新布局模型提升文档解析精度，最高达78% mAP
Advanced Layout Analysis Models for Docling
- 基于RT-DETR等架构训练多模型，适配15万份异构文档
- 最佳模型heron-101达78% mAP，单图推理仅28毫秒
- 开源完整代码与模型，适合文档转换开发者使用
本技术报告记录了将新型版面分析模型集成到Docling文档转换流程的进展。我们在包含15万份文档（公开及私有）的异构语料上，基于RT-DETR、RT-DETRv2和DFINE架构训练了多个前沿目标检测器。对原始检测结果进行了后处理，以更适用于文档转换任务。通过多种评估方法在不同文档基准上验证了版面分析效果，并测量了在CPU、Nvidia和Apple GPU环境下的运行性能。引入五种新文档版面模型，相比Docling旧基线实现20.6%至23.9%的mAP提升，且推理时间相当或更优。最佳模型heron-101在单块NVIDIA A100 GPU上达到78% mAP，单图推理时间为28毫秒。大量定量与定性实验确立了训练、评估与部署文档版面检测器的最佳实践，为文档转换社区提供可操作指导。所有训练好的检查点、代码与文档均在HuggingFace以宽松许可证发布。
原文摘要 · Abstract (English)
This technical report documents the development of novel Layout Analysis models integrated into the Docling document-conversion pipeline. We trained several state-of-the-art object detectors based on the RT-DETR, RT-DETRv2 and DFINE architectures on a heterogeneous corpus of 150,000 documents (both openly available and proprietary). Post-processing steps were applied to the raw detections to make them more applicable to the document conversion task. We evaluated the effectiveness of the layout analysis on various document benchmarks using different methodologies while also measuring the runtime performance across different environments (CPU, Nvidia and Apple GPUs). We introduce five new document layout models achieving 20.6% - 23.9% mAP improvement over Docling's previous baseline, with comparable or better runtime. Our best model, "heron-101", attains 78% mAP with 28 ms/image inference time on a single NVIDIA A100 GPU. Extensive quantitative and qualitative experiments establish best practices for training, evaluating, and deploying document-layout detectors, providing actionable guidance for the document conversion community. All trained checkpoints, code, and documentation are released under a permissive license on HuggingFace.
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。