Copertina

Making MLLMs Blind: Adversarial Smuggling Attacks in MLLM Content Moderation

Fuente: arXiv

Salvato in:

Dettagli Bibliografici
Autori principali:	Li, Zhiheng, Ma, Zongyang, Pan, Yuntong, Zhang, Ziqi, Lv, Xiaolei, Li, Bo, Gao, Jun, Zhang, Jianing, Yuan, Chunfeng, Li, Bing, Hu, Weiming
Natura:	Preprint
Pubblicazione:	2026
Soggetti:	Computer Vision and Pattern Recognition
Accesso online:	Acceder al recurso
Tags:	Aggiungi Tag Nessun Tag, puoi essere il primo ad aggiungerne!!

Documenti analoghi

How Far Is Document Parsing from Solved? PureDocBench: A Source-TraceableBenchmark across Clean, Degraded, and Real-World Settings
di: Li, Zhiheng, et al.
Pubblicazione: (2026)

How to Make Cross Encoder a Good Teacher for Efficient Image-Text Retrieval?
di: Chen, Yuxin, et al.
Pubblicazione: (2024)

Beyond Semantic Search: Towards Referential Anchoring in Composed Image Retrieval
di: Yang, Yuxin, et al.
Pubblicazione: (2026)

DetailFusion: A Dual-branch Framework with Detail Enhancement for Composed Image Retrieval
di: Yang, Yuxin, et al.
Pubblicazione: (2025)

MMhops-R1: Multimodal Multi-hop Reasoning
di: Zhang, Tao, et al.
Pubblicazione: (2025)

Set Prediction Guided by Semantic Concepts for Diverse Video Captioning
di: Lu, Yifan, et al.
Pubblicazione: (2023)

EA-VTR: Event-Aware Video-Text Retrieval
di: Ma, Zongyang, et al.
Pubblicazione: (2024)

Mitigating Hallucinations in Large Vision-Language Models by Self-Injecting Hallucinations
di: Lu, Yifan, et al.
Pubblicazione: (2025)

mR$^2$AG: Multimodal Retrieval-Reflection-Augmented Generation for Knowledge-Based VQA
di: Zhang, Tao, et al.
Pubblicazione: (2024)

Adversarial Attacks against Closed-Source MLLMs via Feature Optimal Alignment
di: Jia, Xiaojun, et al.
Pubblicazione: (2025)

Visual Contextual Attack: Jailbreaking MLLMs with Image-Driven Context Injection
di: Miao, Ziqi, et al.
Pubblicazione: (2025)

Content-decoupled Contrastive Learning-based Implicit Degradation Modeling for Blind Image Super-Resolution
di: Yuan, Jiang, et al.
Pubblicazione: (2024)

Visual-Advantage On-Policy Distillation for Vision-Language Models
di: Liu, Ruiqi, et al.
Pubblicazione: (2026)

Simple and Efficient Partial Graph Adversarial Attack: A New Perspective
di: Zhu, Guanghui, et al.
Pubblicazione: (2023)

Xuanwu: Evolving General Multimodal Models into an Industrial-Grade Foundation for Content Ecosystems
di: Zhang, Zhiqian, et al.
Pubblicazione: (2026)

The Routledge Handbook of Smuggling
Pubblicazione: (2021)

EventBench: Towards Comprehensive Benchmarking of Event-based MLLMs
di: Liu, Shaoyu, et al.
Pubblicazione: (2025)

Content-Agnostic Moderation for Stance-Neutral Recommendation
di: Li, Nan, et al.
Pubblicazione: (2024)

EventFlash: Towards Efficient MLLMs for Event-Based Vision
di: Liu, Shaoyu, et al.
Pubblicazione: (2026)

Filter-And-Refine: A MLLM Based Cascade System for Industrial-Scale Video Content Moderation
di: Wang, Zixuan, et al.
Pubblicazione: (2025)

Adaptive Perturbation for Adversarial Attack
di: Yuan, Zheng, et al.
Pubblicazione: (2021)

LaMPE: Length-aware Multi-grained Positional Encoding for Adaptive Long-context Scaling Without Training
di: Zhang, Sikui, et al.
Pubblicazione: (2025)

Weighted-Sampling Audio Adversarial Example Attack
di: Liu, Xiaolei, et al.
Pubblicazione: (2019)

InfoTok: Information-Theoretic Regularization for Capacity-Constrained Shared Visual Tokenization in Unified MLLMs
di: Tang, Lv, et al.
Pubblicazione: (2026)

PromptIQA: Boosting the Performance and Generalization for No-Reference Image Quality Assessment via Prompts
di: Chen, Zewen, et al.
Pubblicazione: (2024)

CrossGuard: Safeguarding MLLMs against Joint-Modal Implicit Malicious Attacks
di: Zhang, Xu, et al.
Pubblicazione: (2025)

Semantic Router: On the Feasibility of Hijacking MLLMs via a Single Adversarial Perturbation
di: Li, Changyue, et al.
Pubblicazione: (2025)

GMC-IQA: Exploiting Global-correlation and Mean-opinion Consistency for No-reference Image Quality Assessment
di: Chen, Zewen, et al.
Pubblicazione: (2024)

ADVEDM:Fine-grained Adversarial Attack against VLM-based Embodied Agents
di: Wang, Yichen, et al.
Pubblicazione: (2025)

CAVALRY-V: A Large-Scale Generator Framework for Adversarial Attacks on Video MLLMs
di: Zhang, Jiaming, et al.
Pubblicazione: (2025)

Dynamic Content Moderation in Livestreams: Combining Supervised Classification with MLLM-Boosted Similarity Matching
di: Yew, Wei Chee, et al.
Pubblicazione: (2025)

Towards Inclusive Toxic Content Moderation: Addressing Vulnerabilities to Adversarial Attacks in Toxicity Classifiers Tackling LLM-generated Content
di: Furniturewala, Shaz, et al.
Pubblicazione: (2025)

Math Blind: Failures in Diagram Understanding Undermine Reasoning in MLLMs
di: Sun, Yanpeng, et al.
Pubblicazione: (2025)

Erosion Attack for Adversarial Training to Enhance Semantic Segmentation Robustness
di: Song, Yufei, et al.
Pubblicazione: (2026)

Dense Connector for MLLMs
di: Yao, Huanjin, et al.
Pubblicazione: (2024)

Adversarial Attack for Explanation Robustness of Rationalization Models
di: Zhang, Yuankai, et al.
Pubblicazione: (2024)

DiffAttack: Evasion Attacks Against Diffusion-Based Adversarial Purification
di: Kang, Mintong, et al.
Pubblicazione: (2023)

Frequency-Domain Regularized Adversarial Alignment for Transferable Attacks against Closed-Source MLLMs
di: Yuan, Leitao, et al.
Pubblicazione: (2026)

The Illusion of Forgetting: Attack Unlearned Diffusion via Initial Latent Variable Optimization
di: Li, Manyi, et al.
Pubblicazione: (2026)

Chapter 6 Implementing the Asiento and Smuggling
di: Garcia-Monton, Alejandro
Pubblicazione: (2021)