Documenting Computer Vision Datasets: An Invitation to Reflexive Data Practices

Kurzanzeige
dc.contributor.authorMiceli, Milagros
dc.contributor.authorYang, Tianling
dc.contributor.authorNaudts, Laurens
dc.contributor.authorSchüßler, Martin
dc.contributor.authorSerbanescu, Diana
dc.contributor.authorHanna, Alex
dc.date.accessioned2025-01-13T12:32:34Z
dc.date.available2025-01-13T12:32:34Z
dc.date.issued2021
dc.description.abstractIn industrial computer vision, discretionary decisions surrounding the production of image training data remain widely undocumented. Recent research taking issue with such opacity has proposed standardized processes for dataset documentation. In this paper, we expand this space of inquiry through fieldwork at two data processing companies and thirty interviews with data workers and computer vision practitioners. We identify four key issues that hinder the documentation of image datasets and the effective retrieval of production contexts. Finally, we propose reflexivity, understood as a collective consideration of social and intellectual factors that lead to praxis, as a necessary precondition for documentation. Reflexive documentation can help to expose the contexts, relations, routines, and power structures that shape data.
dc.identifier.citationMiceli, M., Yang, T., Naudts, L., Schüßler, M., Serbanescu, D., & Hanna, A. (2021). Documenting Computer Vision Datasets: An Invitation to Reflexive Data Practices. Proceedings of the 2021 ACM Conference on Fairness, Accountability, and Transparency, 161–172. https://doi.org/10.1145/3442188.3445880
dc.identifier.doi10.1145/3442188.3445880
dc.identifier.urihttps://www.weizenbaum-library.de/handle/id/802
dc.language.isoeng
dc.rightsopen access
dc.rights.urihttps://creativecommons.org/licenses/by/4.0/
dc.subjectdatasheets for datasetsen
dc.subjectdataset documentationen
dc.subjectreflexivityen
dc.subjectdata annotationen
dc.subjecttraining dataen
dc.subjecttransparencyen
dc.subjectaccountabilityen
dc.subjectauditsen
dc.subjectmachine learningen
dc.titleDocumenting Computer Vision Datasets: An Invitation to Reflexive Data Practices
dc.typeConferencePaper
dc.type.statuspublishedVersion
dcmi.typeText
dcterms.bibliographicCitation.urlhttps://doi.org/10.1145/3442188.3445880
local.researchgroupKritikalität KI-basierter Systeme
local.researchtopicVerantwortung – Vertrauen – Governance
Dateien
Originalbündel
Gerade angezeigt 1 - 1 von 1
Vorschaubild
Name:
Miceli-Yang-ea_2021_Documenting-Computer-Vision.pdf
Größe:
539.55 KB
Format:
Adobe Portable Document Format
Beschreibung:
Herunterladen
Sammlungen
Conference Papers