KG-ViP: Bridging Knowledge Grounding and Visual Perception in Multi-modal LLMs for Visual Question Answering

Fuente: arXiv
Saved in:
Bibliographic Details
Main Authors: Li, Zhiyang, Ke, Ao, Cao, Yukun, Xie, Xike
Format: Preprint
Published: 2026
Subjects:
Online Access:
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1866917537124450304
author Li, Zhiyang
Ke, Ao
Cao, Yukun
Xie, Xike
author_facet Li, Zhiyang
Ke, Ao
Cao, Yukun
Xie, Xike
contents Multi-modal Large Language Models (MLLMs) for Visual Question Answering (VQA) often suffer from dual limitations: knowledge hallucination and insufficient fine-grained visual perception. Crucially, we identify that commonsense graphs and scene graphs provide precisely complementary solutions to these respective deficiencies by providing rich external knowledge and capturing fine-grained visual details. However, prior works typically treat them in isolation, overlooking their synergistic potential. To bridge this gap, we propose KG-ViP, a unified framework that empowers MLLMs by fusing scene graphs and commonsense graphs. The core of the KG-ViP framework is a novel retrieval-and-fusion pipeline that utilizes the query as a semantic bridge to progressively integrate both graphs, synthesizing a unified structured context that facilitates reliable multi-modal reasoning. Extensive experiments on FVQA 2.0+ and MVQA benchmarks demonstrate that KG-ViP significantly outperforms existing VQA methods.
format Preprint
id arxiv_https___arxiv_org_abs_2601_11632
institution arXiv
publishDate 2026
record_format arxiv
spellingShingle KG-ViP: Bridging Knowledge Grounding and Visual Perception in Multi-modal LLMs for Visual Question Answering
Li, Zhiyang
Ke, Ao
Cao, Yukun
Xie, Xike
Computer Vision and Pattern Recognition
Multi-modal Large Language Models (MLLMs) for Visual Question Answering (VQA) often suffer from dual limitations: knowledge hallucination and insufficient fine-grained visual perception. Crucially, we identify that commonsense graphs and scene graphs provide precisely complementary solutions to these respective deficiencies by providing rich external knowledge and capturing fine-grained visual details. However, prior works typically treat them in isolation, overlooking their synergistic potential. To bridge this gap, we propose KG-ViP, a unified framework that empowers MLLMs by fusing scene graphs and commonsense graphs. The core of the KG-ViP framework is a novel retrieval-and-fusion pipeline that utilizes the query as a semantic bridge to progressively integrate both graphs, synthesizing a unified structured context that facilitates reliable multi-modal reasoning. Extensive experiments on FVQA 2.0+ and MVQA benchmarks demonstrate that KG-ViP significantly outperforms existing VQA methods.
title KG-ViP: Bridging Knowledge Grounding and Visual Perception in Multi-modal LLMs for Visual Question Answering
topic Computer Vision and Pattern Recognition
url https://arxiv.org/abs/2601.11632