </role>
        </name>
        <name type="personal">
            <namePart type="given">Naoaki</namePart>
            <namePart type="family">Okazaki</namePart>
            <role>
                <roleTerm authority="marcrelator" type="text">editor</roleTerm>
            </role>
        </name>
        <originInfo>
            <publisher>Association for Computational Linguistics</publisher>
            <place>
                <placeTerm type="text">Toronto, Canada</placeTerm>
            </place>
        </originInfo>
        <genre authority="marcgt">conference publication</genre>
    </relatedItem>
    <abstract>In recent years, joint Vision-Language (VL) models have increased in popularity and capability. Very few studies have attempted to investigate bias in VL models, even though it is a well-known issue in both individual modalities. This paper presents the first multi-dimensional analysis of bias in English VL models, focusing on gender, ethnicity, and age as dimensions. When subjects are input as images, pre-trained VL models complete a neutral template with a hurtful word 5% of the time, with higher percentages for female and young subjects. Bias presence in downstream models has been tested on Visual Question Answering. We developed a novel bias metric called the Vision-Language Association Test based on questions designed to elicit biased associations between stereotypical concepts and targets. Our findings demonstrate that pre-trained VL models contain biases that are perpetuated in downstream tasks.</abstract>
    <identifier type="citekey">ruggeri-nozza-2023-multi</identifier>
    <identifier type="doi">10.18653/v1/2023.findings-acl.403</identifier>
    <location>
        <url>https://aclanthology.org/2023.findings-acl.403</url>
    </location>
    <part>
        <date>2023-07</date>
        <extent unit="page">
            <start>6445</start>
            <end>6455</end>
        </extent>
    </part>
</mods>
</modsCollection>

%0 Conference Proceedings
%T A Multi-dimensional study on Bias in Vision-Language models
%A Ruggeri, Gabriele
%A Nozza, Debora
%Y Rogers, Anna
%Y Boyd-Graber, Jordan
%Y Okazaki, Naoaki
%S Findings of the Association for Computational Linguistics: ACL 2023
%D 2023
%8 July
%I Association for Computational Linguistics
%C Toronto, Canada
%F ruggeri-nozza-2023-multi
%X In recent years, joint Vision-Language (VL) models have increased in popularity and capability. Very few studies have attempted to investigate bias in VL models, even though it is a well-known issue in both individual modalities. This paper presents the first multi-dimensional analysis of bias in English VL models, focusing on gender, ethnicity, and age as dimensions. When subjects are input as images, pre-trained VL models complete a neutral template with a hurtful word 5% of the time, with higher percentages for female and young subjects. Bias presence in downstream models has been tested on Visual Question Answering. We developed a novel bias metric called the Vision-Language Association Test based on questions designed to elicit biased associations between stereotypical concepts and targets. Our findings demonstrate that pre-trained VL models contain biases that are perpetuated in downstream tasks.
%R 10.18653/v1/2023.findings-acl.403
%U https://aclanthology.org/2023.findings-acl.403
%U https://doi.org/10.18653/v1/2023.findings-acl.403
%P 6445-6455

Markdown (Informal)

[A Multi-dimensional study on Bias in Vision-Language models](https://aclanthology.org/2023.findings-acl.403) (Ruggeri & Nozza, Findings 2023)

• A Multi-dimensional study on Bias in Vision-Language models (Ruggeri & Nozza, Findings 2023)

ACL

• Gabriele Ruggeri and Debora Nozza. 2023. A Multi-dimensional study on Bias in Vision-Language models. In Findings of the Association for Computational Linguistics: ACL 2023, pages 6445–6455, Toronto, Canada. Association for Computational Linguistics.