Do AI Models "like" Black Dogs? Towards Exploring Perceptions of Dogs with Vision-Language Models

Marcelo Feighelstein; Einat Kovalyo; Jennifer Abrams; Sarah Elisabeth Byosiere; Anna Zamansky

doi:10.1145/3565995.3566022

Do AI Models "like" Black Dogs? Towards Exploring Perceptions of Dogs with Vision-Language Models

Marcelo Feighelstein, Einat Kovalyo, Jennifer Abrams, Sarah Elisabeth Byosiere, Anna Zamansky

Department of Information Systems

Research output: Chapter in Book/Report/Conference proceeding › Conference contribution › peer-review

Abstract

Large-scale, pretrained vision-language models such as OpenAI's CLIP are a game changer in Computer Vision due to their unprecedented gzero-shot' image classification capabilities. As they are pretrained on huge amounts of unsupervised web-scraped data, they suffer from inherent biases reflecting human perceptions, norms and beliefs. This position paper aims to highlight the potential of studying models such as CLIP in the context of human-Animal relationships, in particular for understanding human perceptions and preferences with respect to physical attributes of pets and their adoptability.

Original language	English
Title of host publication	ACI 2022 - 9th International Conference on Animal-Computer Interaction
Subtitle of host publication	Defining Tomorrow
Publisher	Association for Computing Machinery
ISBN (Electronic)	9781450398312
DOIs	https://doi.org/10.1145/3565995.3566022
State	Published - 5 Dec 2022
Event	9th International Conference on Animal-Computer Interaction: Defining Tomorrow, ACI 2022 - Newcastle upon Tyne, United Kingdom Duration: 5 Dec 2022 → 8 Dec 2022

Publication series

Name	ACM International Conference Proceeding Series

Conference

Conference	9th International Conference on Animal-Computer Interaction: Defining Tomorrow, ACI 2022
Country/Territory	United Kingdom
City	Newcastle upon Tyne
Period	5/12/22 → 8/12/22

Bibliographical note

Publisher Copyright:
© 2022 Owner/Author.

Keywords

animal-Assisted reading
animal-computer interaction
app design
child
support dog

ASJC Scopus subject areas

Human-Computer Interaction
Computer Networks and Communications
Computer Vision and Pattern Recognition
Software

Access to Document

10.1145/3565995.3566022

Cite this

Feighelstein, M., Kovalyo, E., Abrams, J., Byosiere, S. E., & Zamansky, A. (2022). Do AI Models "like" Black Dogs? Towards Exploring Perceptions of Dogs with Vision-Language Models. In ACI 2022 - 9th International Conference on Animal-Computer Interaction: Defining Tomorrow Article 7 (ACM International Conference Proceeding Series). Association for Computing Machinery. https://doi.org/10.1145/3565995.3566022

@inproceedings{eae96ccba37c49d9b5bdac0e58d9de72,

title = "Do AI Models {"}like{"} Black Dogs? Towards Exploring Perceptions of Dogs with Vision-Language Models",

abstract = "Large-scale, pretrained vision-language models such as OpenAI's CLIP are a game changer in Computer Vision due to their unprecedented gzero-shot' image classification capabilities. As they are pretrained on huge amounts of unsupervised web-scraped data, they suffer from inherent biases reflecting human perceptions, norms and beliefs. This position paper aims to highlight the potential of studying models such as CLIP in the context of human-Animal relationships, in particular for understanding human perceptions and preferences with respect to physical attributes of pets and their adoptability.",

keywords = "animal-Assisted reading, animal-computer interaction, app design, child, support dog",

author = "Marcelo Feighelstein and Einat Kovalyo and Jennifer Abrams and Byosiere, {Sarah Elisabeth} and Anna Zamansky",

note = "Publisher Copyright: {\textcopyright} 2022 Owner/Author.; 9th International Conference on Animal-Computer Interaction: Defining Tomorrow, ACI 2022 ; Conference date: 05-12-2022 Through 08-12-2022",

year = "2022",

month = dec,

day = "5",

doi = "10.1145/3565995.3566022",

language = "English",

series = "ACM International Conference Proceeding Series",

publisher = "Association for Computing Machinery",

booktitle = "ACI 2022 - 9th International Conference on Animal-Computer Interaction",

}

Feighelstein, M, Kovalyo, E, Abrams, J, Byosiere, SE & Zamansky, A 2022, Do AI Models "like" Black Dogs? Towards Exploring Perceptions of Dogs with Vision-Language Models. in ACI 2022 - 9th International Conference on Animal-Computer Interaction: Defining Tomorrow., 7, ACM International Conference Proceeding Series, Association for Computing Machinery, 9th International Conference on Animal-Computer Interaction: Defining Tomorrow, ACI 2022, Newcastle upon Tyne, United Kingdom, 5/12/22. https://doi.org/10.1145/3565995.3566022

Do AI Models "like" Black Dogs? Towards Exploring Perceptions of Dogs with Vision-Language Models. / Feighelstein, Marcelo; Kovalyo, Einat; Abrams, Jennifer et al.
ACI 2022 - 9th International Conference on Animal-Computer Interaction: Defining Tomorrow. Association for Computing Machinery, 2022. 7 (ACM International Conference Proceeding Series).

Research output: Chapter in Book/Report/Conference proceeding › Conference contribution › peer-review

TY - GEN

T1 - Do AI Models "like" Black Dogs? Towards Exploring Perceptions of Dogs with Vision-Language Models

AU - Feighelstein, Marcelo

AU - Kovalyo, Einat

AU - Abrams, Jennifer

AU - Byosiere, Sarah Elisabeth

AU - Zamansky, Anna

PY - 2022/12/5

Y1 - 2022/12/5

N2 - Large-scale, pretrained vision-language models such as OpenAI's CLIP are a game changer in Computer Vision due to their unprecedented gzero-shot' image classification capabilities. As they are pretrained on huge amounts of unsupervised web-scraped data, they suffer from inherent biases reflecting human perceptions, norms and beliefs. This position paper aims to highlight the potential of studying models such as CLIP in the context of human-Animal relationships, in particular for understanding human perceptions and preferences with respect to physical attributes of pets and their adoptability.

AB - Large-scale, pretrained vision-language models such as OpenAI's CLIP are a game changer in Computer Vision due to their unprecedented gzero-shot' image classification capabilities. As they are pretrained on huge amounts of unsupervised web-scraped data, they suffer from inherent biases reflecting human perceptions, norms and beliefs. This position paper aims to highlight the potential of studying models such as CLIP in the context of human-Animal relationships, in particular for understanding human perceptions and preferences with respect to physical attributes of pets and their adoptability.

KW - animal-Assisted reading

KW - animal-computer interaction

KW - app design

KW - child

KW - support dog

UR - http://www.scopus.com/inward/record.url?scp=85152193746&partnerID=8YFLogxK

U2 - 10.1145/3565995.3566022

DO - 10.1145/3565995.3566022

M3 - Conference contribution

AN - SCOPUS:85152193746

T3 - ACM International Conference Proceeding Series

BT - ACI 2022 - 9th International Conference on Animal-Computer Interaction

PB - Association for Computing Machinery

T2 - 9th International Conference on Animal-Computer Interaction: Defining Tomorrow, ACI 2022

Y2 - 5 December 2022 through 8 December 2022

ER -

Feighelstein M, Kovalyo E, Abrams J, Byosiere SE, Zamansky A. Do AI Models "like" Black Dogs? Towards Exploring Perceptions of Dogs with Vision-Language Models. In ACI 2022 - 9th International Conference on Animal-Computer Interaction: Defining Tomorrow. Association for Computing Machinery. 2022. 7. (ACM International Conference Proceeding Series). doi: 10.1145/3565995.3566022

Do AI Models "like" Black Dogs? Towards Exploring Perceptions of Dogs with Vision-Language Models

Abstract

Publication series

Conference

Bibliographical note

Keywords

ASJC Scopus subject areas

Access to Document

Other files and links

Fingerprint

Cite this