用扩散模型并行生成文档文字，速度更快更准。
MinerU-Diffusion: Rethinking Document OCR as Inverse Rendering via Diffusion Decoding
- 将文档识别看作逆向渲染，用扩散模型并行生成文本
- 相比自回归方法提速3.2倍，长文档错误更少
- 适合需要高速高精度文档解析的场景
光学字符识别（OCR）已从行级转录发展到结构化文档解析，需恢复包含版式、表格和公式的长序列。尽管视觉语言模型取得进展，现有系统多依赖自回归解码，带来串行延迟并放大错误传播。本文从逆向渲染视角重思文档OCR，认为从左到右的生成是序列化的结果而非任务本质。为此提出MinerU-Diffusion，一种统一的基于扩散的框架，以视觉条件下的并行扩散去噪替代自回归串行解码。该框架采用分块扩散解码器与不确定性驱动的课程学习策略，实现稳定训练与高效长序列推理。大量实验表明，MinerU-Diffusion在保持鲁棒性的同时，解码速度比自回归基线快达3.2倍。在新提出的语义打乱基准测试中，其对语言先验依赖更低，视觉识别能力更强。
原文摘要 · Abstract (English)
Optical character recognition (OCR) has evolved from line-level transcription to structured document parsing, requiring models to recover long-form sequences containing layout, tables, and formulas. Despite recent advances in vision-language models, most existing systems rely on autoregressive decoding, which introduces sequential latency and amplifies error propagation in long documents. In this work, we revisit document OCR from an inverse rendering perspective, arguing that left-to-right causal generation is an artifact of serialization rather than an intrinsic property of the task. Motivated by this insight, we propose MinerU-Diffusion, a unified diffusion-based framework that replaces autoregressive sequential decoding with parallel diffusion denoising under visual conditioning. MinerU-Diffusion employs a block-wise diffusion decoder and an uncertainty-driven curriculum learning strategy to enable stable training and efficient long-sequence inference. Extensive experiments demonstrate that MinerU-Diffusion consistently improves robustness while achieving up to 3.2x faster decoding compared to autoregressive baselines. Evaluations on the proposed Semantic Shuffle benchmark further confirm its reduced dependence on linguistic priors and stronger visual OCR capability.
Thank you to arXiv for use of its open access interoperability. PaperDance 不是 arXiv 官方产品；中文卡片由大模型生成，请以原文为准。