WeVisDoc is a two‑stage data‑centric framework designed to improve end‑to‑end document parsing. Stage I expands coverage by adding heterogeneous data and applying structure‑preserving degradation synthesis, while Stage II evaluates residual errors with a held‑out probe and uses those diagnostics to target data construction and token budget reallocation. The resulting WeVisDoc‑4B model achieves an overall score of 95.38 on OmniDocBench v1.6 and outperforms competing parsers across all evaluated settings, with Stage II delivering notable gains on degraded tracks.
arXiv:2605.19309v4 Announce Type: replace Abstract: Document Layout Analysis (DLA) pipelines provide structured page representations for retrieval-augmented generation, long-document question answeri...
The paper introduces the Structural Semantic Unit (SSU) and the Coverage, Overlap, Trespass, and Excess (COTe) score as a new framework for evaluating Document Layout Analysis (DLA) models. Unlike traditional metrics such as IoU, F1, or mAP, which are tailored to 2D projections of 3D space, COTe focuses on the semantic structure of printed media and is decomposable to reveal specific failure modes like breaching semantic boundaries or redundant parsing. Experiments on five common DLA models across three datasets show that COTe is more informative and robust—especially under granularity mismatches—than F1, and the authors provide an SSU-labelled dataset and a Python library to facilitate adoption.
arXiv:2607. 10400v1 Announce Type: cross Abstract: Vision language models (VLMs) have achieved strong performance on visual document understanding benchmarks such as DocVQA, ChartQA, and MMLongBench-Doc.
arXiv:2606. 01393v1 Announce Type: cross Abstract: Document parsing and recognition are fundamental capabilities for vision-language models (VLMs) and document processing systems.
TeleOCR is a unified framework for document parsing that tackles challenges in both decoupled and end-to-end Vision‑Language Models. It introduces deformation‑aware learning to handle geometric distortions, an adaptive sampling mechanism for complex layouts, and a content‑structure decoupled strategy to model formula grammars and table structures. The approach achieves state‑of‑the‑art results on multiple benchmarks, including top placement in the ICDAR 2026 Sci‑ImageMiner Challenge.
arXiv:2607. 13639v1 Announce Type: cross Abstract: We introduce OvisOCR2, a 0.
arXiv:2608. 15064v1 Announce Type: new Abstract: Parsing visual documents into machine-readable representations is fundamental to document intelligence.
Transforming table-form documents into machine-processable records requires recovering not only their visible content but also the multilevel structure that organizes it. However, existing benchmarks evaluate either holistic document outputs or conventional table grids, and their aggregate scores provide little insight into where structural failures occur.
ShredBench is a new benchmark for evaluating multimodal large language models (MLLMs) on the task of reconstructing documents from shredded fragments. It uses an automated pipeline that generates fragmented documents from Markdown, covering English, Chinese, code, and table scenarios with varying fragmentation levels. Experiments show that while MLLMs perform well on intact documents, their performance drops sharply when documents are shredded, revealing a gap in fine‑grained cross‑modal reasoning.
Document understanding systems are typically evaluated as separate components via layout analysis, optical text recognition, reading-order prediction, and image captioning. This separation leaves scie...
arXiv:2610.20335v1 Announce Type: new Abstract: Document understanding systems are typically evaluated as separate components via layout analysis, optical text recognition, reading-order prediction,...
The paper introduces MOTe, a decomposable error metric for Document Layout Analysis (DLA) that addresses shortcomings of traditional IoU‑based F1 scores. It proposes the Structural Semantic Unit (SSU) to focus on semantic structure and demonstrates that MOTe correlates more strongly with parsing‑induced Character Error Rates (CER) than F1, providing reliable upper and lower bounds on CER for model predictions. A Python library is released to facilitate the adoption of MOTe in DLA projects.