PP-OCRv5：一个 500 万参数的专用模型在 OCR 任务上媲美十亿参数视觉语言模型
PP-OCRv5: A Specialized 5M-Parameter Model Rivaling Billion-Parameter Vision-Language Models on OCR Tasks
对做视觉研究的人而言，这篇论文给出了一个可复用的结论：在 OCR 这类结构化文本任务上，用数据难度、标注准确度、数据多样性三个维度做数据筛选与配比，可以让 5M 参数的两阶段检测+识别流水线逼近十亿参数 VLM，同时定位更准、幻觉更少；做 VLM 文本定位或幻觉抑制时，可把数据侧优化作为对照基线。
原文依据 with a sufficient volume of high-quality, accurately labeled, and diverse data, the performance ceiling for traditional, efficient two-stage OCR pipelines is far higher than commonly assumed
摘要译文
由模型翻译 · 术语统一 · 请对照原文核对关键结论「OCR 2.0」与大规模视觉语言模型（VLM）的出现，为文本识别树立了新的基准。然而，这些统一架构往往伴随着显著的计算开销、在复杂版式中精确定位文本的困难，以及产生文本幻觉的倾向。本文重新审视「模型规模是通往高精度的唯一路径」这一流行观念，提出 PP-OCRv5——一个经过细致优化、仅有 500 万参数的轻量级 OCR 系统。我们证明，PP-OCRv5 在标准 OCR 基准上取得了与许多十亿参数 VLM 相当的性能，同时具备更优的定位精度并减少了幻觉。我们成功的基石不在于架构扩张，而在于以数据为中心的探究。我们通过量化三个关键维度——数据难度、数据准确度与数据多样性——系统地剖析了训练数据的作用。大量实验表明，在具备足量高质量、标注准确且多样的数据时，传统高效两阶段 OCR 流水线的性能上限远高于通常的预期。这项工作为轻量级专用模型在大模型时代的可行性提供了有力证据，并为 OCR 的数据整理提供了实用洞见。源代码与模型已公开于 https://github.com/PaddlePaddle/PaddleOCR。
英文原文摘要 1479 字符
The advent of "OCR 2.0" and large-scale vision-language models (VLMs) has set new benchmarks in text recogni- tion. However, these unified architectures often come with significant computational demands, challenges in precise text localization within complex layouts, and a propen- sity for textual hallucinations. Revisiting the prevailing notion that model scale is the sole path to high accu- racy, this paper introduces PP-OCRv5, a meticulously op- timized, lightweight OCR system with merely 5 million pa- rameters. We demonstrate that PP-OCRv5 achieves per- formance competitive with many billion-parameter VLMs on standard OCR benchmarks, while offering superior lo- calization precision and reduced hallucinations. The cor- nerstone of our success lies not in architectural expansion but in a data-centric investigation. We systematically dis- sect the role of training data by quantifying three critical dimensions: data difficulty, data accuracy, and data diver- sity. Our extensive experiments reveal that with a sufficient volume of high-quality, accurately labeled, and diverse data, the performance ceiling for traditional, efficient two- stage OCR pipelines is far higher than commonly assumed. This work provides compelling evidence for the viability of lightweight, specialized models in the large-model era and offers practical insights into data curation for OCR. The source code and models are publicly available at https: //github.com/PaddlePaddle/PaddleOCR.