Attention Grounded Enhancement for Visual Document Retrieval

• URL: http://arxiv.org/abs/2511.13415v1
• Date: Mon, 17 Nov 2025 14:28:41 GMT
• Title: Attention Grounded Enhancement for Visual Document Retrieval
• Authors: Wanqing Cui, Wei Huang, Yazhi Guo, Yibo Hu, Meiguang Jin, Junfeng Ma, Keping Bi,
• Abstract summary: We propose a textbfAttention-textbfGrounded textbfREtriever textbfEnhancement (AGREE) framework for visual document retrieval.<n>AGREE combines cross-modal attention from large language models as proxy local supervision to guide the identification of relevant document regions.<n> Experiments on the challenging ViDoRe V2 benchmark show that AGREE significantly outperforms the global-supervision-only baseline.
• Score: 12.602988404893305
• License: http://creativecommons.org/licenses/by/4.0/
• Abstract: Visual document retrieval requires understanding heterogeneous and multi-modal content to satisfy information needs. Recent advances use screenshot-based document encoding with fine-grained late interaction, significantly improving retrieval performance. However, retrievers are still trained with coarse global relevance labels, without revealing which regions support the match. As a result, retrievers tend to rely on surface-level cues and struggle to capture implicit semantic connections, hindering their ability to handle non-extractive queries. To alleviate this problem, we propose a \textbf{A}ttention-\textbf{G}rounded \textbf{RE}triever \textbf{E}nhancement (AGREE) framework. AGREE leverages cross-modal attention from multimodal large language models as proxy local supervision to guide the identification of relevant document regions. During training, AGREE combines local signals with the global signals to jointly optimize the retriever, enabling it to learn not only whether documents match, but also which content drives relevance. Experiments on the challenging ViDoRe V2 benchmark show that AGREE significantly outperforms the global-supervision-only baseline. Quantitative and qualitative analyses further demonstrate that AGREE promotes deeper alignment between query terms and document regions, moving beyond surface-level matching toward more accurate and interpretable retrieval. Our code is available at: https://anonymous.4open.science/r/AGREE-2025.

Related papers

• Retrieval or Representation? Reassessing Benchmark Gaps in Multilingual and Visually Rich RAG [1.4425299138308667]
  BM25 rank documents by term overlap with corpus-level weighting.<n>End-to-end multimodal retrievers trained on large query-document datasets claim substantial improvements over these approaches.<n>We demonstrate that better document representation is the primary driver of benchmark improvements.
  arXiv  Detail & Related papers  (2026-03-04T16:21:20Z)
• RegionRAG: Region-level Retrieval-Augumented Generation for Visually-Rich Documents [40.107303323097646]
  modelname is a novel framework that shifts the retrieval paradigm from the document level to the region level.<n> Experiments on six benchmarks demonstrate that RegionRAG achieves state-of-the-art performance.
  arXiv  Detail & Related papers  (2025-10-31T08:00:32Z)
• Cross-Granularity Hypergraph Retrieval-Augmented Generation for Multi-hop Question Answering [49.43814054718318]
  Multi-hop question answering (MHQA) requires integrating knowledge scattered across multiple passages to derive the correct answer.<n>Traditional retrieval-augmented generation (RAG) methods primarily focus on coarse-grained textual semantic similarity.<n>We propose a novel RAG approach called HGRAG for MHQA that achieves cross-granularity integration of structural and semantic information via hypergraphs.
  arXiv  Detail & Related papers  (2025-08-15T06:36:13Z)
• Hierarchical Lexical Graph for Enhanced Multi-Hop Retrieval [22.33550491040999]
  RAG grounds large language models in external evidence, yet it still falters when answers must be pieced together across semantically distant documents.<n>We build two plug-and-play retrievers: StatementGraphRAG and TopicGraphRAG.<n>Our methods outperform naive chunk-based RAG achieving an average relative improvement of 23.1% in retrieval recall and correctness.
  arXiv  Detail & Related papers  (2025-06-09T17:58:35Z)
• Cognitive-Aligned Document Selection for Retrieval-augmented Generation [2.9060210098040855]
  We propose GGatrieval to dynamically update queries and filter high-quality, reliable retrieval documents.<n>We parse the user query into its syntactic components and perform fine-grained grounded alignment with the retrieved documents.<n>Our approach introduces a novel criterion for filtering retrieved documents, closely emulating human strategies for acquiring targeted information.
  arXiv  Detail & Related papers  (2025-02-17T13:00:15Z)
• GeAR: Generation Augmented Retrieval [82.20696567697016]
  This paper introduces a novel method, $textbfGe$neration.<n>It improves the global document-Query similarity through contrastive learning, but also integrates well-designed fusion and decoding modules.<n>When used as a retriever, GeAR does not incur any additional computational cost over bi-encoders.
  arXiv  Detail & Related papers  (2025-01-06T05:29:00Z)
• Generative Retrieval Meets Multi-Graded Relevance [104.75244721442756]
  We introduce a framework called GRaded Generative Retrieval (GR$2$) GR$2$ focuses on two key components: ensuring relevant and distinct identifiers, and implementing multi-graded constrained contrastive training. Experiments on datasets with both multi-graded and binary relevance demonstrate the effectiveness of GR$2$.
  arXiv  Detail & Related papers  (2024-09-27T02:55:53Z)
• Leveraging Inter-Chunk Interactions for Enhanced Retrieval in Large Language Model-Based Question Answering [12.60063463163226]
  IIER captures the internal connections between document chunks by considering three types of interactions: structural, keyword, and semantic. It identifies multiple seed nodes based on the target question and iteratively searches for relevant chunks to gather supporting evidence. It refines the context and reasoning chain, aiding the large language model in reasoning and answer generation.
  arXiv  Detail & Related papers  (2024-08-06T02:39:55Z)
• BERM: Training the Balanced and Extractable Representation for Matching to Improve Generalization Ability of Dense Retrieval [54.66399120084227]
  We propose a novel method to improve the generalization of dense retrieval via capturing matching signal called BERM. Dense retrieval has shown promise in the first-stage retrieval process when trained on in-domain labeled datasets.
  arXiv  Detail & Related papers  (2023-05-18T15:43:09Z)
• UnifieR: A Unified Retriever for Large-Scale Retrieval [84.61239936314597]
  Large-scale retrieval is to recall relevant documents from a huge collection given a query. Recent retrieval methods based on pre-trained language models (PLM) can be coarsely categorized into either dense-vector or lexicon-based paradigms. We propose a new learning framework, UnifieR which unifies dense-vector and lexicon-based retrieval in one model with a dual-representing capability.
  arXiv  Detail & Related papers  (2022-05-23T11:01:59Z)
• Augmenting Document Representations for Dense Retrieval with Interpolation and Perturbation [49.940525611640346]
  Document Augmentation for dense Retrieval (DAR) framework augments the representations of documents with their Dense Augmentation and perturbations. We validate the performance of DAR on retrieval tasks with two benchmark datasets, showing that the proposed DAR significantly outperforms relevant baselines on the dense retrieval of both the labeled and unlabeled documents.
  arXiv  Detail & Related papers  (2022-03-15T09:07:38Z)

This list is automatically generated from the titles and abstracts of the papers in this site.

This site does not guarantee the quality of this site (including all information) and is not responsible for any consequences.