RECALLED: An Unbounded Resource Consumption Attack on Large Vision-Language Models

Authors: Haoran Gao, Yuanhe Zhang, Zhenhong Zhou, Lei Jiang, Fanyu Meng, Yujia Xiao, Kun Wang, Yang Liu, Junlan Feng
Published: 2025-07-24

Source: https://arxiv.org/abs/2507.18053

PDF: https://arxiv.org/pdf/2507.18053

AIにより推定されたラベル

攻撃戦略分析 セキュリティ戦略生成 視覚的手法

※ こちらのラベルはAIによって自動的に追加されました。そのため、正確でないことがあります。
詳細は文献データベースについてをご覧ください。

Abstract

Resource Consumption Attacks (RCAs) have emerged as a significant threat to the deployment of Large Language Models (LLMs). With the integration of vision modalities, additional attack vectors exacerbate the risk of RCAs in large vision-language models (LVLMs). However, existing red-teaming studies have largely overlooked visual inputs as a potential attack surface, resulting in insufficient mitigation strategies against RCAs in LVLMs. To address this gap, we propose RECALLED (REsource Consumption Attack on Large Vision-LanguagE MoDels), the first approach for exploiting visual modalities to trigger unbounded RCAs red-teaming. First, we present Vision Guided Optimization, a fine-grained pixel-level optimization, to obtain Output Recall adversarial perturbations, which can induce repeating output. Then, we inject the perturbations into visual inputs, triggering unbounded generations to achieve the goal of RCAs. Additionally, we introduce Multi-Objective Parallel Losses to generate universal attack templates and resolve optimization conflicts when intending to implement parallel attacks. Empirical results demonstrate that RECALLED increases service response latency by over 26 ↑, resulting in an additional 20% increase in GPU utilization and memory consumption. Our study exposes security vulnerabilities in LVLMs and establishes a red-teaming framework that can facilitate future defense development against RCAs.