Enhancing Adversarial Transferability in Visual-Language Pre-training Models via Local Shuffle and Sample-based Attack

Fuente: arXiv
Saved in:
Bibliographic Details
Main Authors: Liu, Xin, Zhou, Aoyang
Format: Preprint
Published: 2025
Subjects:
Online Access:
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1866912683285020672
author Liu, Xin
Zhou, Aoyang
Zhou, Aoyang
author_facet Liu, Xin
Zhou, Aoyang
Zhou, Aoyang
contents Visual-Language Pre-training (VLP) models have achieved significant performance across various downstream tasks. However, they remain vulnerable to adversarial examples. While prior efforts focus on improving the adversarial transferability of multimodal adversarial examples through cross-modal interactions, these approaches suffer from overfitting issues, due to a lack of input diversity by relying excessively on information from adversarial examples in one modality when crafting attacks in another. To address this issue, we draw inspiration from strategies in some adversarial training methods and propose a novel attack called Local Shuffle and Sample-based Attack (LSSA). LSSA randomly shuffles one of the local image blocks, thus expanding the original image-text pairs, generating adversarial images, and sampling around them. Then, it utilizes both the original and sampled images to generate the adversarial texts. Extensive experiments on multiple models and datasets demonstrate that LSSA significantly enhances the transferability of multimodal adversarial examples across diverse VLP models and downstream tasks. Moreover, LSSA outperforms other advanced attacks on Large Vision-Language Models.
format Preprint
id arxiv_https___arxiv_org_abs_2511_00831
institution arXiv
publishDate 2025
record_format arxiv
spellingShingle Enhancing Adversarial Transferability in Visual-Language Pre-training Models via Local Shuffle and Sample-based Attack
Liu, Xin
Zhou, Aoyang
Zhou, Aoyang
Computer Vision and Pattern Recognition
Artificial Intelligence
Visual-Language Pre-training (VLP) models have achieved significant performance across various downstream tasks. However, they remain vulnerable to adversarial examples. While prior efforts focus on improving the adversarial transferability of multimodal adversarial examples through cross-modal interactions, these approaches suffer from overfitting issues, due to a lack of input diversity by relying excessively on information from adversarial examples in one modality when crafting attacks in another. To address this issue, we draw inspiration from strategies in some adversarial training methods and propose a novel attack called Local Shuffle and Sample-based Attack (LSSA). LSSA randomly shuffles one of the local image blocks, thus expanding the original image-text pairs, generating adversarial images, and sampling around them. Then, it utilizes both the original and sampled images to generate the adversarial texts. Extensive experiments on multiple models and datasets demonstrate that LSSA significantly enhances the transferability of multimodal adversarial examples across diverse VLP models and downstream tasks. Moreover, LSSA outperforms other advanced attacks on Large Vision-Language Models.
title Enhancing Adversarial Transferability in Visual-Language Pre-training Models via Local Shuffle and Sample-based Attack
topic Computer Vision and Pattern Recognition
Artificial Intelligence
url https://arxiv.org/abs/2511.00831