mlsanityData to 7 Oct 2026

Read Like Humans: Autonomous, Bidirectional and Iterative Language Modeling for Scene Text Recognition

arXiv 2103.06495 ↗First posted 11 Mar 2021cs.CVcited by 194

Not on the map. Why?

Abstract (arXiv)

Linguistic knowledge is of great benefit to scene text recognition. However, how to effectively model linguistic rules in end-to-end deep networks remains a research challenge. In this paper, we argue that the limited capacity of language models comes from: 1) implicitly language modeling; 2) unidirectional feature representation; and 3) language model with noise input. Correspondingly, we propose an autonomous, bidirectional and iterative ABINet for scene text recognition. Firstly, the autonomous suggests to block gradient flow between vision and language models to enforce explicitly language modeling. Secondly, a novel bidirectional cloze network (BCN) as the language model is proposed based on bidirectional feature representation. Thirdly, we propose an execution manner of iterative correction for language model which can effectively alleviate the impact of noise input. Additionally, based on the ensemble of iterative predictions, we propose a self-training method which can learn from unlabeled images effectively. Extensive experiments indicate that ABINet has superiority on low-quality images and achieves state-of-the-art results on several mainstream benchmarks. Besides, the ABINet trained with ensemble self-training shows promising improvement in realizing human-level recognition. Code is available at https://github.com/FangShancheng/ABINet.

Key facts. Cited by 194 papers in the map (data through 7 Oct 2026). Builds on 36 of its 56 references that are in the map. First posted on arXiv 11 Mar 2021.

Builds on

36 of 56 references.

References · 36

  1. 2017Attention Is All You Needcited by 59,932
  2. 2015Deep Residual Learning for Image Recognitioncited by 51,809
  3. 2018BERT: Pre-training of Deep Bidirectional Transformers for Language Understandingcited by 40,204
  4. 2015U-Net: Convolutional Networks for Biomedical Image Segmentationcited by 19,604
  5. 2016Layer Normalizationcited by 5,579
  6. 2016Pointer Sentinel Mixture Modelscited by 2,992
  7. 2018Deep contextualized word representationscited by 2,285
  8. 2019Self-training with Noisy Student improves ImageNet classificationcited by 1,304
  9. 2015An End-to-End Trainable Neural Network for Image-based Sequence Recognition and Its Application to Scene Text Recognitioncited by 519
  10. 2016Synthetic Data for Text Localisation in Natural Imagescited by 396
  11. 2014Synthetic Data and Artificial Neural Networks for Natural Scene Text Recognitioncited by 214
  12. 2019What Is Wrong With Scene Text Recognition Model Comparisons? Dataset and Model Analysiscited by 191
24 more
  1. 2019Mask TextSpotter: An End-to-End Trainable Neural Network for Spotting Text with Arbitrary Shapescited by 171
  2. 2020Towards Accurate Scene Text Recognition with Semantic Reasoning Networkscited by 131
  3. 2018Show, Attend and Read: A Simple and Strong Baseline for Irregular Text Recognitioncited by 130
  4. 2014Reading Text in the Wild with Convolutional Neural Networkscited by 122
  5. 2016Fully Convolutional Instance-aware Semantic Segmentationcited by 122
  6. 2016Robust Scene Text Recognition with Automatic Rectificationcited by 121
  7. 2020SEED: Semantics Enhanced Encoder-Decoder Framework for Scene Text Recognitioncited by 119
  8. 2018Scene Text Detection and Recognition: The Deep Learning Eracited by 114
  9. 2017Focusing Attention: Towards Accurate Text Recognition in Natural Imagescited by 110
  10. 2019Decoupled Attention Network for Text Recognitioncited by 102
  11. 2020RobustScanner: Dynamically Enhancing Positional Clues for Robust Text Recognitioncited by 85
  12. 2018ESIR: End-to-end Scene Text Recognition via Iterative Image Rectificationcited by 79
  13. 2016Recursive Recurrent Nets with Attention Modeling for OCR in the Wildcited by 74
  14. 2018NRTR: A No-Recurrence Sequence-to-Sequence Model For Scene Text Recognitioncited by 73
  15. 2020ContourNet: Taking a Further Step toward Accurate Arbitrary-shaped Scene Text Detectioncited by 69
  16. 2019TextScanner: Reading Characters in Order for Robust Scene Text Recognitioncited by 68
  17. 2018Scene Text Recognition from Two-Dimensional Perspectivecited by 57
  18. 2020GTC: Guided Training of CTC Towards Efficient and Accurate Scene Text Recognitioncited by 52
  19. 2015Reading Scene Text in Deep Convolutional Sequencescited by 44
  20. 2020On Vocabulary Reliance in Scene Text Recognitioncited by 38
  21. 20192D Attentional Irregular Scene Text Recognizercited by 18
  22. 2017Attention-based Extraction of Structured Information from Street View Imagerycited by 14
  23. 2014Deep Structured Output Learning for Unconstrained Text Recognitioncited by 13
  24. 2019A Holistic Representation Guided Attention Network for Scene Text Recognitioncited by 9

Built on by

Cited by year (2026 to 7 Oct)

Citing papers per year: 2021: 9; 2022: 37; 2023: 52; 2024: 50; 2025: 29; 2026: 17.

194 papers.

Sort:

By area:

  1. 2022GIT: A Generative Image-to-text Transformer for Vision and Languagecited by 536
  2. 2021TrOCR: Transformer-based Optical Character Recognition with Pre-trained Modelscited by 278
  3. 2023TextDiffuser: Diffusion Models as Text Painterscited by 223
  4. 2024OCRBench v2: An Improved Benchmark for Evaluating Large Multimodal Models on Visual Text Localization and Reasoningcited by 165
  5. 2022Scene Text Recognition with Permuted Autoregressive Sequence Modelscited by 148
  6. 2022SVTR: Scene Text Recognition with a Single Visual Modelcited by 86
  7. 2021LaTr: Layout-Aware Transformer for Scene-Text VQAcited by 72
  8. 2022SwinTextSpotter: Scene Text Spotting via Better Synergy between Text Detection and Text Recognitioncited by 70
  9. 2023SPTS v2: Single-Point Scene Text Spottingcited by 66
  10. 2023Conditional Text Image Generation with Diffusion Modelscited by 57
  11. 2023Multi-modal In-Context Learning Makes an Ego-evolving Scene Text Recognizercited by 53
  12. 2021Benchmarking Chinese Text Recognition: Datasets, Baselines, and an Empirical Studycited by 52
38 more
  1. 2023Exploring OCR Capabilities of GPT-4V(ision) : A Quantitative and In-depth Evaluationcited by 46
  2. 2022Reading and Writing: Discriminative and Generative Modeling for Self-Supervised Text Recognitioncited by 45
  3. 2022Toward Understanding WordArt: Corner-Guided Transformer for Scene Text Recognitioncited by 45
  4. 2023Revisiting Scene Text Recognition: A Data Perspectivecited by 45
  5. 2023Improving Diffusion Models for Scene Text Editing with Dual Encoderscited by 44
  6. 2023DiffUTE: Universal Text Editing Diffusion Modelcited by 44
  7. 2024TextCtrl: Diffusion-based Scene Text Editing with Prior Guidance Controlcited by 44
  8. 2022ABINet++: Autonomous, Bidirectional and Iterative Language Modeling for Scene Text Spottingcited by 40
  9. 2023DTrOCR: Decoder-only Transformer for Optical Character Recognitioncited by 40
  10. 2021CDistNet: Perceiving Multi-Domain Character Distance for Robust Text Recognitioncited by 36
  11. 2022Multi-Granularity Prediction for Scene Text Recognitioncited by 36
  12. 2022OCR-IDL: OCR Annotations for Industry Document Library Datasetcited by 35
  13. 2021Multi-modal Text Recognition Networks: Interactive Enhancements between Visual and Semantic Featurescited by 34
  14. 2022MaskOCR: Text Recognition with Masked Encoder-Decoder Pretrainingcited by 34
  15. 2025A Survey on Knowledge-Oriented Retrieval-Augmented Generationcited by 34
  16. 2023CLIP4STR: A Simple Baseline for Scene Text Recognition with Pre-trained Vision-Language Modelcited by 30
  17. 2023ESTextSpotter: Towards Better Scene Text Spotting with Explicit Synergy in Transformercited by 29
  18. 2022Self-supervised Character-to-Character Distillation for Text Recognitioncited by 27
  19. 2022C3-STISR: Scene Text Image Super-resolution with Triple Cluescited by 26
  20. 2021Visual Semantics Allow for Textual Reasoning Better in Scene Text Recognitioncited by 25
  21. 2022Self-supervised Implicit Glyph Attention for Text Recognitioncited by 25
  22. 2022Levenshtein OCRcited by 24
  23. 2023Chinese Text Recognition with A Pre-Trained CLIP-Like Model Through Image-IDS Aligningcited by 24
  24. 2024SVTRv2: CTC Beats Encoder-Decoder Models in Scene Text Recognitioncited by 23
  25. 2024Visual Text Generation in the Wildcited by 22
  26. 2022SimAN: Exploring Self-Supervised Representation Learning of Scene Text via Similarity-Aware Normalizationcited by 21
  27. 2023ICDAR 2023 Competition on Structured Text Extraction from Visually-Rich Document Imagescited by 20
  28. 2023Context Perception Parallel Decoder for Scene Text Recognitioncited by 19
  29. 2023Linguistic More: Taking a Further Step toward Efficient and Accurate Scene Text Recognitioncited by 18
  30. 2024Instruction-Guided Scene Text Recognitioncited by 17
  31. 2022Multimodal Semi-Supervised Learning for Text Recognitioncited by 16
  32. 2023CLIPTER: Looking at the Bigger Picture in Scene Text Recognitioncited by 16
  33. 2024First Creating Backgrounds Then Rendering Texts: A New Paradigm for Visual Text Blendingcited by 16
  34. 2023Improving Scene Text Image Super-resolution via Dual Prior Modulation Networkcited by 15
  35. 2023Scene Text Image Super-resolution based on Text-conditional Diffusion Modelscited by 15
  36. 2023Symmetrical Linguistic Feature Distillation with CLIP for Scene Text Recognitioncited by 14
  37. 2023LISTER: Neighbor Decoding for Length-Insensitive Scene Text Recognitioncited by 13
  38. 2023IPAD: Iterative, Parallel, and Diffusion-based Network for Scene Text Recognitioncited by 12

50 most cited of 194 All 194 (JSON)