Pixels versus Privacy: Leveraging Vision-Language Models for Sensitive Information Extraction
S. Schultenkämper, F.S. Bäumer, International Journal On Advances in Security 17 (2024) 1–10.
Download (ext.)
Artikel
| Veröffentlicht
| Englisch
Autor*in
Abstract
Threats to user privacy in Web 2.0 are abundant and can arise from various sources, including texts, geoinformation,
images, videos, or combinations of these. To alert users of potential threats, it is crucial to gather all relevant information. However, aggregating user-specific information from various web platforms, including social networks, can be challenging due to the vast amount of data available, as well as issues with data quality and the numerous possible variants. This paper examines the capability of current Vision-Language Models to accurately identify relevant image data and extract sensitive information. To accomplish this, we developed our own dataset with diverse
expressions for privacy attributes, based on the VISPR dataset. Furthermore, we address the challenge of synthetic images of people and its impact on our approach. Our findings suggest that these models are effective in pre-selecting relevant images, but there are limitations in information extraction.
Keywords—Computer Vision; Privacy; Social Networks
Erscheinungsjahr
Zeitschriftentitel
International Journal On Advances in Security
Band
17
Zeitschriftennummer
1+2
Seite
1-10
ISSN
FH-PUB-ID
Zitieren
Schultenkämper, Sergej ; Bäumer, Frederik S.: Pixels versus Privacy: Leveraging Vision-Language Models for Sensitive Information Extraction. In: International Journal On Advances in Security Bd. 17, IARIA XPS Press (2024), Nr. 1+2, S. 1–10
Schultenkämper S, Bäumer FS. Pixels versus Privacy: Leveraging Vision-Language Models for Sensitive Information Extraction. International Journal On Advances in Security. 2024;17(1+2):1-10.
Schultenkämper, S., & Bäumer, F. S. (2024). Pixels versus Privacy: Leveraging Vision-Language Models for Sensitive Information Extraction. International Journal On Advances in Security, 17(1+2), 1–10.
@article{Schultenkämper_Bäumer_2024, title={Pixels versus Privacy: Leveraging Vision-Language Models for Sensitive Information Extraction}, volume={17}, number={1+2}, journal={International Journal On Advances in Security}, publisher={IARIA XPS Press}, author={Schultenkämper, Sergej and Bäumer, Frederik S.}, year={2024}, pages={1–10} }
Schultenkämper, Sergej, and Frederik S. Bäumer. “Pixels versus Privacy: Leveraging Vision-Language Models for Sensitive Information Extraction.” International Journal On Advances in Security 17, no. 1+2 (2024): 1–10.
S. Schultenkämper and F. S. Bäumer, “Pixels versus Privacy: Leveraging Vision-Language Models for Sensitive Information Extraction,” International Journal On Advances in Security, vol. 17, no. 1+2, pp. 1–10, 2024.
Schultenkämper, Sergej, and Frederik S. Bäumer. “Pixels versus Privacy: Leveraging Vision-Language Models for Sensitive Information Extraction.” International Journal On Advances in Security, vol. 17, no. 1+2, IARIA XPS Press, 2024, pp. 1–10.
Link(s) zu Volltext(en)
Access Level
Closed Access
Google Scholar