Cover Image

PA-Attack: Guiding Gray-Box Attacks on LVLM Vision Encoders with Prototypes and Attention

Fuente: arXiv

Saved in:

Bibliographic Details
Main Authors:	Mei, Hefei, Wang, Zirui, Xu, Chang, Guo, Jianyuan, Dong, Minjing
Format:	Preprint
Published:	2026
Subjects:	Computer Vision and Pattern Recognition
Online Access:	Acceder al recurso
Tags:	Add Tag No Tags, Be the first to tag this record!

Similar Items

VEAttack: Downstream-agnostic Vision Encoder Attack against Large Vision Language Models
by: Mei, Hefei, et al.
Published: (2025)

Efficient Image-to-Image Diffusion Classifier for Adversarial Robustness
by: Mei, Hefei, et al.
Published: (2024)

VALD: Multi-Stage Vision Attack Detection for Efficient LVLM Defense
by: Kadvil, Nadav, et al.
Published: (2026)

Mitigating Object Hallucinations in Large Vision-Language Models via Attention Calibration
by: Zhu, Younan, et al.
Published: (2025)

Harnessing Vision Foundation Models for High-Performance, Training-Free Open Vocabulary Segmentation
by: Shi, Yuheng, et al.
Published: (2024)

VSSD: Vision Mamba with Non-Causal State Space Duality
by: Shi, Yuheng, et al.
Published: (2024)

Improving the JPEG-resistance of Adversarial Attacks on Face Recognition by Interpolation Smoothing
by: Guo, Kefu, et al.
Published: (2024)

When and Where to Attack? Stage-wise Attention-Guided Adversarial Attack on Large Vision Language Models
by: Kwak, Jaehyun, et al.
Published: (2026)

Pushing the Frontier of Black-Box LVLM Attacks via Fine-Grained Detail Targeting
by: Zhao, Xiaohan, et al.
Published: (2026)

Adversarial Attacks on Both Face Recognition and Face Anti-spoofing Models
by: Zhou, Fengfan, et al.
Published: (2024)

Improving the Transferability of Adversarial Attacks on Face Recognition with Diverse Parameters Augmentation
by: Zhou, Fengfan, et al.
Published: (2024)

Multi-Scale VMamba: Hierarchy in Hierarchy Visual State Space Model
by: Shi, Yuheng, et al.
Published: (2024)

Query-Efficient Hard-Label Black-Box Attack against Vision Transformers
by: Zhou, Chao, et al.
Published: (2024)

Vision Superalignment: Weak-to-Strong Generalization for Vision Foundation Models
by: Guo, Jianyuan, et al.
Published: (2024)

SGHA-Attack: Semantic-Guided Hierarchical Alignment for Transferable Targeted Attacks on Vision-Language Models
by: Wang, Haobo, et al.
Published: (2026)

Transferable Adversarial Attacks on Black-Box Vision-Language Models
by: Hu, Kai, et al.
Published: (2025)

Stealthy Backdoor Attack in Self-Supervised Learning Vision Encoders for Large Vision Language Models
by: Liu, Zhaoyi, et al.
Published: (2025)

VEGAS: Mitigating Hallucinations in Large Vision-Language Models via Vision-Encoder Attention Guided Adaptive Steering
by: Wang, Zihu, et al.
Published: (2025)

Improving Visual Quality and Transferability of Adversarial Attacks on Face Recognition Simultaneously with Adversarial Restoration
by: Zhou, Fengfan, et al.
Published: (2023)

Breaking the Black-Box: Confidence-Guided Model Inversion Attack for Distribution Shift
by: Liu, Xinhao, et al.
Published: (2024)

Feature Clipping for Uncertainty Calibration
by: Tao, Linwei, et al.
Published: (2024)

Attention-Guided Patch-Wise Sparse Adversarial Attacks on Vision-Language-Action Models
by: Zhang, Naifu, et al.
Published: (2025)

Catching the Details: Self-Distilled RoI Predictors for Fine-Grained MLLM Perception
by: Shi, Yuheng, et al.
Published: (2025)

SHIELD: Suppressing Hallucinations In LVLM Encoders via Bias and Vulnerability Defense
by: Huang, Yiyang, et al.
Published: (2025)

PG-Attack: A Precision-Guided Adversarial Attack Framework Against Vision Foundation Models for Autonomous Driving
by: Fu, Jiyuan, et al.
Published: (2024)

Black-Box Adversarial Attack on Vision Language Models for Autonomous Driving
by: Wang, Lu, et al.
Published: (2025)

OpenLVLM-MIA: A Controlled Benchmark Revealing the Limits of Membership Inference Attacks on Large Vision-Language Models
by: Miyamoto, Ryoto, et al.
Published: (2025)

SSyncOA: Self-synchronizing Object-aligned Watermarking to Resist Cropping-paste Attacks
by: Zhao, Chengxin, et al.
Published: (2024)

Hard-Label Black-Box Attacks on 3D Point Clouds
by: Liu, Daizong, et al.
Published: (2024)

AFTER: Mitigating the Object Hallucination of LVLM via Adaptive Factual-Guided Activation Editing
by: Wang, Tianbo, et al.
Published: (2026)

DAMRO: Dive into the Attention Mechanism of LVLM to Reduce Object Hallucination
by: Gong, Xuan, et al.
Published: (2024)

Vision Remember: Recovering Visual Information in Efficient LVLM with Vision Feature Resampling
by: Feng, Ze, et al.
Published: (2025)

Low-Frequency Black-Box Backdoor Attack via Evolutionary Algorithm
by: Qiao, Yanqi, et al.
Published: (2024)

Consistency Calibration: Improving Uncertainty Calibration via Consistency among Perturbed Neighbors
by: Tao, Linwei, et al.
Published: (2024)

Rethinking Impersonation and Dodging Attacks on Face Recognition Systems
by: Zhou, Fengfan, et al.
Published: (2024)

Securing Vision-Language Models with a Robust Encoder Against Jailbreak and Adversarial Attacks
by: Hossain, Md Zarif, et al.
Published: (2024)

Thermal Topology Collapse: Universal Physical Patch Attacks on Infrared Vision Systems
by: Hu, Chengyin, et al.
Published: (2026)

Distortion-Aware Adversarial Attacks on Bounding Boxes of Object Detectors
by: Phuc, Pham, et al.
Published: (2024)

When Surfaces Lie: Exploiting Wrinkle-Induced Attention Shift to Attack Vision-Language Models
by: Hu, Chengyin, et al.
Published: (2026)

AdaIAT: Adaptively Increasing Attention to Generated Text to Alleviate Hallucinations in LVLM
by: Zhong, Li'an, et al.
Published: (2026)