Evaluating Precise Geolocation Inference Capabilities of Vision Language Models

Fuente: arXiv
Saved in:
Bibliographic Details
Main Authors: Jay, Neel, Nguyen, Hieu Minh, Hoang, Trung Dung, Haimes, Jacob
Format: Preprint
Published: 2025
Subjects:
Online Access:
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1866913700423663616
author Jay, Neel
Nguyen, Hieu Minh
Hoang, Trung Dung
Haimes, Jacob
author_facet Jay, Neel
Nguyen, Hieu Minh
Hoang, Trung Dung
Haimes, Jacob
contents The prevalence of Vision-Language Models (VLMs) raises important questions about privacy in an era where visual information is increasingly available. While foundation VLMs demonstrate broad knowledge and learned capabilities, we specifically investigate their ability to infer geographic location from previously unseen image data. This paper introduces a benchmark dataset collected from Google Street View that represents its global distribution of coverage. Foundation models are evaluated on single-image geolocation inference, with many achieving median distance errors of <300 km. We further evaluate VLM "agents" with access to supplemental tools, observing up to a 30.6% decrease in distance error. Our findings establish that modern foundation VLMs can act as powerful image geolocation tools, without being specifically trained for this task. When coupled with increasing accessibility of these models, our findings have greater implications for online privacy. We discuss these risks, as well as future work in this area.
format Preprint
id arxiv_https___arxiv_org_abs_2502_14412
institution arXiv
publishDate 2025
record_format arxiv
spellingShingle Evaluating Precise Geolocation Inference Capabilities of Vision Language Models
Jay, Neel
Nguyen, Hieu Minh
Hoang, Trung Dung
Haimes, Jacob
Computer Vision and Pattern Recognition
Cryptography and Security
Machine Learning
The prevalence of Vision-Language Models (VLMs) raises important questions about privacy in an era where visual information is increasingly available. While foundation VLMs demonstrate broad knowledge and learned capabilities, we specifically investigate their ability to infer geographic location from previously unseen image data. This paper introduces a benchmark dataset collected from Google Street View that represents its global distribution of coverage. Foundation models are evaluated on single-image geolocation inference, with many achieving median distance errors of <300 km. We further evaluate VLM "agents" with access to supplemental tools, observing up to a 30.6% decrease in distance error. Our findings establish that modern foundation VLMs can act as powerful image geolocation tools, without being specifically trained for this task. When coupled with increasing accessibility of these models, our findings have greater implications for online privacy. We discuss these risks, as well as future work in this area.
title Evaluating Precise Geolocation Inference Capabilities of Vision Language Models
topic Computer Vision and Pattern Recognition
Cryptography and Security
Machine Learning
url https://arxiv.org/abs/2502.14412