Guardado en:
Detalles Bibliográficos
Autores principales: Park, Hyeongcheol, Seo, Jiyoung, Mun, Jaewon, Park, Hogun, Byeon, Wonmin, Kim, Sung June, Im, Hyeonsoo, Lee, JeungSub, Kim, Sangpil
Formato: Preprint
Publicado: 2025
Materias:
Acceso en línea:https://arxiv.org/abs/2512.20136
Etiquetas: Agregar Etiqueta
Sin Etiquetas, Sea el primero en etiquetar este registro!
_version_ 1866914465547550720
author Park, Hyeongcheol
Seo, Jiyoung
Mun, Jaewon
Park, Hogun
Byeon, Wonmin
Kim, Sung June
Im, Hyeonsoo
Lee, JeungSub
Kim, Sangpil
author_facet Park, Hyeongcheol
Seo, Jiyoung
Mun, Jaewon
Park, Hogun
Byeon, Wonmin
Kim, Sung June
Im, Hyeonsoo
Lee, JeungSub
Kim, Sangpil
contents Retrieval-Augmented Generation (RAG) has recently been extended to multimodal settings, connecting multimodal large language models (MLLMs) with vast corpora of external knowledge such as multimodal knowledge graphs (MMKGs). Despite their recent success, multimodal RAG in the audio-visual domain remains challenging due to 1) limited modality coverage and multi-hop connectivity of existing MMKGs, and 2) retrieval based solely on similarity in a shared multimodal embedding space, which fails to filter out off-topic or redundant knowledge. To address these limitations, we propose M$^3$KG-RAG, a Multi-hop Multimodal Knowledge Graph-enhanced RAG that retrieves query-aligned audio-visual knowledge from MMKGs, improving reasoning depth and answer faithfulness in MLLMs. Specifically, we devise a lightweight multi-agent pipeline to construct multi-hop MMKG (M$^3$KG), which contains context-enriched triplets of multimodal entities, enabling modality-wise retrieval based on input queries. Furthermore, we introduce GRASP (Grounded Retrieval And Selective Pruning), which ensures precise entity grounding to the query, evaluates answer-supporting relevance, and prunes redundant context to retain only knowledge essential for response generation. Extensive experiments across diverse multimodal benchmarks demonstrate that M$^3$KG-RAG significantly enhances MLLMs' multimodal reasoning and grounding over existing approaches. Project website: https://kuai-lab.github.io/cvpr2026m3kgrag/
format Preprint
id arxiv_https___arxiv_org_abs_2512_20136
institution arXiv
publishDate 2025
record_format arxiv
spellingShingle M$^3$KG-RAG: Multi-hop Multimodal Knowledge Graph-enhanced Retrieval-Augmented Generation
Park, Hyeongcheol
Seo, Jiyoung
Mun, Jaewon
Park, Hogun
Byeon, Wonmin
Kim, Sung June
Im, Hyeonsoo
Lee, JeungSub
Kim, Sangpil
Computation and Language
Artificial Intelligence
Retrieval-Augmented Generation (RAG) has recently been extended to multimodal settings, connecting multimodal large language models (MLLMs) with vast corpora of external knowledge such as multimodal knowledge graphs (MMKGs). Despite their recent success, multimodal RAG in the audio-visual domain remains challenging due to 1) limited modality coverage and multi-hop connectivity of existing MMKGs, and 2) retrieval based solely on similarity in a shared multimodal embedding space, which fails to filter out off-topic or redundant knowledge. To address these limitations, we propose M$^3$KG-RAG, a Multi-hop Multimodal Knowledge Graph-enhanced RAG that retrieves query-aligned audio-visual knowledge from MMKGs, improving reasoning depth and answer faithfulness in MLLMs. Specifically, we devise a lightweight multi-agent pipeline to construct multi-hop MMKG (M$^3$KG), which contains context-enriched triplets of multimodal entities, enabling modality-wise retrieval based on input queries. Furthermore, we introduce GRASP (Grounded Retrieval And Selective Pruning), which ensures precise entity grounding to the query, evaluates answer-supporting relevance, and prunes redundant context to retain only knowledge essential for response generation. Extensive experiments across diverse multimodal benchmarks demonstrate that M$^3$KG-RAG significantly enhances MLLMs' multimodal reasoning and grounding over existing approaches. Project website: https://kuai-lab.github.io/cvpr2026m3kgrag/
title M$^3$KG-RAG: Multi-hop Multimodal Knowledge Graph-enhanced Retrieval-Augmented Generation
topic Computation and Language
Artificial Intelligence
url https://arxiv.org/abs/2512.20136