Task-Agnostic Attacks Against Vision Foundation Models

Fuente: arXiv
Saved in:
Bibliographic Details
Main Authors: Pulfer, Brian, Belousov, Yury, Kinakh, Vitaliy, Furon, Teddy, Voloshynovskiy, Slava
Format: Preprint
Published: 2025
Subjects:
Online Access:
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1866916645024301056
author Pulfer, Brian
Belousov, Yury
Kinakh, Vitaliy
Furon, Teddy
Voloshynovskiy, Slava
author_facet Pulfer, Brian
Belousov, Yury
Kinakh, Vitaliy
Furon, Teddy
Voloshynovskiy, Slava
contents The study of security in machine learning mainly focuses on downstream task-specific attacks, where the adversarial example is obtained by optimizing a loss function specific to the downstream task. At the same time, it has become standard practice for machine learning practitioners to adopt publicly available pre-trained vision foundation models, effectively sharing a common backbone architecture across a multitude of applications such as classification, segmentation, depth estimation, retrieval, question-answering and more. The study of attacks on such foundation models and their impact to multiple downstream tasks remains vastly unexplored. This work proposes a general framework that forges task-agnostic adversarial examples by maximally disrupting the feature representation obtained with foundation models. We extensively evaluate the security of the feature representations obtained by popular vision foundation models by measuring the impact of this attack on multiple downstream tasks and its transferability between models.
format Preprint
id arxiv_https___arxiv_org_abs_2503_03842
institution arXiv
publishDate 2025
record_format arxiv
spellingShingle Task-Agnostic Attacks Against Vision Foundation Models
Pulfer, Brian
Belousov, Yury
Kinakh, Vitaliy
Furon, Teddy
Voloshynovskiy, Slava
Computer Vision and Pattern Recognition
Artificial Intelligence
Cryptography and Security
Machine Learning
The study of security in machine learning mainly focuses on downstream task-specific attacks, where the adversarial example is obtained by optimizing a loss function specific to the downstream task. At the same time, it has become standard practice for machine learning practitioners to adopt publicly available pre-trained vision foundation models, effectively sharing a common backbone architecture across a multitude of applications such as classification, segmentation, depth estimation, retrieval, question-answering and more. The study of attacks on such foundation models and their impact to multiple downstream tasks remains vastly unexplored. This work proposes a general framework that forges task-agnostic adversarial examples by maximally disrupting the feature representation obtained with foundation models. We extensively evaluate the security of the feature representations obtained by popular vision foundation models by measuring the impact of this attack on multiple downstream tasks and its transferability between models.
title Task-Agnostic Attacks Against Vision Foundation Models
topic Computer Vision and Pattern Recognition
Artificial Intelligence
Cryptography and Security
Machine Learning
url https://arxiv.org/abs/2503.03842