mlsanityData to 7 Oct 2026

Post-OCR Paragraph Recognition by Graph Convolutional Networks

arXiv 2101.12741 ↗First posted 29 Jan 2021cs.CV cs.LGcited by 16

Not on the map. Why?

Abstract (arXiv)

We propose a new approach for paragraph recognition in document images by spatial graph convolutional networks (GCN) applied on OCR text boxes. Two steps, namely line splitting and line clustering, are performed to extract paragraphs from the lines in OCR results. Each step uses a beta-skeleton graph constructed from bounding boxes, where the graph edges provide efficient support for graph convolution operations. With only pure layout input features, the GCN model size is 3~4 orders of magnitude smaller compared to R-CNN based models, while achieving comparable or better accuracies on PubLayNet and other datasets. Furthermore, the GCN models show good generalization from synthetic training data to real-world images, and good adaptivity for variable document styles.

Key facts. Cited by 16 papers in the map (data through 7 Oct 2026). Builds on 17 of its 49 references that are in the map. First posted on arXiv 29 Jan 2021.

Builds on

17 of 49 references.

References · 17

  1. 2017Attention Is All You Needcited by 59,932
  2. 2015Deep Residual Learning for Image Recognitioncited by 51,809
  3. 2015Faster R-CNN: Towards Real-Time Object Detection with Region Proposal Networkscited by 9,617
  4. 2017Graph Attention Networkscited by 8,854
  5. 2017Mask R-CNNcited by 7,259
  6. 2017Inductive Representation Learning on Large Graphscited by 6,284
  7. 2017Neural Message Passing for Quantum Chemistrycited by 3,460
  8. 2019A Comprehensive Survey on Graph Neural Networkscited by 2,906
  9. 2018Deeper Insights into Graph Convolutional Networks for Semi-Supervised Learningcited by 1,099
  10. 2018Link Prediction Based on Graph Neural Networkscited by 829
  11. 2015Convolutional Networks on Graphs for Learning Molecular Fingerprintscited by 751
  12. 2019DeepGCNs: Can GCNs Go as Deep as CNNs?cited by 527
5 more
  1. 2019PubLayNet: largest dataset ever for document layout analysiscited by 349
  2. 2017Learning to Extract Semantic Structure from Documents Using Multimodal Fully Convolutional Neural Networkcited by 92
  3. 2020Deep Relational Reasoning Graph Network for Arbitrary Shape Text Detectioncited by 77
  4. 2019Deep Visual Template-Free Form Parsingcited by 24
  5. 2017Sequence-to-Label Script Identification for Multilingual OCRcited by 12

Built on by

Cited by year

Citing papers per year: 2021: 1; 2022: 4; 2023: 6; 2024: 4; 2025: 1.

16 papers.

Sort:

  1. 2022Towards End-to-End Unified Scene Text Detection and Layout Analysiscited by 96
  2. 2022FormNet: Structural Encoding beyond Sequential Modeling in Form Document Information Extractioncited by 64
  3. 2023ICDAR 2023 Competition on Hierarchical Text Detection and Recognitioncited by 23
  4. 2021ROPE: Reading Order Equivariant Positional Encoding for Graph-based Document Information Extractioncited by 22
  5. 2022Doc-GCN: Heterogeneous Graph Convolutional Networks for Document Layout Analysiscited by 17
  6. 2024Detect-Order-Construct: A Tree Construction based Approach for Hierarchical Document Structure Analysiscited by 14
  7. 2024DLAFormer: An End-to-End Transformer For Document Layout Analysiscited by 11
  8. 2023Hierarchical Text Spotter for Joint Text Spotting and Layout Analysiscited by 8
  9. 2022Unified Line and Paragraph Detection by Graph Convolutional Networkscited by 7
  10. 2023Text Reading Order in Uncontrolled Conditions by Sparse Graph Segmentationcited by 5
  11. 2023CTE: A Dataset for Contextualized Table Extractioncited by 3
  12. 2023PARAGRAPH2GRAPH: A GNN-based framework for layout paragraph analysiscited by 2
4 more
  1. 2023GVdoc: Graph-based Visual Document Classificationcited by 1
  2. 2024Dynamic Relation Transformer for Contextual Text Block Detectioncited by 1
  3. 2024Block-level Text Spotting with LLMscited by 0
  4. 2025LIGHT: Multi-Modal Text Linking on Historical Mapscited by 0