paper-with-me

홈 › Papers

SHIELD: Suppressing Hallucinations In LVLM Encoders via Bias and Vulnerability Defense

2025-10-18 · Yiyang Huang, Liang Shi, Yitian Zhang, Yi Xu, Yun Fu arxiv

Large Vision-Language Models (LVLMs) excel in diverse cross-modal tasks. However, object hallucination, where models produce plausible but inaccurate object descriptions, remains a significant challenge. In contrast to previous work focusing on LLM components, this paper is the first to trace LVLM hallucinations to visual encoders and identifies three key issues: statistical bias, inherent bias, and vulnerability. To address these challenges, we propose SHIELD, a training-free framework that mitigates hallucinations through three strategies: re-weighting visual tokens to reduce statistical bias, introducing noise-derived tokens to counter inherent bias, and applying adversarial attacks with contrastive decoding to address vulnerability. Experiments demonstrate that SHIELD effectively mitigates object hallucinations across diverse benchmarks and LVLM families. Moreover, SHIELD achieves strong performance on the general LVLM benchmark, highlighting its broad applicability. Code is available at https://github.com/hukcc/SHIELD.

📄 PDF Abstract BibTeX arXiv:2510.16596

Code (0)

등록된 구현이 없습니다.

Similar Papers 제목 키워드 기반

Towards Interpretable Hallucination Analysis and Mitigation in LVLMs via Contrastive Neuron Steering

2026-01-31 · Guangtao Lyu, Xinyi Cheng, Qi Liu, Chenghao Xu 외 arxiv

LVLMs achieve remarkable multimodal understanding and generation but remain susceptible to hallucinations. Existing mitigation methods predominantly focus on output-level adjustments, leaving the internal mechanisms that…

Visual Grounding

IKOD: Mitigating Visual Attention Degradation in Large Vision-Language Models

2025-08-05 · Jiabing Yang, Chenhang Cui, Yiyang Zhou, Yixiang Chen 외 arxiv

Recent advancements in Large Vision-Language Models (LVLMs) have demonstrated significant progress across multiple domains. However, these models still face the inherent challenge of integrating vision and language for c…

Efficient Contrastive Decoding with Probabilistic Hallucination Detection - Mitigating Hallucinations in Large Vision Language Models -

2025-04-16 · Laura Fieback, Nishilkumar Balar, Jakob Spiegelberg, Hanno Gottschalk

Despite recent advances in Large Vision Language Models (LVLMs), these models still suffer from generating hallucinatory responses that do not align with the visual input provided. To mitigate such hallucinations, we int…

Hallucination

SDCD: Structure-Disrupted Contrastive Decoding for Mitigating Hallucinations in Large Vision-Language Models

2026-01-07 · Yuxuan Xia, Siheng Wang, Peng Li arxiv

Large Vision-Language Models (LVLMs) demonstrate significant progress in multimodal understanding and reasoning, yet object hallucination remains a critical challenge. While existing research focuses on mitigating langua…

Automating construction safety inspections using a multi-modal vision-language RAG framework

2025-10-05 · Chenxin Wang, Elyas Asadi Shamsabadi, Zhaohui Chen, Luming Shen 외 arxiv

Conventional construction safety inspection methods are often inefficient as they require navigating through large volume of information. Recent advances in large vision-language models (LVLMs) provide opportunities to a…

Information Retrieval