无需训练，用图模型推断复杂文档的阅读顺序，准确率超现有方法。
Reading Order Inference for Complex Document Layouts
- 构建文本行间的有向图，用语言模型信号评分边。
- 在仿古文献布局中正确率95%，远超传统方法的50%。
- 对页面翻转不变，适合历史文献数字化场景。
阅读顺序推断仍是复杂历史手稿数字化的关键瓶颈，尤其在多路径交错排布的版式中，如‘普通注释’布局（Glossa Ordinaria），中心文本被非矩形、非凸区域环绕的注释包围。本文提出一种无训练、基于图的框架：每个OCR文本行作为节点，边由轻量级语言模型信号加权组合生成（包括因果语言模型条件似然与BERT的下一句预测，第三种句嵌入信号未提升效果），全局阅读顺序通过度约束有向路径覆盖求解。为避免贪心选择导致的“边盗用”失败，引入最大遗憾推理规则，优先确定机会成本高的边。在合成的Glossa Ordinaria网格布局、23个ALTO页面几何（10个历史页及镜像翻转变体）和140页多栏英文子集OmniDocBench上评估，相比经典递归XY-cut（PaddleOCR PP-StructureV3）与两种LayoutReader变体（仅布局、布局+文本），本方法在环绕布局平均恢复95%真实后继边，远超XY-cut的50%；在OmniDocBench多栏子集达88%宏边准确率，优于XY-cut的75%与LayoutReader的25%。布局模型因词级与行级粒度不匹配，迁移性能差。额外验证镜像不变性：本方法受水平或垂直反射影响小于1个百分点，传统XY-cut为2点，LayoutReader-T最高达8点。
原文摘要 · Abstract (English)
Reading order inference remains a critical bottleneck in the digitization of complex historical manuscripts, where pages contain multiple spatially interleaved reading streams, the canonical example being the Glossa Ordinaria layout, in which a central text is surrounded by commentaries that wrap around it in non-rectangular, non-convex regions. We present a training-free, graph-based framework: each OCR text line becomes a node in a directed candidate-transition graph, edges are scored by a weighted additive ensemble of two lightweight language-model signals (causal language model conditional likelihood and BERT next-sentence prediction, NSP; a third sentence-embedding signal was evaluated but did not improve reading order), and the global reading order is recovered as a degree-constrained directed path cover. To avoid the cascading "edge-theft" failures of greedy edge selection, we propose a max-regret inference rule that prioritizes commitments with high opportunity cost. We evaluate on synthetic Glossa Ordinaria grid layouts, on 23 ALTO page geometries (10 historical source pages plus mirrored and flipped variants), and on a 140-page multi-column English subset of OmniDocBench, comparing our method against the canonical recursive XY-cut (PaddleOCR PP-StructureV3) and two LayoutReader variants (layout-only and text+layout) on identical inputs. On wrap-around Glossa layouts our method recovers 95% of ground-truth successor edges on average vs. XY-cut's 50%; on the OmniDocBench multi-column subset it reaches 88% macro edge accuracy versus XY-cut's 75% and LayoutReader's 25%. The LayoutReader baselines transfer poorly due to a word-level vs. line-level granularity mismatch. We additionally verify mirror-invariance under horizontal and vertical page reflections: Our method changes by less than 1 percentage point, classical XY-cut by 2 points, and LayoutReader-T by up to 8 points.
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。