arXivDaily arXiv每日学术速递 周一至周五更新
arXiv 2502.02501cs.CV
收藏

基于图的文档结构分析

Graph-based Document Structure Analysis

Karlsruhe Institute of Technology(卡尔斯鲁厄理工学院)

Yufan Chen, Ruiping Liu, Junwei Zheng, Di Wen, Kunyu Peng, Jiaming Zhang, Rainer Stiefelhagen

AI总结:

本文提出基于图的文档结构分析任务gDSA,构建含80K图像和4.13M关系标注的GraphDoc数据集,并设计DRGG生成器,以图结构同时建模文档元素的空间与逻辑关系,在基准上达到57.6% mAP$_g$@0.5。

摘要:

阅读文档时,浏览文档的空间布局是初步理解文档的第一步。然而,传统的文档布局分析(DLA)方法仅提供对文档的浅层解析,侧重于基础实例检测,往往无法捕捉实例之间细微的空间和逻辑关系。这些局限性阻碍了基于DLA的模型实现类似于人类阅读的逐步深入理解。在本工作中,我们提出了一种新颖的基于图的文档结构分析(gDSA)任务。该任务要求模型不仅检测文档元素,还要以图结构的形式生成空间和逻辑关系,从而能够以整体和直观的方式理解文档。针对这一新任务,我们构建了一个基于关系图的文档结构分析数据集(GraphDoc),包含80K文档图像和4.13M关系标注,使训练模型能够完成多项任务,如阅读顺序、层次结构分析以及复杂的元素间关系推理。此外,我们提出了文档关系图生成器(DRGG)来解决gDSA任务,在这一新任务和数据集上,该生成器以57.6%的mAP$_g$@0.5实现了强大的基准基线性能。我们希望这种文档结构的图形化表示能够标志着文档结构分析与理解领域的创新性进步。新数据集和代码将在https://yufanchen96.github.io/projects/GraphDoc公开发布。

英文摘要:

When reading a document, glancing at the spatial layout of a document is an initial step to understand it roughly. Traditional document layout analysis (DLA) methods, however, offer only a superficial parsing of documents, focusing on basic instance detection and often failing to capture the nuanced spatial and logical relations between instances. These limitations hinder DLA-based models from achieving a gradually deeper comprehension akin to human reading. In this work, we propose a novel graph-based Document Structure Analysis (gDSA) task. This task requires that model not only detects document elements but also generates spatial and logical relations in form of a graph structure, allowing to understand documents in a holistic and intuitive manner. For this new task, we construct a relation graph-based document structure analysis dataset (GraphDoc) with 80K document images and 4.13M relation annotations, enabling training models to complete multiple tasks like reading order, hierarchical structures analysis, and complex inter-element relation inference. Furthermore, a document relation graph generator (DRGG) is proposed to address the gDSA task, which achieves performance with 57.6% at mAP$_g$@0.5 for a strong benchmark baseline on this novel task and dataset. We hope this graphical representation of document structure can mark an innovative advancement in document structure analysis and understanding. The new dataset and code will be made publicly available at https://yufanchen96.github.io/projects/GraphDoc.

机构、总结、中文摘要均由AI分析而来,可能存在错误,请以论文正文为准。

↑