AraMS-28k: The Largest Publicly Released Line-Level Dataset of Historical Arabic Manuscripts with Margin and Insertion-Anchor Annotations

📄 arXiv: 2608.26921v1 📥 PDF

作者: Mohamed Guechaoui, Mohamed Diaa Zellagui, Souleyman Chaib, Sahraoui Dhelim

分类: cs.CV, cs.CL

发布日期: 2026-08-27

备注: Data and code available at this https URL and this https URL. Dataset: https://doi.org/10.5281/zenodo.22095333 Code: https://github.com/ArchaText/AraMS-28k-Dataset

💡 一句话要点

提出AraMS-28k数据集以促进历史阿拉伯手稿的识别与分析

🎯 匹配领域: 支柱九：具身大模型 (Embodied Foundation Models)

关键词: 阿拉伯手稿 数据集 行级标注 非线性阅读 文本识别 布局分析 机器学习

📋 核心要点

  1. 现有的历史阿拉伯手稿数据集缺乏足够的标注和多样性，限制了相关研究的进展。
  2. AraMS-28k数据集通过结合手抄本和印刷版，提供了详细的行级标注和插入锚点，解决了非线性阅读顺序的问题。
  3. 实验结果表明，使用该数据集进行手稿识别的基线模型在不同书籍间的泛化能力显著提升，验证了数据集的有效性。

📝 摘要（中文）

本文介绍了AraMS-28k，这是迄今为止公开发布的最大历史阿拉伯手稿行级数据集，包含14本书、3043页和28600条注释文本行。数据集涵盖了手抄本和印刷版，提供了主文本和边注的标注，并为边注提供了插入锚点，以恢复手稿的非线性阅读顺序。此外，数据集还提供了完整的元音标注和去标注的转录版本，支持阿拉伯手稿识别和布局分析的可重复研究。

🔬 方法详解

问题定义：本文旨在解决历史阿拉伯手稿数据集标注不足和缺乏多样性的问题，现有方法无法有效处理手稿的非线性阅读顺序。

核心思路：通过构建AraMS-28k数据集，结合手抄本和印刷版，提供详细的行级标注和插入锚点，以恢复手稿的真实阅读顺序。

技术框架：数据集构建采用RefLAM注释管道，首先将多模态LLM OCR与独立来源的干净转录对齐，然后每行经过人工审核，结合自动验证与专家监督。

关键创新：首次为历史阿拉伯手稿语料库发布了行级插入锚点注释，显著提升了对非线性阅读顺序的理解和处理能力。

关键设计：数据集提供了完整的元音标注和去标注的转录版本，确保了标注的一致性和准确性，同时采用了固定的训练、验证和测试划分。

🖼️ 关键图片

fig_0
fig_1
fig_2

📊 实验亮点

实验结果显示，使用AraMS-28k数据集的基线模型在手稿识别任务中取得了显著提升，特别是在跨书籍的泛化能力上，验证了数据集的有效性和实用性。

🎯 应用场景

AraMS-28k数据集在阿拉伯手稿识别、布局分析和阅读顺序恢复等领域具有广泛的应用潜力。其发布将为相关研究提供重要的基础数据，促进对历史文献的数字化和分析，推动人文学科与计算机科学的交叉研究。

📄 摘要（原文）

We introduce AraMS-28k, the largest publicly released line-level dataset of genuine historical Arabic manuscripts, comprising 14 books, 3,043 pages, and 28,600 annotated text lines (27,971 main-text, 629 margin). Thirteen books are hand-copied manuscripts spanning three script traditions -- Naskh, Ruq'ah, and Maghrebi -- and one is a lithographed printed edition included to broaden format diversity. Each line is labelled as main-text or margin, and margin lines that have an unambiguous attachment point in the main text are further annotated with an insertion anchor, recovering the manuscript's true non-linear reading order at line-level granularity -- to our knowledge the first such annotation released for a historical Arabic manuscript corpus. Because reference transcriptions are fully vocalised while manuscript hands are typically undiacritised, we release both the raw diacritised transcription and a diacritic-normalised counterpart for every line. The dataset was constructed with RefLAM, a reference-grounded annotation pipeline that aligns multimodal-LLM OCR against independently sourced clean transcriptions and routes every line through human review, combining automatic verification with expert oversight. We describe the construction and quality-control process, present the annotation schema, report dataset statistics at both the corpus and per-book level, and provide baseline HTR results using Kraken and HATFormer, including a cross-script generalisation gradient from in-distribution pages to fully unseen books. AraMS-28k is released with page images, line-level annotations, and fixed train/val/test splits under CC BY-NC-SA 4.0 to support reproducible research on Arabic manuscript recognition, layout analysis, and reading-order recovery.