DocAtlas: Multilingual Document Understanding Across 80+ Languages

• URL: http://arxiv.org/abs/2605.12623v2
• Date: Thu, 21 May 2026 05:33:02 GMT
• Title: DocAtlas: Multilingual Document Understanding Across 80+ Languages
• Abstract summary: We introduce DocAtlas, a framework that constructs high-fidelity OCR datasets and benchmarks covering 82 languages.<n>Our dual pipelines, differential rendering of native DOCX documents and synthetic-based generation for right-to-left scripts produce precise structural annotations.
• Score: 58.715440331861295
• License: http://creativecommons.org/licenses/by-nc-sa/4.0/
• Abstract: Multilingual document understanding remains limited for low-resource languages due to scarce training data and model-based annotation pipelines that perpetuate existing biases. We introduce DocAtlas, a framework that constructs high-fidelity OCR datasets and benchmarks covering 82 languages and 9 evaluation tasks. Our dual pipelines, differential rendering of native DOCX documents and synthetic LaTeX-based generation for right-to-left scripts produce precise structural annotations in a unified DocTag format encoding layout, text, and component types, without learned models for core annotation. Evaluating 16 state-of-the-art models reveals persistent gaps in low-resource scripts. We show that Direct Preference Optimization (DPO) using rendering-derived ground truth as positive signal achieves stable multilingual adaptation, improving both in-domain (+1.9%) and out-of-domain (+1.8%) accuracy without measurable base-language degradation, where supervised fine-tuning degrades out-of-domain performance by up to 21%. Our best variant, DocAtlas-DeepSeek, improves +1.7% over the strongest baseline. Code is available at https://github.com/ahmedheakl/DocAtlas .

Related papers

• MORE: A Multilingual Document Parsing Benchmark and Evaluation [13.789730225958877]
  MORE is a large-scale benchmark designed for multilingual document parsing evaluation.<n>It covers 149 languages, making it the most linguistically diverse benchmark to date.<n>All samples are curated from real-world documents via a model-assisted, human-refined annotation pipeline.
  arXiv  Detail & Related papers  (2026-07-03T04:59:12Z)
• BabelDOC: Better Layout-Preserving PDF Translation via Intermediate Representation [9.979324579114392]
  Existing document translation pipelines face a tension between linguistic processing and layout preservation.<n>We introduce BabelDOC, an Intermediate Representation (IR)-based framework for layout-preserving PDF translation.<n>BabelDOC decouples visual layout metadata from semantic content, enabling document-level translation operations.
  arXiv  Detail & Related papers  (2026-05-11T16:56:44Z)
• MDPBench: A Benchmark for Multilingual Document Parsing in Real-World Scenarios [72.8160644291677]
  We introduce Multilingual Document Parsing Benchmark, the first benchmark for multilingual digital and photographed document parsing.<n> MDPBench comprises 3,400 document images spanning 17 languages, diverse scripts, and varied photographic conditions.
  arXiv  Detail & Related papers  (2026-03-30T07:47:46Z)
• Bilingual Text-to-Motion Generation: A New Benchmark and Baselines [52.71312720094036]
  We introduce BiHumanML3D, the first bilingual text-to-motion benchmark, constructed via LLM-assisted annotation and rigorous manual correction.<n>We also propose a simple yet effective baseline, Bilingual Motion Diffusion synthesis (BiMD), featuring Cross-Lingual Alignment (CLA)<n>BiMD with CLA achieves an FID of 0.045 vs. 0.169 and R@3 of 82.8% vs. 80.8%, significantly outperforms monolingual diffusion models and translation baselines on BiHumanML3D.
  arXiv  Detail & Related papers  (2026-03-26T08:48:27Z)
• Diffusion-Pretrained Dense and Contextual Embeddings [3.1271035100823643]
  pplx-embed is a family of multilingual embedding models that employ multi-stage contrastive learning on a diffusion-pretrained language model backbone for web-scale retrieval.<n>We release two model types: pplx-embed-v1 for standard retrieval, and pplx-embed-context-v1 for contextualized embeddings.
  arXiv  Detail & Related papers  (2026-02-11T18:59:08Z)
• LEMUR: A Corpus for Robust Fine-Tuning of Multilingual Law Embedding Models for Retrieval [18.46710400838861]
  Large language models (LLMs) are increasingly used to access legal information.<n>Yet, their deployment in multilingual legal settings is constrained by unreliable retrieval and the lack of domain-adapted, open-embedding models.<n>We introduce LEMUR, a large-scale multilingual corpus of EU environmental legislation constructed from 24,953 official EUR-Lex PDF documents covering 25 languages.
  arXiv  Detail & Related papers  (2026-02-10T09:20:24Z)
• IndicDLP: A Foundational Dataset for Multi-Lingual and Multi-Domain Document Layout Parsing [8.447520506307567]
  IndicDLP is a large-scale foundational document layout dataset spanning 11 representative Indic languages alongside English and 12 common document domains.<n>Our experiments demonstrate that fine-tuning existing English models on IndicDLP significantly boosts performance.<n>Models trained on IndicDLP generalize well beyond Indic layouts, making it a valuable resource for document digitization.
  arXiv  Detail & Related papers  (2025-12-23T10:49:37Z)
• DCAD-2000: A Multilingual Dataset across 2000+ Languages with Data Cleaning as Anomaly Detection [71.97939405401961]
  We introduce DCAD-2000 (Data Cleaning as Anomaly Detection), a large-scale multilingual corpus constructed from Common Crawl data and existing multilingual sources.<n>DCAD-2000 covers 2,282 languages, 46.72TB of text, and 8.63 billion documents, spanning 155 high- and medium-resource languages and 159 writing scripts.<n>By fine-tuning LLMs on DCAD-2000, we demonstrate notable improvements in data quality, robustness of the cleaning pipeline, and downstream performance.
  arXiv  Detail & Related papers  (2025-02-17T08:28:29Z)
• RustRepoTrans: Repository-level Code Translation Benchmark Targeting Rust [50.65321080814249]
  RustRepoTrans is the first repository-level context code translation benchmark targeting incremental translation.<n>We evaluate seven representative LLMs, analyzing their errors to assess limitations in complex translation scenarios.
  arXiv  Detail & Related papers  (2024-11-21T10:00:52Z)
• Translation and Fusion Improves Zero-shot Cross-lingual Information Extraction [18.926993352330797]
  We propose TransFusion, a framework in which models are fine-tuned to use English translations of low-resource language data. GoLLIE-TF, a cross-lingual instruction-tuned LLM for IE tasks, is designed to close the performance gap between high and low-resource languages.
  arXiv  Detail & Related papers  (2023-05-23T01:23:22Z)
• OneAligner: Zero-shot Cross-lingual Transfer with One Rich-Resource Language Pair for Low-Resource Sentence Retrieval [91.76575626229824]
  We present OneAligner, an alignment model specially designed for sentence retrieval tasks. When trained with all language pairs of a large-scale parallel multilingual corpus (OPUS-100), this model achieves the state-of-the-art result. We conclude through empirical results and analyses that the performance of the sentence alignment task depends mostly on the monolingual and parallel data size.
  arXiv  Detail & Related papers  (2022-05-17T19:52:42Z)
• LAWDR: Language-Agnostic Weighted Document Representations from Pre-trained Models [8.745407715423992]
  Cross-lingual document representations enable language understanding in multilingual contexts. Large pre-trained language models such as BERT, XLM and XLM-RoBERTa have achieved great success when fine-tuned on sentence-level downstream tasks.
  arXiv  Detail & Related papers  (2021-06-07T07:14:00Z)
• Unsupervised Domain Adaptation of a Pretrained Cross-Lingual Language Model [58.27176041092891]
  Recent research indicates that pretraining cross-lingual language models on large-scale unlabeled texts yields significant performance improvements. We propose a novel unsupervised feature decomposition method that can automatically extract domain-specific features from the entangled pretrained cross-lingual representations. Our proposed model leverages mutual information estimation to decompose the representations computed by a cross-lingual model into domain-invariant and domain-specific parts.
  arXiv  Detail & Related papers  (2020-11-23T16:00:42Z)

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.