CAI: Caption-Sensitive Attention Intervention for Mitigating Object Hallucination in Large Vision-Language Models

Fuente: arXiv
Gespeichert in:
Bibliographische Detailangaben
Hauptverfasser: Li, Qiming, Ye, Zekai, Feng, Xiaocheng, Zhong, Weihong, Qin, Libo, Chen, Ruihan, Li, Baohang, Jiang, Kui, Wang, Yaowei, Liu, Ting, Qin, Bing
Format: Preprint
Veröffentlicht: 2025
Schlagworte:
Online-Zugang:
Tags: Tag hinzufügen
Keine Tags, Fügen Sie den ersten Tag hinzu!
_version_ 1866915365428133888
author Li, Qiming
Ye, Zekai
Feng, Xiaocheng
Zhong, Weihong
Qin, Libo
Chen, Ruihan
Li, Baohang
Jiang, Kui
Wang, Yaowei
Liu, Ting
Qin, Bing
author_facet Li, Qiming
Ye, Zekai
Feng, Xiaocheng
Zhong, Weihong
Qin, Libo
Chen, Ruihan
Li, Baohang
Jiang, Kui
Wang, Yaowei
Liu, Ting
Qin, Bing
contents Although Large Vision-Language Models (LVLMs) have demonstrated powerful capabilities in interpreting visual information, they frequently produce content that deviates from visual information, leading to object hallucination. To tackle this, recent works mostly depend on expensive manual annotations and training cost, or significantly increase inference time. In this work, we observe that LVLMs' attention to visual information is significantly stronger when answering caption queries compared to non-caption queries. Inspired by this phenomenon, we propose Caption-sensitive Attention Intervention (CAI), a training-free, plug-and-play hallucination mitigation method that leverages the attention activation pattern in response to caption queries to enhance LVLMs' visual perception capability. Extensive experimental results across four benchmarks covering both discriminative and generative tasks, demonstrate that CAI achieves state-of-the-art (SOTA) hallucination mitigating performance only with minimal additional inference cost.
format Preprint
id arxiv_https___arxiv_org_abs_2506_23590
institution arXiv
publishDate 2025
record_format arxiv
spellingShingle CAI: Caption-Sensitive Attention Intervention for Mitigating Object Hallucination in Large Vision-Language Models
Li, Qiming
Ye, Zekai
Feng, Xiaocheng
Zhong, Weihong
Qin, Libo
Chen, Ruihan
Li, Baohang
Jiang, Kui
Wang, Yaowei
Liu, Ting
Qin, Bing
Computer Vision and Pattern Recognition
Although Large Vision-Language Models (LVLMs) have demonstrated powerful capabilities in interpreting visual information, they frequently produce content that deviates from visual information, leading to object hallucination. To tackle this, recent works mostly depend on expensive manual annotations and training cost, or significantly increase inference time. In this work, we observe that LVLMs' attention to visual information is significantly stronger when answering caption queries compared to non-caption queries. Inspired by this phenomenon, we propose Caption-sensitive Attention Intervention (CAI), a training-free, plug-and-play hallucination mitigation method that leverages the attention activation pattern in response to caption queries to enhance LVLMs' visual perception capability. Extensive experimental results across four benchmarks covering both discriminative and generative tasks, demonstrate that CAI achieves state-of-the-art (SOTA) hallucination mitigating performance only with minimal additional inference cost.
title CAI: Caption-Sensitive Attention Intervention for Mitigating Object Hallucination in Large Vision-Language Models
topic Computer Vision and Pattern Recognition
url https://arxiv.org/abs/2506.23590