Vanish into Thin Air: Cross-prompt Universal Adversarial Attacks for SAM2

Ziqi Zhou (Huazhong University of Science and Technology) · Shengshan Hu (Huazhong University of Science and Technology) · Leo Yu Zhang (Griffith University) · Hai Jin (Huazhong University of Science and Technology) · Yifan Hu (Heilongjiang University) · Yufei Song (Huazhong University of Science and Technology) · Zijing Li (Huazhong University of Science and Technology) · Dezhong Yao (Huazhong University of Science and Technology) · Long Zheng (Huazhong University of Science and Technology, Tsinghua University)

adversarial examples architectural differencescross-prompt universal adversarial attackdirectional guidancedual semantic deviationfoundation model generalization capability image segmentation performance gapprompt dependencyrobustness segmentation tasks semantic consistencysemantic entanglementtarget-scanning strategy

Recent studies reveal the vulnerability of the image segmentation foundation model SAM to adversarial examples. Its successor, SAM2, has attracted significant attention due to its strong generalization capability in video segmentation. However, its robustness remains unexplored, and it is unclear whether existing attacks on SAM can be directly transferred to SAM2. In this paper, we first analyze the performance gap of existing attacks between SAM and SAM2 and highlight two key challenges arising from their architectural differences: directional guidance from the prompt and semantic entanglement across consecutive frames. To address these issues, we propose UAP-SAM2, the first cross-prompt universal adversarial attack against SAM2 driven by dual semantic deviation. For cross-prompt transferability, we begin by designing a target-scanning strategy that divides each frame into k regions, each randomly assigned a prompt, to reduce prompt dependency during optimization. For effectiveness, we design a dual semantic deviation framework that optimizes a UAP by distorting the semantics within the current frame and disrupting the semantic consistency across consecutive frames. Extensive experiments on six datasets across two segmentation tasks demonstrate the effectiveness of the proposed method for SAM2. The comparative results show that UAP-SAM2 significantly outperforms state-of-the-art (SOTA) attacks by a large margin.