Structured Attention Matters to Multimodal LLMs in Document Understanding

• URL: http://arxiv.org/abs/2506.21600v1
• Date: Thu, 19 Jun 2025 07:16:18 GMT
• Title: Structured Attention Matters to Multimodal LLMs in Document Understanding
• Authors: Chang Liu, Hongkai Chen, Yujun Cai, Hang Wu, Qingwen Ye, Ming-Hsuan Yang, Yiwei Wang,
• Abstract summary: We investigate how input format influences document comprehension performance.<n>We discover that raw OCR text often impairs rather than improves MLLMs' performance.<n>We propose a novel structure-preserving approach that encodes document elements using the LaTex paradigm.
• Score: 52.37530640460363
• License: http://arxiv.org/licenses/nonexclusive-distrib/1.0/
• Abstract: Document understanding remains a significant challenge for multimodal large language models (MLLMs). While previous research has primarily focused on locating evidence pages through precise multimodal queries, our work investigates a fundamental yet overlooked aspect: how input format influences document comprehension performance. Through systematic analysis, we discover that raw OCR text often impairs rather than improves MLLMs' performance, which is a counterintuitive finding we attribute to attention dispersion and structure loss. To further substantiate our hypothesis, we propose a novel structure-preserving approach that encodes document elements using the LaTex paradigm, maintaining the hierarchical organization and spatial relationships critical for comprehension. Our attention analysis reveals that structured text induces structured attention patterns on both textual and visual content, directing models to focus on semantically meaningful regions while reducing attention waste. This approach significantly enhances MLLMs' document question answering performance across diverse document types without requiring architectural modifications or additional training.

Related papers

• DISRetrieval: Harnessing Discourse Structure for Long Document Retrieval [51.89673002051528]
  DISRetrieval is a novel hierarchical retrieval framework that leverages linguistic discourse structure to enhance long document understanding.<n>Our studies confirm that discourse structure significantly enhances retrieval effectiveness across different document lengths and query types.
  arXiv  Detail & Related papers  (2025-05-26T14:45:12Z)
• How do Large Language Models Understand Relevance? A Mechanistic Interpretability Perspective [64.00022624183781]
  Large language models (LLMs) can assess relevance and support information retrieval (IR) tasks.<n>We investigate how different LLM modules contribute to relevance judgment through the lens of mechanistic interpretability.
  arXiv  Detail & Related papers  (2025-04-10T16:14:55Z)
• QID: Efficient Query-Informed ViTs in Data-Scarce Regimes for OCR-free Visual Document Understanding [53.69841526266547]
  Fine-tuning a pre-trained Vision-Language Model with new datasets often falls short in optimizing the vision encoder.<n>We introduce QID, a novel, streamlined, architecture-preserving approach that integrates query embeddings into the vision encoder.
  arXiv  Detail & Related papers  (2025-04-03T18:47:16Z)
• DoPTA: Improving Document Layout Analysis using Patch-Text Alignment [3.3181276611945267]
  We present a novel image-text alignment technique specially designed for leveraging the textual information in document images to improve performance on visual tasks.<n>Our document encoder model DoPTA - trained with this technique demonstrates strong performance on a wide range of document image understanding tasks, without requiring OCR during inference.<n>DoPTA also sets new state-of-the art results on D4LA, and FUNSD, two challenging document visual analysis benchmarks.
  arXiv  Detail & Related papers  (2024-12-17T13:26:31Z)
• Hierarchical Visual Feature Aggregation for OCR-Free Document Understanding [41.43688559565315]
  We present a novel OCR-free document understanding framework based on pretrained Multimodal Large Language Models (MLLMs) Our approach employs multi-scale visual features to handle various font sizes within document images. We introduce a novel instruction tuning task, which facilitates the model's text-reading capability by learning to predict the relative positions of input text.
  arXiv  Detail & Related papers  (2024-11-08T00:58:12Z)
• PDF-MVQA: A Dataset for Multimodal Information Retrieval in PDF-based Visual Question Answering [13.625303311724757]
  Document Question Answering (QA) presents a challenge in understanding visually-rich documents (VRD) We propose PDF-MVQA, which is tailored for research journal articles, encompassing multiple pages and multimodal information retrieval.
  arXiv  Detail & Related papers  (2024-04-19T09:00:05Z)
• LayoutLLM: Large Language Model Instruction Tuning for Visually Rich Document Understanding [0.0]
  This paper proposes LayoutLLM, a more flexible document analysis method for understanding imaged documents. Existing methods have been developed to enhance document comprehension by incorporating pre-training awareness of images, text, and layout structure. Our experiments demonstrate improvement over the baseline model in various document analysis tasks.
  arXiv  Detail & Related papers  (2024-03-21T09:25:24Z)
• Enhancing Visual Document Understanding with Contrastive Learning in Large Visual-Language Models [56.76307866160105]
  We propose a contrastive learning framework, termed Document Object COntrastive learning (DoCo) DoCo leverages an auxiliary multimodal encoder to obtain the features of document objects and align them to the visual features generated by the vision encoder of Large Visual-Language Models (LVLMs) We demonstrate that the proposed DoCo serves as a plug-and-play pre-training method, which can be employed in the pre-training of various LVLMs without inducing any increase in computational complexity during the inference process.
  arXiv  Detail & Related papers  (2024-02-29T10:17:27Z)
• Unified Pretraining Framework for Document Understanding [52.224359498792836]
  We present UDoc, a new unified pretraining framework for document understanding. UDoc is designed to support most document understanding tasks, extending the Transformer to take multimodal embeddings as input. An important feature of UDoc is that it learns a generic representation by making use of three self-supervised losses.
  arXiv  Detail & Related papers  (2022-04-22T21:47:04Z)

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.