Assessing Privacy Preservation and Utility in Online Vision-Language Models
Fuente:
arXiv
Enregistré dans:
| Auteurs principaux: | , , , , |
|---|---|
| Format: | Preprint |
| Publié: |
2026
|
| Sujets: | |
| Accès en ligne: | |
| Tags: |
Ajouter un tag
Pas de tags, Soyez le premier à ajouter un tag!
|
| _version_ | 1866918439396835328 |
|---|---|
| author | Chaudhari, Karmesh Siddharam Zhu, Youxiang Feng, Amy Liang, Xiaohui Zhang, Honggang |
| author_facet | Chaudhari, Karmesh Siddharam Zhu, Youxiang Feng, Amy Liang, Xiaohui Zhang, Honggang |
| contents | The increasing use of Online Vision Language Models (OVLMs) for processing images has introduced significant privacy risks, as individuals frequently upload images for various utilities, unaware of the potential for privacy violations. Images contain relationships that relate to Personally Identifiable Information (PII), where even seemingly harmless details can indirectly reveal sensitive information through surrounding clues. This paper explores the critical issue of PII disclosure in images uploaded to OVLMs and its implications for user privacy. We investigate how the extraction of contextual relationships from images can lead to direct (explicit) or indirect (implicit) exposure of PII, significantly compromising personal privacy. Furthermore, we propose methods to protect privacy while preserving the intended utility of the images in Vision Language Model (VLM)-based applications. Our evaluation demonstrates the efficacy of these techniques, highlighting the delicate balance between maintaining utility and protecting privacy in online image processing environments. Index Terms-Personally Identifiable Information (PII), Privacy, Utility, privacy concerns, sensitive information |
| format | Preprint |
| id |
arxiv_https___arxiv_org_abs_2604_09695 |
| institution | arXiv |
| publishDate | 2026 |
| record_format | arxiv |
| spellingShingle | Assessing Privacy Preservation and Utility in Online Vision-Language Models Chaudhari, Karmesh Siddharam Zhu, Youxiang Feng, Amy Liang, Xiaohui Zhang, Honggang Computer Vision and Pattern Recognition Artificial Intelligence The increasing use of Online Vision Language Models (OVLMs) for processing images has introduced significant privacy risks, as individuals frequently upload images for various utilities, unaware of the potential for privacy violations. Images contain relationships that relate to Personally Identifiable Information (PII), where even seemingly harmless details can indirectly reveal sensitive information through surrounding clues. This paper explores the critical issue of PII disclosure in images uploaded to OVLMs and its implications for user privacy. We investigate how the extraction of contextual relationships from images can lead to direct (explicit) or indirect (implicit) exposure of PII, significantly compromising personal privacy. Furthermore, we propose methods to protect privacy while preserving the intended utility of the images in Vision Language Model (VLM)-based applications. Our evaluation demonstrates the efficacy of these techniques, highlighting the delicate balance between maintaining utility and protecting privacy in online image processing environments. Index Terms-Personally Identifiable Information (PII), Privacy, Utility, privacy concerns, sensitive information |
| title | Assessing Privacy Preservation and Utility in Online Vision-Language Models |
| topic | Computer Vision and Pattern Recognition Artificial Intelligence |
| url | https://arxiv.org/abs/2604.09695 |