https://ym-xu.github.io/survey/A daily-updated map of document understanding research — layout, OCR, tables, charts, document QA, long documents, retrieval and agents — with short editor's notes, classic papers and core benchmarks.enWed, 11 Nov 2026 15:23:29 +0000https://arxiv.org/abs/2611.13752https://arxiv.org/abs/2611.13752Tue, 10 Nov 2026 00:00:00 +0000OCR & Text RecognitionHandwriting & Historical<p><b>Fonseca Cacho, Jorge Ramón</b> · <a href="https://github.com/SniperJF/zone-robust-ocr-align">code</a></p><p>OCR post-processing needs OCR text aligned word by word with its ground truth. Data sets rarely provide that mapping, and a monotone aligner can lose its place when zoning errors permute the reading order. We revisit this problem, the Zoning Problem of our 2019 paper, with a protocol that tests correctness under such p…</p>https://arxiv.org/abs/2611.13609https://arxiv.org/abs/2611.13609Tue, 10 Nov 2026 00:00:00 +0000OCR & Text Recognition<p><b>Mannon Ochilov, Oybek Narzullayev, Bahodir Zaripov</b> · <a href="https://github.com/OybekNarzullaev/byt5-uzbek-postocr">code</a></p><p>In Uzbek official and technical documents, a large share of OCR errors is caused not by poor image quality but by characters that are visually identical across the two alphabets in use. Because the script transition is still under way, Cyrillic and Latin appear on the same page and often within a single sentence, and i…</p>https://arxiv.org/abs/2611.13217https://arxiv.org/abs/2611.13217Mon, 09 Nov 2026 00:00:00 +0000Agents & ReasoningHandwriting & Historical<p><b>Rayed AlGhamdi</b></p><p>This study examines the reliability of Claude Code, an agentic deployment of the Claude Opus 4.6 large language model, as a grading system for handwritten university exams. Four research questions addressed agreement with expert human grading on total scores (RQ1), agreement across five question types from multiple cho…</p>https://arxiv.org/abs/2611.11886https://arxiv.org/abs/2611.11886Mon, 09 Nov 2026 00:00:00 +0000Agents & ReasoningDocument QA & VLMs<p><b>Zhefan Wang, Da Li, Keping Bi, Nanqing Dong</b></p><p>Agricultural documents, such as prefectural technical reports and crop calendars, are critical knowledge resources. However, their complex table layouts, significant variations in document length, and dense multi-modal elements pose substantial challenges for automated understanding. To address these issues, the NTCIR-…</p>https://arxiv.org/abs/2611.12957https://arxiv.org/abs/2611.12957Sat, 07 Nov 2026 00:00:00 +0000Agents & ReasoningDocument QA & VLMsLayout & Reading Order<p><b>Bingdian Zhang, Zixuan Li, Long Bai, Xiaolong Jin, et al.</b></p><p>VLM-based agents solve complex tasks by iteratively interacting with environments, and their effectiveness depends heavily on how the environment exposes information and provides tools for acting on it. However, constructing an effective environment for complex multimodal document question answering remains challenging…</p>https://arxiv.org/abs/2611.10679https://arxiv.org/abs/2611.10679Sat, 07 Nov 2026 00:00:00 +0000Parsing & Conversion<p><b>Halimat Afolabi, Daeyoung Lee</b></p><p>Document parsing benchmarks are often treated as stable measures of model quality, yet rankings may depend substantially on benchmark design choices. In this work, we present RankShift, a systematic study comparing OCR and vision-language models (VLMs) under controlled variation in dataset composition, matching strateg…</p>https://arxiv.org/abs/2611.10475https://arxiv.org/abs/2611.10475Sat, 07 Nov 2026 00:00:00 +0000Agents & Reasoning<p><b>Jinyoung Park, Yupan Huang, Shimin Zhang, Jiayu Ding, et al.</b> · <i>NeurIPS 2026</i></p><p>Microsoft Word is a central environment for office automation, and the faithful execution of its editing operations is an important capability for document agents. Unlike plain-text editing, editing in Word modifies a structured object model, where the requested effect is realized in object properties. Existing benchma…</p>https://arxiv.org/abs/2611.08091https://arxiv.org/abs/2611.08091Thu, 05 Nov 2026 00:00:00 +0000Handwriting & HistoricalOCR & Text Recognition<p><b>Ismail El Jamiy, Khadija Msalek, Guillaume Loizelet, Farah Benamara, et al.</b></p><p>Despite recent progress in handwritten text recognition, medieval Arabic scientific manuscripts remain an open challenge for current vision-language models (VLMs) due to cursive ligatures, variable diacritization, and more importantly the lack of training data. This paper aims to address this gap, by proposing the firs…</p>https://arxiv.org/abs/2611.07810https://arxiv.org/abs/2611.07810Thu, 05 Nov 2026 00:00:00 +0000Handwriting & Historical<p><b>Peirong Zhang, Zhenhua Yang, Jiaxin Zhang, Moises Diaz, et al.</b></p><p>Handwritten signature verification (HSV) is a critical biometric technique in legal and forensic authentication, whose efficacy hinges on the discriminability of signature representations. Current HSV methods predominantly adopt supervised learning to enhance representation quality, leaving self-supervised learning lar…</p>https://arxiv.org/abs/2611.06449https://arxiv.org/abs/2611.06449Wed, 04 Nov 2026 00:00:00 +0000Parsing & ConversionDocument QA & VLMs<p><b>Boyu Niu, Yingjie Xiong, Shurui Li, Linhan Pang, et al.</b></p><p>Traditional document parsing systems use Markdown as their output representation, which is effective for linear text but does not fully preserve fine-grained spatial structure, table geometry, visual hierarchy, or rendering semantics. We show that HTML is advantageous for heterogeneous and layout-sensitive documents wh…</p>https://arxiv.org/abs/2611.05539https://arxiv.org/abs/2611.05539Tue, 03 Nov 2026 00:00:00 +0000OCR & Text Recognition<p><b>Jorge Ramón Fonseca Cacho</b> · <a href="https://github.com/SniperJF/ocr-edit-gate-release">code</a></p><p>LLMs can repair noisy OCR transcripts of historical documents, but they also change text that was already correct. We propose EditVeto, a veto gate placed after a full-context LLM rewrite. The rewrite is character-aligned to the OCR input, its differences are split into blocks, and a 0.6B-parameter gate keeps or revert…</p>https://arxiv.org/abs/2611.04874https://arxiv.org/abs/2611.04874Tue, 03 Nov 2026 00:00:00 +0000OCR & Text Recognition<p><b>Zhiheng Li, Zongyang Ma, Jiaxian Chen, Cheng Cui, et al.</b> · <a href="https://github.com/zhihengli-casia/Awesome-OCR-in-the-Foundation-Model-Era">code</a></p><p>Optical Character Recognition (OCR) is among the oldest and most widely deployed applications of artificial intelligence. This survey reviews its development, focusing on the Foundation Model Era. We define OCR through faithful transcription: within a specified visual scope, OCR should completely reconstruct the text, …</p>https://arxiv.org/abs/2611.03339https://arxiv.org/abs/2611.03339Mon, 02 Nov 2026 00:00:00 +0000OCR & Text RecognitionAgents & Reasoning<p><b>Yakoub Bazi, Sarah Alsaleh, Mohamad M. Al Rahhal, Mohamed A. Mekhtiche, et al.</b> · <a href="https://github.com/yakoubbazi/ArabGeoText">code</a></p><p>Geospatial text spotting involves locating and transcribing geographic labels, such as road names, district names, and landmarks, in raster maps. These labels are often small, densely arranged, and arbitrarily oriented. Although Arabic text recognition has been studied extensively in documents and natural scenes, Arabi…</p>https://arxiv.org/abs/2611.03097https://arxiv.org/abs/2611.03097Mon, 02 Nov 2026 00:00:00 +0000Agents & ReasoningOCR & Text Recognition<p><b>Tiansheng Hu, Yichuan Wang, James V. Talwar, Nicholas Tomlin, et al.</b></p><p>Complex documents such as scientific papers contain rich information in tables, figures, and equations; however, existing deep research (DR) agents typically reduce papers to OCR-extracted text. We instead ask whether vision-language models (VLMs) can enable page-native deep research: directly retrieving and reasoning …</p>https://arxiv.org/abs/2611.03023https://arxiv.org/abs/2611.03023Mon, 02 Nov 2026 00:00:00 +0000Charts, Figures & Infographics<p><b>Haiyang Zhang, Sicheng Song, Zixin Chen, Jie Gu, et al.</b></p><p>Misleading charts can lead vision-language models to conclusions that conflict with underlying data. Existing work focuses on inference-time strategies, while training-time supervision remains less explored. We present VizDAGuide, a DAG-structured supervision method for misleading-chart question answering. The method r…</p>https://arxiv.org/abs/2611.00242https://arxiv.org/abs/2611.00242Fri, 30 Oct 2026 00:00:00 +0000Retrieval & RAG<p><b>Sheikh Tahmim Mehdi, Muminul Islam Jami</b></p><p>A deep understanding of scientific literature often relies on interpreting figures that contain key information beyond the main text. Leveraging Large Language Models (LLMs) to extract insights from these figures can significantly accelerate knowledge discovery. Existing works on Scientific Visual Question Answering re…</p>https://arxiv.org/abs/2610.41426https://arxiv.org/abs/2610.41426Fri, 30 Oct 2026 00:00:00 +0000OCR & Text Recognition<p><b>Zhibin Ma, Pengwen Dai, Zhigang Yao, Xugong Qin, et al.</b></p><p>Existing Scene Text Recognition (STR) models are inherently coupled with a specific decoding strategy by their training paradigms, limiting their flexibility in deployment across diverse recognition scenarios. In this paper, we propose FlexReader, an STR model that supports multiple decoding strides, enabling a single …</p>https://arxiv.org/abs/2610.41360https://arxiv.org/abs/2610.41360Fri, 30 Oct 2026 00:00:00 +0000Retrieval & RAGDocument QA & VLMs<p><b>Chongzhi Zhang, Yuqi Zuo, Aixin Sun</b></p><p>In modern information retrieval (IR), documents are increasingly encountered as structured sequences of visually rich, paginated units, rather than as flat text. Example documents encountered in our daily life include Word files, PDFs, and slide decks. Free text-centric retrieval systems either ignore this structure en…</p>https://arxiv.org/abs/2610.41335https://arxiv.org/abs/2610.41335Fri, 30 Oct 2026 00:00:00 +0000Retrieval & RAG<p><b>Ziniu Liu, Xiaoyang Chen, Dezhi Ye, Jie Liu, et al.</b></p><p>Retrieval-augmented generation (RAG) over visually rich documents must find both relevant content and the evidence that supports an answer. Screenshot retrieval preserves images and page structure, but written facts within screenshots also need to be searchable. Moreover, topical relevance alone does not identify which…</p>https://arxiv.org/abs/2610.41254https://arxiv.org/abs/2610.41254Fri, 30 Oct 2026 00:00:00 +0000OCR & Text Recognition<p><b>Brian Tang, Kang G. Shin</b></p><p>While modern Vision-Language Models (VLMs) have been remarkably successful in general scene text recognition, they consistently struggle with the long-tail distribution of text that is degraded by motion blur or low resolution. These limitations are particularly apparent in egocentric applications like smart glasses, w…</p>https://arxiv.org/abs/2610.40233https://arxiv.org/abs/2610.40233Thu, 29 Oct 2026 00:00:00 +0000Handwriting & HistoricalOCR & Text Recognition<p><b>Chutong Meng, Alec J. Bateman, Antonios Anastasopoulos</b></p><p>While mining parallel sentences from web pages has proven effective for low-resource machine translation, the majority of the world's languages have little to no web presence. What they do have, though, are resources that are the result of centuries of efforts by linguists, such as grammars, dictionaries, and other doc…</p>https://arxiv.org/abs/2610.39308https://arxiv.org/abs/2610.39308Thu, 29 Oct 2026 00:00:00 +0000OCR & Text RecognitionParsing & ConversionLong & Multi-page Documents<p><b>Shu Wei, Jingjing Wu, Qunyi Xie, Lingshu Zhang, et al.</b></p><p>Unlimited OCR ushers the era of long-horizon parsing, substantially mitigating the growth of working memory as document lengthens. However, as document length scales to hundreds of pages, the visual context grows substantially, sharply increasing the computational cost of each decoding step and limiting the scalability…</p>https://arxiv.org/abs/2610.39301https://arxiv.org/abs/2610.39301Thu, 29 Oct 2026 00:00:00 +0000Agents & Reasoning<p><b>Yunhang Shen, Zhangquan Chen, Xiaotian Li, Zhongpeng Cai, et al.</b></p><p>We present Youtu-Parsing-Omni, an omni-modal parser that converts documents, images, audio, and video into structured records grounded in space and time, providing a perception layer for agents that read, locate, and cite evidence across modalities. Its design rests on one encoder for every modality and one schema for …</p>https://arxiv.org/abs/2610.38924https://arxiv.org/abs/2610.38924Wed, 28 Oct 2026 00:00:00 +0000Agents & Reasoning<p><b>Ritiz Tambi, Sahar Shayegan, Joshua Meyer, Sun Kim, et al.</b></p><p>Organisations keep decades of financial records as scanned PDFs on shared drives, and AI agents can now reach those drives through Model Context Protocol (MCP) servers. Existing benchmarks do not mirror this production setting because they simplify the task. In some, the relevant documents are handed to the agent, so i…</p>https://arxiv.org/abs/2610.36956https://arxiv.org/abs/2610.36956Tue, 27 Oct 2026 00:00:00 +0000Handwriting & Historical<p><b>Tamas Kovacs, Daniel Luger</b></p><p>Handwritten text recognition (HTR) for the registers of Pope Honorius III depends on modern expert transcriptions, which exist only for regnal years 1-2 of Reg. Vat. 9. For the remaining registers, nineteenth-century scholarship offers Pressutti's regesta and partial editions. On fol. 1-203, where both layers exist, we…</p>https://arxiv.org/abs/2610.36929https://arxiv.org/abs/2610.36929Tue, 27 Oct 2026 00:00:00 +0000Document QA & VLMs<p><b>Saifullah Saifullah, David Dembinsky, Stefan Agne, Andreas Dengel, et al.</b> · <a href="https://github.com/saifullah3396/docxeval/">code</a></p><p>Multimodal transformer models have driven unprecedented performance breakthroughs in modern document understanding. However, the inherent black-box nature of these models remains a major obstacle to their safe deployment in high-stakes sectors. Despite significant progress in explainability for vision and language task…</p>https://arxiv.org/abs/2610.36850https://arxiv.org/abs/2610.36850Tue, 27 Oct 2026 00:00:00 +0000Handwriting & HistoricalOCR & Text Recognition<p><b>Preethi S, Meeradevi T, Arun Prakash A</b></p><p>Evaluation of Tamil handwritten character recognition (HCR) has largely focused on clean test images, and a standardized multi-corruption robustness benchmark for isolated Tamil HCR has not been established. We introduce uTHCD-C, a corruption-robustness benchmark based on the unconstrained Tamil Handwritten Character D…</p>https://arxiv.org/abs/2610.36759https://arxiv.org/abs/2610.36759Tue, 27 Oct 2026 00:00:00 +0000Agents & ReasoningDocument QA & VLMs<p><b>Jing Huang, Zhiya Tan, Changtao Miao, Shutao Gong, et al.</b></p><p>Multilingual multi-hop document question answering (MMHop-DocQA) requires locating and integrating evidence across complex layouts, document pages, and diverse languages. Existing approaches include single-pass processing and adaptive tool use, but learning to coordinate evidence acquisition, cross-page integration, an…</p>https://arxiv.org/abs/2610.36540https://arxiv.org/abs/2610.36540Tue, 27 Oct 2026 00:00:00 +0000Handwriting & HistoricalOCR & Text Recognition<p><b>Ousmane Hamadou, Jean Louis Kedieng Ebongue Fendji</b> · <i>BMVC 2026 Workshop</i></p><p>Handwritten civil-status registers across Sub-Saharan Africa are being photographed, but rarely transcribed, which leaves their content unsearchable. Local administrations that hold these records typically lack GPUs, reliable connectivity, and the budget for external digitization services. We study whether a frugal pip…</p>https://arxiv.org/abs/2610.36444https://arxiv.org/abs/2610.36444Tue, 27 Oct 2026 00:00:00 +0000Agents & ReasoningRetrieval & RAG<p><b>Zekai Chen, Hao Zhu, Yu Liu, Xunkai Li, et al.</b> · <a href="https://github.com/Ereshkigal1/RouteDRAG">code</a></p><p>Web-based applications increasingly rely on understanding long multimodal documents, where answering questions often requires combining information across pages, tables, and figures. Agentic retrieval-augmented generation addresses these tasks through iterative retrieval and reasoning. However, Agentic execution is not…</p>https://arxiv.org/abs/2610.35542https://arxiv.org/abs/2610.35542Mon, 26 Oct 2026 00:00:00 +0000Agents & ReasoningRetrieval & RAG<p><b>Wenxin Zhao, Jiachuan Wang, Xinyi Zhu, Jiake Ge, et al.</b></p><p>Long-context multimodal document question answering (QA) has become an important yet challenging task for modern vision-language models (VLMs), as it requires reasoning over evidence dispersed across pages and modalities within limited input budgets. Current pipelines adopt retrieval-augmented generation (RAG) to selec…</p>https://arxiv.org/abs/2610.35240https://arxiv.org/abs/2610.35240Mon, 26 Oct 2026 00:00:00 +0000OCR & Text RecognitionRetrieval & RAGAgents & Reasoning<p><b>Radha Krishna Deshpande, Ajoy Mondal, Akash Manna, C V Jawahar</b></p><p>India's historical public archives present severe computer vision and document image analysis challenges: decades of parliamentary records survive only as degraded, multi-column scanned pages featuring mixed scripts, complex layouts, and dense visual noise where OCR failure is the primary obstacle to information retrie…</p>https://arxiv.org/abs/2610.35085https://arxiv.org/abs/2610.35085Mon, 26 Oct 2026 00:00:00 +0000Handwriting & Historical<p><b>Tamas Kovacs, Daniel Luger</b></p><p>Unlocking medieval archives at scale is bottlenecked by the cost of line-level diplomatic transcriptions and manual abbreviation expansion. Focusing on the registers of Pope Honorius III (1216-1227) over 5,000 letters surviving in full only as manuscript images - we show that existing, line-break-free expert transcript…</p>https://arxiv.org/abs/2610.34879https://arxiv.org/abs/2610.34879Mon, 26 Oct 2026 00:00:00 +0000Layout & Reading Order<p><b>Kehua Chen, Mingyi Chen</b></p><p>Document layout analysis (DLA) faces a long-standing dilemma: existing methods either capture rich local details at prohibitive computational cost (e.g., Transformer-based approaches with quadratic complexity) or achieve efficient global modeling but sacrifice fine-grained structural awareness (e.g., state-space models…</p>https://arxiv.org/abs/2610.34148https://arxiv.org/abs/2610.34148Sun, 25 Oct 2026 00:00:00 +0000OCR & Text Recognition<p><b>Majid Adibian, Hossein Saberi, Hossein Zeinali</b></p><p>Optical character recognition (OCR) systems depend strongly on the scale and diversity of their training data. One approach to document OCR first analyzes the page layout and extracts text-line images, which are then transcribed by a recognition model. Multilingual line-level recognition involving Persian, Arabic, and …</p>https://arxiv.org/abs/2610.32463https://arxiv.org/abs/2610.32463Fri, 23 Oct 2026 00:00:00 +0000OCR & Text Recognition<p><b>Sugam Karki, Raúl Vázquez, Yova Kementchedjhieva</b></p><p>While large vision-language models excel at optical character recognition, it remains unclear how their underlying vision encoders process text and its meaning, and whether they support anything resembling reading comprehension. We systematically study transformer-based vision encoders, using cross-modal alignment betw…</p>https://arxiv.org/abs/2610.32423https://arxiv.org/abs/2610.32423Fri, 23 Oct 2026 00:00:00 +0000Agents & Reasoning<p><b>Somraj Gautam, Gaurav Harit</b></p><p>Recent advances in Vision-Language Models (VLMs) have significantly improved Visual Question Answering (VQA) performance, yet their reasoning over structured content such as tables remains inconsistent, often plagued by spatial misalignment and logical hallucinations. Existing approaches for improving table VQA typical…</p>https://arxiv.org/abs/2611.09953https://arxiv.org/abs/2611.09953Thu, 22 Oct 2026 00:00:00 +0000Handwriting & Historical<p><b>Azam Nouri</b></p><p>Handwritten character recognition systems can achieve high accuracy while retaining concentrated errors among visually similar classes. Motivated by residual letter confusions in earlier geometry-based recognition work, we investigate whether these errors can be addressed through the training objective. We introduce Co…</p>https://arxiv.org/abs/2611.00092https://arxiv.org/abs/2611.00092Thu, 22 Oct 2026 00:00:00 +0000Charts, Figures & InfographicsParsing & Conversion<p><b>Aman Rangapur, Antonio Jimeno Yepes, Daniel Bustamante, Nutchapol Dendumrongsup, et al.</b></p><p>We introduce OCRia-1, a 4B vision-language model (VLM) for document parsing that excels at preserving table structure, extracting chart data and retaining text formatting. OCRia-1 handles content that has traditionally been challenging in document parsing: merged cells and spanning headers, tables with thousands of cel…</p>https://arxiv.org/abs/2610.31385https://arxiv.org/abs/2610.31385Thu, 22 Oct 2026 00:00:00 +0000Long & Multi-page DocumentsDocument QA & VLMs<p><b>Yufei Tong, Guanjie Cheng, Tianle Wu, Leilei Zhang, et al.</b></p><p>Visual document understanding is essential for acquiring knowledge from visually rich documents and is attracting growing research interest. Existing surveys emphasize architectures or technical approaches, with limited attention to how task bottlenecks shape technical choices and research priorities. We present a task…</p>https://arxiv.org/abs/2610.30863https://arxiv.org/abs/2610.30863Thu, 22 Oct 2026 00:00:00 +0000Imaging & Generation<p><b>Yasas Savinda, Maheshi Buddhinee Dissanayake</b></p><p>Historical document binarization remains challenging when faded ink, bleed-through, and background degradation exhibit wavelength-dependent behavior across visible and near-infrared observations. This study investigates whether adaptive cross-spectral feature fusion and targeted structural guidance can improve multispe…</p>https://arxiv.org/abs/2610.30072https://arxiv.org/abs/2610.30072Wed, 21 Oct 2026 00:00:00 +0000Agents & Reasoning<p><b>Xinzhe Wang, Xudong Weng, Zang Li, Shaolei Zhang</b></p><p>Data agents built on ReAct-style loops now execute complex analytical workflows over structured data, yet their intermediate reasoning still runs almost entirely through text. Human analysts instead move between numbers, tables, and charts, reading visual evidence to expose trends and anomalies that text summaries drop…</p>https://arxiv.org/abs/2610.30026https://arxiv.org/abs/2610.30026Wed, 21 Oct 2026 00:00:00 +0000Agents & Reasoning<p><b>Yikai Zhao, Saurabh Pandey, Pradeep Kumar Misra</b></p><p>Enterprise agents act on documents. Before an agent can follow a policy, answer a compliance question or file a ticket, some ingestion layer must turn a Word file or PDF into tokens - and that layer is where an agent deployed in the wild silently loses information. We show that one layer is lost by every mainstream con…</p>https://arxiv.org/abs/2610.29523https://arxiv.org/abs/2610.29523Wed, 21 Oct 2026 00:00:00 +0000Retrieval & RAG<p><b>Markus Krimmel, Louis Milliken, Ekaterina Glazkova, Maximilian Schall, et al.</b> · <a href="https://huggingface.co/collections/perplexity-ai/pplx-embed-v2">code</a></p><p>In this report, we introduce pplx-embed-v2-late, a pair of 0.6B and 9B multilingual and multimodal late-interaction models. Both models achieve state-of-the-art results within their respective size classes across a range of text, image, and visual document retrieval benchmarks. On visual document retrieval tasks, pplx-…</p>https://arxiv.org/abs/2610.29188https://arxiv.org/abs/2610.29188Wed, 21 Oct 2026 00:00:00 +0000Parsing & ConversionTables<p><b>Eunbin Kim, Hyeonseok Kang, Sangkeun Jung</b> · <i>EMNLP 2026</i></p><p>Tables in real-world documents, including scientific papers, financial reports, and statistical summaries, frequently use merged cells, multi-level headers, and sectioned layouts. Yet the lightweight formats LLMs handle best (Markdown, CSV) cannot express these structures, while expressive formats (HTML, LaTeX) impose …</p>https://arxiv.org/abs/2610.33066https://arxiv.org/abs/2610.33066Tue, 20 Oct 2026 00:00:00 +0000OCR & Text Recognition<p><b>Yu Xia, Yinglun Li, Yuhui Chen, Jinlong Huang, et al.</b> · <a href="https://huggingface.co/ATH-MaaS/OvisOCR3">code</a></p><p>We introduce OvisOCR3, a 0.8B model for a broad range of OCR-related tasks. OvisOCR3 unifies document parsing, general OCR, key information extraction (KIE), document figure captioning, and cross-page table reconstruction in a single model. We train the model on real and synthetic data through continued pretraining and…</p>https://arxiv.org/abs/2610.28957https://arxiv.org/abs/2610.28957Tue, 20 Oct 2026 00:00:00 +0000Long & Multi-page DocumentsRetrieval & RAG<p><b>Juan Terven, Diana-Margarita Córdova-Esparza, Julio-Alejandro Romero-González, Edgar Chavez-Urbiola, et al.</b></p><p>Scanning every passage in a long document seems thorough, but does it preserve more useful evidence or produce better answers? We compare 11 retrieval configurations spanning lexical and dense search, hybrid reranking, and full-document scanning with Jev, OpenAI Decisions, and Cloudflare Clef. Our evaluation uses 396 Q…</p>https://arxiv.org/abs/2610.28899https://arxiv.org/abs/2610.28899Tue, 20 Oct 2026 00:00:00 +0000Handwriting & Historical<p><b>Elona Shatri, Jiří Mayer, Jan Hajič</b></p><p>Synthesised data is a common strategy to address the lack of annotated handwritten music scores in Optical Music Recognition (OMR). Smashcima is a recent synthesis framework that assembles pages of handwritten scores by sampling glyph images from MUSCIMA++ and positioning them with hard-coded rules. In this work we ask…</p>https://arxiv.org/abs/2610.28282https://arxiv.org/abs/2610.28282Tue, 20 Oct 2026 00:00:00 +0000Agents & ReasoningCharts, Figures & Infographics<p><b>Qi Wang, Peijie Wang, Jun Cao, Haoyang Tong, et al.</b></p><p>Multimodal large language models (MLLMs) have substantially advanced chart understanding and reasoning, yet their performance remains highly uneven across chart types. While modern MLLMs perform well on common charts, their accuracy drops sharply on long-tail charts with diverse visual structures and reasoning requirem…</p>https://arxiv.org/abs/2610.28171https://arxiv.org/abs/2610.28171Tue, 20 Oct 2026 00:00:00 +0000Layout & Reading Order<p><b>Zhiyuan Zhao, Hengrui Kang, Guang Liang, Bin Wang, et al.</b> · <a href="https://github.com/opendatalab/DocLayout-YOLO/tree/release-v2">code</a></p><p>Accurate and efficient document layout analysis is critical for modern document parsing systems. However, although the YOLO framework represents a leading paradigm for the real-time detector, its adaptation to heterogeneous documents remains underexplored. To address this gap, we propose DocLayout-YOLO, a document-spec…</p>https://arxiv.org/abs/2610.27829https://arxiv.org/abs/2610.27829Tue, 20 Oct 2026 00:00:00 +0000Retrieval & RAG<p><b>Weijian Jian, Siyang Chen, Yuguang Yang, Yajun Zhang, et al.</b></p><p>Multimodal retrieval-augmented generation over visually rich documents requires evidence at a finer granularity than entire pages. Existing visual retrievers typically index or return full pages, whereas training region-level retrievers often requires costly query-to-region annotations or noisy model-generated labels. …</p>https://arxiv.org/abs/2610.27592https://arxiv.org/abs/2610.27592Tue, 20 Oct 2026 00:00:00 +0000Imaging & Generation<p><b>Xingsong Ye, Yongkun Du, Jiaxin Zhang, Chong Sun, et al.</b> · <i>NeurIPS 2026</i> · <a href="https://github.com/YesianRohn/TextRIRO">code</a></p><p>Aiming to modify textual content while preserving the original style, Scene Text Editing (STE) has significant practical value in many applications. STE typically requires paired images before and after editing for training. However, recent approaches rely heavily on synthetic paired data and complex disentanglement st…</p>https://arxiv.org/abs/2610.30177https://arxiv.org/abs/2610.30177Mon, 19 Oct 2026 00:00:00 +0000Retrieval & RAG<p><b>Haowen Xiang, Yibo Yan, Xin Zou, Shuliang Liu, et al.</b></p><p>Visual document retrieval relies on rich multi-vector representations to capture fine-grained evidence in document pages, but encoding and matching these representations can be computationally expensive. Existing compression and elastic representation methods expose smaller configurations, yet making multiple budgets a…</p>https://arxiv.org/abs/2610.27309https://arxiv.org/abs/2610.27309Mon, 19 Oct 2026 00:00:00 +0000OCR & Text RecognitionHandwriting & Historical<p><b>Paraskevi Platanou, John Pavlopoulos</b></p><p>Optical Character Recognition (OCR) and handwritten character recognition (HCR) in low-resource historical material collections often rely on data augmentation, yet augmentation strategies are typically evaluated almost exclusively through predictive accuracy. This leaves open whether a model's improved performance is …</p>https://arxiv.org/abs/2610.26362https://arxiv.org/abs/2610.26362Mon, 19 Oct 2026 00:00:00 +0000Information Extraction & FormsOCR & Text Recognition<p><b>Islam Mesabah, Ali Elhosary, Priyabanta Sandulu, Darko Obradovic, et al.</b></p><p>Document annotation is a fundamental yet labor-intensive step in building document information extraction systems. While Optical Character Recognition (OCR) and Large Language Models (LLMs) can automate parts of this process, we argue that annotation systems should be judged not by how completely they remove human effo…</p>https://arxiv.org/abs/2611.00078https://arxiv.org/abs/2611.00078Sun, 18 Oct 2026 00:00:00 +0000OCR & Text Recognition<p><b>Faizan Iqbal, Haq Nawaz Malik, Hafsah Pervaiz</b> · <a href="https://huggingface.co/Faizaniqbal/KoshurOCR-CRNN-Baseline">code</a></p><p>Optical character recognition (OCR) remains inaccessible to many low-resource scripts because annotated document images are unavailable and synthetic training data can fail to preserve script-specific visual structure. We present a data-centric framework that builds end-to-end OCR for such scripts from only an unannota…</p>https://arxiv.org/abs/2610.25440https://arxiv.org/abs/2610.25440Sun, 18 Oct 2026 00:00:00 +0000OCR & Text RecognitionHandwriting & Historical<p><b>Téa Wright, Laura X. Ma, Claudia Iron Hawk, Hannah J. Haynie, et al.</b></p><p>Building and expanding access to digital archives supports language reclamation and creates lasting resources for community revitalization. Designing effective tools to support these efforts requires taking into account the community-, language-, and document-specific context. We investigate how small ($\leq$3B) vision…</p>https://arxiv.org/abs/2610.25235https://arxiv.org/abs/2610.25235Sun, 18 Oct 2026 00:00:00 +0000OCR & Text RecognitionParsing & ConversionDocument QA & VLMs<p><b>Jumana Alsubhi, Ahmed Alhusayni, Abdulrahman Gharawi, Israa Hamdine, et al.</b></p><p>Choosing an OCR system for a multilingual archive requires balancing accuracy against deployment constraints. This study compares cloud and self-hosted systems on three tasks: bounding box detection on 48 Arabic, English, and French pages, text recognition on 315 pages across nine document types, and mathematical expre…</p>https://arxiv.org/abs/2610.24822https://arxiv.org/abs/2610.24822Sat, 17 Oct 2026 00:00:00 +0000Retrieval & RAGDocument QA & VLMs<p><b>Omar El Bachyr, Laura Maria Bernardy, Saad Ezzini, Jacques Klein, et al.</b> · <i>SIGIR</i></p><p>Enterprise documents such as financial reports are increasingly processed by automated systems, yet they are primarily designed for human reading and exhibit complex visual structure, combining dense text with tables, charts, diagrams, and rich layouts. These characteristics pose significant challenges for Retrieval-Au…</p>https://arxiv.org/abs/2610.24060https://arxiv.org/abs/2610.24060Fri, 16 Oct 2026 00:00:00 +0000Handwriting & HistoricalOCR & Text Recognition<p><b>Robin Hollifeldt, Tianru Zhang, Jessica Ström, Ekta Vats</b> · <i>TACL</i></p><p>Recovering obscured text in palimpsests, i.e. the manuscripts where the original writing (under-text) was erased and overwritten for reuse, is vital for preserving lost information. Due to the nature of the material, such a task can be highly challenging and time-consuming. This paper presents an approach towards autom…</p>