Pith. sign in

REVIEW 2 major objections 5 minor 55 references

Boosting Semi-Supervised Scene Text Recognition via Viewing and Summarizing

T0 review · 2 major / 5 minor · reviewed 2026-08-12 · deepseek-v4-flash

Challenge this review Re-run · record.json Download PDF Read on arXiv ↗ Review JSON

Pith's one-line read A viewing-and-summarizing semi-supervised framework reads artistic and distorted scene text without new human labels, by generating diverse character styles and correcting a contrastive-loss error.

desk verdict Strong empirical results for semi-supervised STR, but the paper's headline theoretical claim about the previous contrastive loss is not actually proven and should be reframed. read the letter →

arxiv 2411.15585 v1 pith:ZT6QZVVO submitted 2024-11-23 cs.CV

Yadong Qu , Yuxin Wang , Bangbang Zhou , Zixiao Wang , Hongtao Xie , Yongdong Zhang This is my paper · ORCID
classification cs.CV
keywords scenetextrecognitionsemi-supervisedlearningcharactermorphologyUnidirectionalAlignmentLossOnlineGenerationStrategymeanteacherUnion14M-Benchmarkcontrastive
verification ladder T0 review T1 audit T2 compute T3 formal

The pith

A machine-rendered reading of the paper's core claim, the machinery that carries it, and where it could break.

The reading

This paper tries to establish that the main barrier to recognizing artistic and distorted scene text is insufficient character-morphology diversity, and that a semi-supervised learner can overcome that barrier without human annotation. Its proposed method, ViSu, combines a viewing stage, in which an online generator produces background-free text samples in varied styles and real unlabeled images are added, with a summarizing stage, in which a new alignment loss clusters features of the same character. The paper also claims to prove that the earlier Character Contrastive loss contains a derivation error that spuriously repels same-class characters, and that its new loss fixes this error. If the claims are right, the result is 94.7% average accuracy on six common benchmarks and 70.9% on Union14M-Benchmark using only synthetic labels and unlabeled real data, with the largest gains on hard categories.

What carries the argument

The load-bearing machinery is a Mean-Teacher student-teacher loop, where the teacher is a slow exponential-moving-average copy of the student, with three purpose-built components. Unified Representation Forms (URF) rotate any tall image so text reads left-to-right, cutting eight possible orientation and reading-order variants down to two stable forms. The Online Generation Strategy (OGS) takes the labels of synthetic images and renders fresh background-free text in randomized fonts and orientations at every step, producing the base images that the teacher sees. The Character Unidirectional Alignment (CUA) Loss aligns a student's strongly augmented character features with the teacher's features from those base images, using a denominator that omits the repulsive same-class term the paper identifies in the earlier contrastive loss. Recognition and character-consistency losses provide the remaining supervision, and the EMA teacher supplies stable targets.

What would settle it

Take a minibatch of features from the trained teacher, compute the exact summed gradient of the previous Character Contrastive loss with respect to a same-class anchor over all of its positives, and measure its projection onto the mean same-class direction. If that projection is positive and the per-term repulsion disappears in aggregate, the paper's central derivation-error claim is falsified, even though CUA might still help empirically.

Watch

Extended reading notes

Core claim

The central discovery the paper argues for is that the hard part of scene text recognition is character morphology, and that a semi-supervised learner can attack it by viewing more diverse glyphs and then summarizing them into compact per-character features. The paper's theoretical contribution is a critique of the earlier Character Contrastive loss: in its gradient, the coefficient $N_2=-\exp(f_i\cdot f_{p'}/\tau)$ multiplies same-class feature vectors, so the authors say the loss mistakenly pushes characters of the same class apart and makes intra-class distributions sparse. Their replacement, the Character Unidirectional Alignment Loss, changes the denominator so that only one direction of alignment is enforced: student features are drawn toward teacher reference features for the same character, and same-class characters no longer act as negatives. With this correction plus the online generation of background-free style-varied samples, the method reports 94.7% average accuracy on common benchmarks and 70.9% on Union14M-Benchmark, and the paper attributes its largest gains on artistic, multi-oriented, and contextless text to these two mechanisms.

Load-bearing premise

The load-bearing premise is that the per-positive gradient term the paper flags, taken as the behavior of the whole loss, actually describes how the previous contrastive loss treats same-class characters; the paper never sums the gradient over all positive targets, where attraction terms could offset the repulsion.

Editorial extensions

If this is right

  • With only synthetic MJSynth and SynthText labels plus unlabeled real images, the method reaches 94.7% average accuracy on six common benchmarks and 70.9% on Union14M-Benchmark, beating fully supervised models on hard text benchmarks.
  • The ablation tables attribute the gain to the two mechanisms: OGS and CUA each add measurable accuracy on artistic, multi-oriented, and contextless categories, and the gains transfer when the framework is attached to other recognizers.
  • The correction to the contrastive loss implies same-class character features become more compact, which should reduce character ambiguity on distorted samples without any additional language model.
  • Because OGS generates background-free samples from existing synthetic annotations, the extra training data costs no human annotation and can be produced online during training.

Reading between the lines

Editorial extensions of the paper, not claims the author makes directly.

  • Beyond the paper, the per-anchor gradient argument suggests a direct test: compute the summed gradient of the old contrastive loss over all same-class pairs and check whether intra-class repulsion actually survives; the proposed CUA loss could be practically useful even if its stated theoretical justification does not hold in aggregate.
  • The same viewing-and-summarizing design, generating style-diverse background-free guidance for a teacher and aligning a student to it, is a generic recipe that could transfer to handwritten text, mathematical expressions, or low-resource scripts, where synthetic fonts are even less representative.
  • The paper's own limitation note locates the remaining failure mode in extreme aspect ratios: long lines are resized too aggressively and lose character information, pointing to segmentation of long text lines as the natural next step.
  • The choice to feed the teacher background-free images codifies the assumption that background is task-irrelevant for character recognition; extending this framework to cluttered or occluded text may require re-introducing context rather than removing it.
Share X Bluesky LinkedIn Reddit HN

Editorial analysis

A structured set of objections, weighed in public.

Desk editor's note, referee report, and a circularity audit.

Referee Report

2 major / 5 minor

Summary. The paper proposes ViSu, a semi-supervised scene text recognition framework built around two ideas. The 'viewing' component uses an Online Generation Strategy (OGS) to render background-free glyph samples from synthetic labels, a Unified Representation Forms (URF) step that reduces reading-order and orientation variants, and a Mean-Teacher setup that exploits real unlabeled data. The 'summarizing' component introduces a Character Unidirectional Alignment (CUA) loss that aligns strongly augmented student features with teacher features extracted from base images. The authors claim to theoretically demonstrate a flaw in the existing Character Contrastive (CC) loss and to correct it with CUA. Experiments report 94.7% average accuracy on six common benchmarks and 70.9% on Union14M-Benchmark, with gains across CRNN, TRBA, and a Transformer baseline.

Significance. The empirical contribution is substantial: the method is evaluated on multiple benchmarks and backbones, with component-wise ablations, and the reported gains on challenging Union-B and WordArt are large. The idea of generating background-free samples to make a semi-supervised STR model focus on character morphology is practically attractive. The theoretical motivation, however, is the weakest part: the claimed proof of the CC-loss defect is not carried out for the summed objective, and part of the reported Union-B gain is tuned on Union-B itself. With a corrected analysis and a clearer evaluation protocol, the paper could be a useful contribution; in its current form the central 'theoretically demonstrate' claim is not supported.

major comments (2)
  1. [Sec. 3.2 / Appendix A, Eq. (6)] The theoretical demonstration of the CC-loss flaw inspects only a single positive summand p. After summing Eq. (6) over all p in P'(i), the coefficient of a fixed positive feature f_q in the gradient-descent update direction is proportional to B(i) - |P'(i)| exp(f_i · f_q / tau), not simply -exp(f_i · f_q / tau). This coefficient is positive for positives that are not much more similar than the batch average, and it becomes negative only for sufficiently above-average positives. For example, when all positives have the same similarity and negatives have low similarity, the net update for every positive is still attractive. The manuscript never evaluates this summed coefficient, so the conclusion that CC loss 'mistakenly causes sparsity in the intra-class distribution' is not demonstrated. The CUA motivation therefore rests on an unproven gradient-sign assertion. A minimal fix is to analyze the summed gradient and to support the sparsity claim with either a toy experiment or empirical sign statistics of the per-sample coefficients.
  2. [B.4 / B.5, Tables 8 and 9] The final OGS configuration (random font, orientation, background, color) and the CUA hyperparameters (lambda, eta_cua, tau) are selected by Union-B A VG, and the same Union-B numbers are then reported as the headline result in Table 1 and the abstract. Because Union-B is one of the two main claims, the reported 70.9% figure is partially fitted to that benchmark rather than predicted. Please separate tuning from evaluation, for example by using a validation split of Union-B or by reporting the chosen configuration's sensitivity on a different benchmark. This does not invalidate the common-benchmark results but weakens the strength of the Union-B claim.
minor comments (5)
  1. [Table 3] The row labeled 'ViSu OGS 1.7 25.1 5.8 4.9 3.0 6.9 10.2 8.2' appears to contain deltas rather than accuracies, and the row is not clearly explained in the text; please relabel and describe these numbers explicitly.
  2. [Table 4] The first two rows use 'CE' in the consistency-loss column, whereas later rows use 'KL-div'; the distinction between recognition loss and consistency loss should be clarified in the table caption and in Section 4.4.2.
  3. [Sec. 3.2 / Appendix A] The statement that 'the detailed proof is included in Appendix A' overstates the content of the appendix: Appendix A derives gradient formulas but contains no statement or proof about intra-class sparsity.
  4. [Sec. 4.2] The symbol 'A VG' is used for both the weighted average on common benchmarks and the average on Union-B; please use distinct names to avoid ambiguity.
  5. [Fig. 4] The t-SNE comparison between CC and CUA losses is qualitative; please report a quantitative metric such as intra-class and inter-class feature distances to substantiate the claimed clustering effect.

Circularity Check

0 steps flagged · score 0.0 of 10

No significant circularity: benchmark claims are externally evaluated; the CUA-Loss motivation is an independent (if contestable) gradient argument, not a reduction to inputs.

full rationale

The paper's main claims are empirical and are checked against external benchmarks (six common benchmarks and Union14M-Benchmark, Tables 1-2) using public checkpoints and same-configuration reproductions, so the reported 94.7% and 70.9% accuracies are not defined in terms of the model's own outputs or derived from the claimed loss correction. The theoretical section 3.2 derives gradients of the defined CC and CUA losses (Eqs. 5-7 and Appendix A) and asserts that the per-summand N2<0 term treats positive samples as negatives; whether this sign argument survives summation over all positives is a mathematical correctness issue, not circularity, because the conclusion is not assumed in the premises. The missing evaluation of the summed-gradient coefficient in Eq. 6 is a real support gap, but it does not make the derivation circular. Self-citations (e.g., [10], [43], [52], [54]) are ordinary prior-work comparisons and do not carry the load of the central argument; no uniqueness theorem or ansatz is imported from the authors' earlier papers. Hyperparameters (lambda, eta_cua, tau) and OGS settings are selected on Union-B in Tables 8-9, which is a benchmark-tuning limitation rather than a circular derivation: the loss is not defined in terms of Union-B accuracy and the reported number is a measured result, not a fitted prediction dressed as a prediction. Section B.7 itself acknowledges a remaining limitation (extreme aspect ratios). Because no load-bearing step reduces to its own inputs, the circularity score is 0.

Assumptions & free parameters 7 free parameters · 5 assumptions · 0 invented entities

No new physical entities, mediators, or conserved quantities are introduced. The new components are a loss function and a data generation strategy, which are methods rather than postulated entities. The main loaded assumptions are the correctness of the CC Loss formula, the transferability of synthetic labels to rendered glyphs, and the value of teacher features from generated and unlabeled images.

free parameters (7)
  • lambda (CUA loss weight) = 0.1
    Tuned on Union-B accuracy in Table 9; lambda=0 gives 68.2% and lambda=1 gives 68.6%.
  • eta_cua confidence threshold = 0.7
    Tuned on Union-B in Table 9; 0.5 gives 70.5% and 0.9 gives 70.6%.
  • tau temperature in CUA and CC losses = 0.1
    Tuned on Union-B in Table 9; 0.05 and 0.2 give slightly lower accuracy.
  • eta_ccr consistency threshold = 0.5
    Set in implementation with no dedicated ablation reported.
  • URF aspect-ratio threshold r = 1.3
    Hand-chosen in Sec. B.1 to decide rotation of tall images.
  • EMA smoothing factor alpha = 0.999
    Standard Mean Teacher choice in Sec. B.1.
  • OGS configuration (random font, orientation, background, color) = character-level font, instance-level orientation, no background, no color
    Selected by Union-B comparison in Table 8; adding background or random color decreases accuracy.
assumptions (5)
  • standard math Gradient calculus and chain rule for the loss derivatives
    Used implicitly in Appendix A; no formal proof system is involved.
  • domain assumption The formula for the previous character contrastive loss in Eq. 3 accurately represents CornerTransformer [45]
    The critique of CC Loss depends on this exact formulation; if [45] used a different normalization or positive-set definition, the theoretical argument does not apply.
  • ad hoc to paper Teacher model features from OGS background-free samples and weakly augmented real images are reliable alignment targets
    CUA Loss assumes these reference features are noise-free and glyph-diverse; this is validated only indirectly by final accuracy.
  • ad hoc to paper Character labels from synthetic data transfer to randomly re-rendered glyphs in OGS
    The method renders the same word with random fonts and orientations, assuming the rendered character still belongs to the original label class.
  • domain assumption Union14M-U is a legitimate unlabeled training distribution for evaluating generalization on Union14M-Benchmark
    Unlabeled data and test benchmark are drawn from the same Union14M collection, so the test set is not domain-disjoint from the unlabeled training set.

how reviews work

0 comments
Cite this review

Pith. "Pith review of Boosting Semi-Supervised Scene Text Recognition via Viewing and Summarizing." pith.science (2026). https://pith.science/paper/ZT6QZVVO

@misc{pith2026241115585,
  author       = {Pith},
  title        = {Pith review of: Boosting Semi-Supervised Scene Text Recognition via Viewing and Summarizing},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/ZT6QZVVO}},
  note         = {Machine review of arXiv:2411.15585}
}
read the original abstract

Existing scene text recognition (STR) methods struggle to recognize challenging texts, especially for artistic and severely distorted characters. The limitation lies in the insufficient exploration of character morphologies, including the monotonousness of widely used synthetic training data and the sensitivity of the model to character morphologies. To address these issues, inspired by the human learning process of viewing and summarizing, we facilitate the contrastive learning-based STR framework in a self-motivated manner by leveraging synthetic and real unlabeled data without any human cost. In the viewing process, to compensate for the simplicity of synthetic data and enrich character morphology diversity, we propose an Online Generation Strategy to generate background-free samples with diverse character styles. By excluding background noise distractions, the model is encouraged to focus on character morphology and generalize the ability to recognize complex samples when trained with only simple synthetic data. To boost the summarizing process, we theoretically demonstrate the derivation error in the previous character contrastive loss, which mistakenly causes the sparsity in the intra-class distribution and exacerbates ambiguity on challenging samples. Therefore, a new Character Unidirectional Alignment Loss is proposed to correct this error and unify the representation of the same characters in all samples by aligning the character features in the student model with the reference features in the teacher model. Extensive experiment results show that our method achieves SOTA performance (94.7\% and 70.9\% average accuracy on common benchmarks and Union14M-Benchmark). Code will be available at https://github.com/qqqyd/ViSu.

Figures

Figures reproduced from arXiv: 2411.15585 by the authors.

Figure 1
Figure 1. (a) shows some images from synthetic datasets MJSynth and SynthText. (b) and (c) show [PITH_FULL_IMAGE:figures/full_fig_p002_1.png] view at source ↗
Figure 2
Figure 2. (a) All possible representations of English text images according to character orientation [PITH_FULL_IMAGE:figures/full_fig_p004_2.png] view at source ↗
Figure 3
Figure 3. Our framework consists of the student and teacher model. [PITH_FULL_IMAGE:figures/full_fig_p005_3.png] view at source ↗
Figures from the paper (2 more)
Figure 4
Figure 4. Figure 4: (a) shows several challenging examples. The four lines from top to bottom represent the [PITH_FULL_IMAGE:figures/full_fig_p019_4.png]
Figure 5
Figure 5. Figure 5: Failure cases. The first line is the ground-truth, and the second line is the recognition [PITH_FULL_IMAGE:figures/full_fig_p019_5.png]

Discussion (0). Continue with ORCID to comment.

Reference graph

Works this paper leans on

55 extracted references · 38 canonical work pages

  1. [1]

    What if we only use real datasets for scene text recognition? toward scene text recognition with fewer labels

    Jeonghun Baek, Yusuke Matsui, and Kiyoharu Aizawa. What if we only use real datasets for scene text recognition? toward scene text recognition with fewer labels. In Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition, pages 3113–3122, 2021

    work page 2021

  2. [2]

    Scene text recognition with permuted autoregressive sequence models

    Darwin Bautista and Rowel Atienza. Scene text recognition with permuted autoregressive sequence models. In European Conference on Computer Vision, pages 178–196. Springer, 2022

    work page 2022

  3. [3]

    Combining labeled and unlabeled data with co-training

    Avrim Blum and Tom Mitchell. Combining labeled and unlabeled data with co-training. In Proceedings of the eleventh annual conference on Computational learning theory, pages 92–100, 1998

    1998

  4. [4]

    Big self-supervised models are strong semi-supervised learners

    Ting Chen, Simon Kornblith, Kevin Swersky, Mohammad Norouzi, and Geoffrey E Hinton. Big self-supervised models are strong semi-supervised learners. Advances in neural information processing systems, 33:22243–22255, 2020

    work page 2020

  5. [5]

    Lister: Neighbor decoding for length-insensitive scene text recognition

    Changxu Cheng, Peng Wang, Cheng Da, Qi Zheng, and Cong Yao. Lister: Neighbor decoding for length-insensitive scene text recognition. In Proceedings of the IEEE/CVF International Conference on Computer Vision, pages 19541–19551, 2023

    work page 2023

  6. [6]

    Icdar2019 robust reading challenge on arbitrary-shaped text-rrc-art

    Chee Kheng Chng, Yuliang Liu, Yipeng Sun, Chun Chet Ng, Canjie Luo, Zihan Ni, ChuanMing Fang, Shuaitao Zhang, Junyu Han, Errui Ding, et al. Icdar2019 robust reading challenge on arbitrary-shaped text-rrc-art. In 2019 International Conference on Document Analysis and Recognition (ICDAR), pages 1571–1576. IEEE, 2019

    work page 2019

  7. [7]

    Semi-supervised learning with context-conditional generative adversarial networks

    Emily Denton, Sam Gross, and Rob Fergus. Semi-supervised learning with context-conditional generative adversarial networks. arXiv preprint arXiv:1611.06430, 2016

    arXiv 2016

  8. [8]

    Tri-net for semi-supervised deep learning

    W Dong-DongChen and ZH WeiGao. Tri-net for semi-supervised deep learning. InProceedings of twenty-seventh international joint conference on artificial intelligence, pages 2014–2020, 2018

    work page 2014

Show all 55 references
  1. [9]

    Svtr: Scene text recognition with a single visual model

    Yongkun Du, Zhineng Chen, Caiyan Jia, Xiaoting Yin, Tianlun Zheng, Chenxia Li, Yuning Du, and Yu-Gang Jiang. Svtr: Scene text recognition with a single visual model. arXiv preprint arXiv:2205.00159, 2022

    2022 arXiv

  2. [10]

    Read like humans: Autonomous, bidirectional and iterative language modeling for scene text recognition

    Shancheng Fang, Hongtao Xie, Yuxin Wang, Zhendong Mao, and Yongdong Zhang. Read like humans: Autonomous, bidirectional and iterative language modeling for scene text recognition. In Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition , pages 7098–...

    2021

  3. [11]

    Semi-supervised scene text recognition

    Yunze Gao, Yingying Chen, Jinqiao Wang, and Hanqing Lu. Semi-supervised scene text recognition. IEEE Transactions on Image Processing, 30:3005–3016, 2021

    2021

  4. [12]

    Synthetic data for text localisation in natural images

    Ankush Gupta, Andrea Vedaldi, and Andrew Zisserman. Synthetic data for text localisation in natural images. In Proceedings of the IEEE conference on computer vision and pattern recognition, pages 2315–2324, 2016

    2016

  5. [13]

    Judging a book by its cover

    Brian Kenji Iwana, Syed Tahseen Raza Rizvi, Sheraz Ahmed, Andreas Dengel, and Seiichi Uchida. Judging a book by its cover. arXiv preprint arXiv:1610.09204, 2016

    2016 arXiv

  6. [14]

    Synthetic data and artificial neural networks for natural scene text recognition

    Max Jaderberg, Karen Simonyan, Andrea Vedaldi, and Andrew Zisserman. Synthetic data and artificial neural networks for natural scene text recognition. arXiv preprint arXiv:1406.2227, 2014. 11

    2014 arXiv

  7. [15]

    Revisiting scene text recognition: A data perspective

    Qing Jiang, Jiapeng Wang, Dezhi Peng, Chongyu Liu, and Lianwen Jin. Revisiting scene text recognition: A data perspective. In Proceedings of the IEEE/CVF international conference on computer vision, pages 20543–20554, 2023

    2023

  8. [16]

    Icdar 2015 competition on robust reading

    Dimosthenis Karatzas, Lluis Gomez-Bigorda, Anguelos Nicolaou, Suman Ghosh, Andrew Bagdanov, Masakazu Iwamura, Jiri Matas, Lukas Neumann, Vijay Ramaseshan Chandrasekhar, Shijian Lu, et al. Icdar 2015 competition on robust reading. In 2015 13th international conference on docume...

    2015

  9. [17]

    Icdar 2013 robust reading competition

    Dimosthenis Karatzas, Faisal Shafait, Seiichi Uchida, Masakazu Iwamura, Lluis Gomez i Big- orda, Sergi Robles Mestre, Joan Mas, David Fernandez Mota, Jon Almazan Almazan, and Lluis Pere De Las Heras. Icdar 2013 robust reading competition. In 2013 12th international conference ...

    2013

  10. [18]

    The open images dataset v4: Unified image classification, object detection, and visual relationship detection at scale

    Alina Kuznetsova, Hassan Rom, Neil Alldrin, Jasper Uijlings, Ivan Krasin, Jordi Pont-Tuset, Shahab Kamali, Stefan Popov, Matteo Malloci, Alexander Kolesnikov, et al. The open images dataset v4: Unified image classification, object detection, and visual relationship detection a...

    1956

  11. [19]

    Temporal ensembling for semi-supervised learning.arXiv preprint arXiv:1610.02242, 2016

    Samuli Laine and Timo Aila. Temporal ensembling for semi-supervised learning.arXiv preprint arXiv:1610.02242, 2016

    2016 arXiv

  12. [20]

    Recognizing handwritten mathematical expressions via paired dual loss attention network and printed mathematical expressions

    Anh Duc Le. Recognizing handwritten mathematical expressions via paired dual loss attention network and printed mathematical expressions. In Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition Workshops, pages 566–567, 2020

    2020

  13. [21]

    Pseudo-label: The simple and efficient semi-supervised learning method for deep neural networks

    Dong-Hyun Lee et al. Pseudo-label: The simple and efficient semi-supervised learning method for deep neural networks. In Workshop on challenges in representation learning, ICML , volume 3, page 896. Atlanta, 2013

    2013

  14. [22]

    Char-net: A character-aware neural network for distorted scene text recognition

    Wei Liu, Chaofeng Chen, and Kwan-Yee Wong. Char-net: A character-aware neural network for distorted scene text recognition. In Proceedings of the AAAI conference on artificial intelligence, volume 32, 2018

    2018

  15. [23]

    Moran: A multi-object rectified attention network for scene text recognition

    Canjie Luo, Lianwen Jin, and Zenghui Sun. Moran: A multi-object rectified attention network for scene text recognition. Pattern Recognition, 90:109–118, 2019

    2019

  16. [24]

    Scene text recognition using higher order language priors

    Anand Mishra, Karteek Alahari, and CV Jawahar. Scene text recognition using higher order language priors. In BMVC-British machine vision conference. BMV A, 2012

    2012

  17. [25]

    Multi-modal text recognition networks: Interactive enhancements between visual and semantic features

    Byeonghu Na, Yoonsik Kim, and Sungrae Park. Multi-modal text recognition networks: Interactive enhancements between visual and semantic features. In European Conference on Computer Vision, pages 446–463. Springer, 2022

    2022

  18. [26]

    Semi-supervised learning with generative adversarial networks.arXiv preprint arXiv:1606.01583, 2016

    Augustus Odena. Semi-supervised learning with generative adversarial networks.arXiv preprint arXiv:1606.01583, 2016

    2016 arXiv

  19. [27]

    Seq-ups: Sequential uncertainty-aware pseudo- label selection for semi-supervised text recognition

    Gaurav Patel, Jan P Allebach, and Qiang Qiu. Seq-ups: Sequential uncertainty-aware pseudo- label selection for semi-supervised text recognition. In Proceedings of the IEEE/CVF Winter Conference on Applications of Computer Vision, pages 6180–6190, 2023

    2023

  20. [28]

    Recognizing text with perspective distortion in natural scenes

    Trung Quy Phan, Palaiahnakote Shivakumara, Shangxuan Tian, and Chew Lim Tan. Recognizing text with perspective distortion in natural scenes. In Proceedings of the IEEE International Conference on Computer Vision, pages 569–576, 2013

    2013

  21. [29]

    Semi- supervised learning with ladder networks

    Antti Rasmus, Mathias Berglund, Mikko Honkala, Harri Valpola, and Tapani Raiko. Semi- supervised learning with ladder networks. Advances in neural information processing systems, 28, 2015

    2015

  22. [30]

    A robust arbitrary text detection system for natural scene images

    Anhar Risnumawan, Palaiahankote Shivakumara, Chee Seng Chan, and Chew Lim Tan. A robust arbitrary text detection system for natural scene images. Expert Systems with Applications , 41(18):8027–8048, 2014. 12

    2014

  23. [31]

    Regularization with stochastic transfor- mations and perturbations for deep semi-supervised learning

    Mehdi Sajjadi, Mehran Javanmardi, and Tolga Tasdizen. Regularization with stochastic transfor- mations and perturbations for deep semi-supervised learning. Advances in neural information processing systems, 29, 2016

    2016

  24. [32]

    Conceptual captions: A cleaned, hypernymed, image alt-text dataset for automatic image captioning

    Piyush Sharma, Nan Ding, Sebastian Goodman, and Radu Soricut. Conceptual captions: A cleaned, hypernymed, image alt-text dataset for automatic image captioning. In Proceedings of the 56th Annual Meeting of the Association for Computational Linguistics (Volume 1: Long Papers), ...

    2018

  25. [33]

    An end-to-end trainable neural network for image- based sequence recognition and its application to scene text recognition

    Baoguang Shi, Xiang Bai, and Cong Yao. An end-to-end trainable neural network for image- based sequence recognition and its application to scene text recognition. IEEE transactions on pattern analysis and machine intelligence, 39(11):2298–2304, 2016

    2016

  26. [34]

    Aster: An attentional scene text recognizer with flexible rectification

    Baoguang Shi, Mingkun Yang, Xinggang Wang, Pengyuan Lyu, Cong Yao, and Xiang Bai. Aster: An attentional scene text recognizer with flexible rectification. IEEE transactions on pattern analysis and machine intelligence, 41(9):2035–2048, 2018

    2018

  27. [35]

    Super-convergence: Very fast training of neural networks using large learning rates

    Leslie N Smith and Nicholay Topin. Super-convergence: Very fast training of neural networks using large learning rates. In Artificial intelligence and machine learning for multi-domain operations applications, volume 11006, pages 369–386. SPIE, 2019

    2019

  28. [36]

    Unsupervised and semi-supervised learning with categorical genera- tive adversarial networks

    Jost Tobias Springenberg. Unsupervised and semi-supervised learning with categorical genera- tive adversarial networks. arXiv preprint arXiv:1511.06390, 2015

    2015 arXiv

  29. [37]

    Mean teachers are better role models: Weight-averaged con- sistency targets improve semi-supervised deep learning results

    Antti Tarvainen and Harri Valpola. Mean teachers are better role models: Weight-averaged con- sistency targets improve semi-supervised deep learning results. Advances in neural information processing systems, 30, 2017

    2017

  30. [38]

    Visualizing data using t-sne

    Laurens Van der Maaten and Geoffrey Hinton. Visualizing data using t-sne. Journal of machine learning research, 9(11), 2008

    2008

  31. [39]

    Coco-text: Dataset and benchmark for text detection and recognition in natural images

    Andreas Veit, Tomas Matera, Lukas Neumann, Jiri Matas, and Serge Belongie. Coco-text: Dataset and benchmark for text detection and recognition in natural images. arXiv preprint arXiv:1601.07140, 2016

    2016 arXiv

  32. [40]

    End-to-end scene text recognition

    Kai Wang, Boris Babenko, and Serge Belongie. End-to-end scene text recognition. In 2011 International conference on computer vision, pages 1457–1464. IEEE, 2011

    2011

  33. [41]

    Multi-granularity prediction for scene text recognition

    Peng Wang, Cheng Da, and Cong Yao. Multi-granularity prediction for scene text recognition. In European Conference on Computer Vision, pages 339–355. Springer, 2022

    2022

  34. [42]

    From two to one: A new scene text recognizer with visual language modeling network

    Yuxin Wang, Hongtao Xie, Shancheng Fang, Jing Wang, Shenggao Zhu, and Yongdong Zhang. From two to one: A new scene text recognizer with visual language modeling network. In Proceedings of the IEEE/CVF International Conference on Computer Vision , pages 14194– 14203, 2021

    2021

  35. [43]

    Symmetrical linguistic feature distillation with clip for scene text recognition

    Zixiao Wang, Hongtao Xie, Yuxin Wang, Jianjun Xu, Boqiang Zhang, and Yongdong Zhang. Symmetrical linguistic feature distillation with clip for scene text recognition. In Proceedings of the 31st ACM International Conference on Multimedia, pages 509–518, 2023

    2023

  36. [44]

    Handwritten mathematical expression recognition via paired adversarial learning

    Jin-Wen Wu, Fei Yin, Yan-Ming Zhang, Xu-Yao Zhang, and Cheng-Lin Liu. Handwritten mathematical expression recognition via paired adversarial learning. International Journal of Computer Vision, 128:2386–2401, 2020

    2020

  37. [45]

    Toward understanding wordart: Corner-guided transformer for scene text recognition

    Xudong Xie, Ling Fu, Zhifei Zhang, Zhaowen Wang, and Xiang Bai. Toward understanding wordart: Corner-guided transformer for scene text recognition. In European Conference on Computer Vision, pages 303–321. Springer, 2022

    2022

  38. [46]

    Symmetry-constrained rectification network for scene text recognition

    Mingkun Yang, Yushuo Guan, Minghui Liao, Xin He, Kaigui Bian, Song Bai, Cong Yao, and Xiang Bai. Symmetry-constrained rectification network for scene text recognition. In Proceedings of the IEEE/CVF international conference on computer vision, pages 9147–9156, 2019. 13

    2019

  39. [47]

    Sequential visual and semantic consistency for semi-supervised text recognition

    Mingkun Yang, Biao Yang, Minghui Liao, Yingying Zhu, and Xiang Bai. Sequential visual and semantic consistency for semi-supervised text recognition. Pattern Recognition Letters, 178:174–180, 2024

    2024

  40. [48]

    A survey on deep semi-supervised learning

    Xiangli Yang, Zixing Song, Irwin King, and Zenglin Xu. A survey on deep semi-supervised learning. IEEE Transactions on Knowledge and Data Engineering, 2022

    2022

  41. [49]

    Towards accurate scene text recognition with semantic reasoning networks

    Deli Yu, Xuan Li, Chengquan Zhang, Tao Liu, Junyu Han, Jingtuo Liu, and Errui Ding. Towards accurate scene text recognition with semantic reasoning networks. In Proceedings of the IEEE/CVF conference on computer vision and pattern recognition, pages 12113–12122, 2020

    2020

  42. [50]

    Robustscanner: Dynamically enhancing positional clues for robust text recognition

    Xiaoyu Yue, Zhanghui Kuang, Chenhao Lin, Hongbin Sun, and Wayne Zhang. Robustscanner: Dynamically enhancing positional clues for robust text recognition. In European Conference on Computer Vision, pages 135–151. Springer, 2020

    2020

  43. [51]

    Esir: End-to-end scene text recognition via iterative image rectification

    Fangneng Zhan and Shijian Lu. Esir: End-to-end scene text recognition via iterative image rectification. In Proceedings of the IEEE/CVF conference on computer vision and pattern recognition, pages 2059–2068, 2019

    2019

  44. [52]

    Linguistic more: Taking a further step toward efficient and accurate scene text recognition

    Boqiang Zhang, Hongtao Xie, Yuxin Wang, Jianjun Xu, and Yongdong Zhang. Linguistic more: Taking a further step toward efficient and accurate scene text recognition. arXiv preprint arXiv:2305.05140, 2023

    2023 arXiv

  45. [53]

    Uber- text: A large-scale dataset for optical character recognition from street-level imagery

    Ying Zhang, Lionel Gueguen, Ilya Zharkov, Peter Zhang, Keith Seifert, and Ben Kadlec. Uber- text: A large-scale dataset for optical character recognition from street-level imagery. In SUNw: Scene Understanding Workshop-CVPR, volume 2017, page 5, 2017

    2017

  46. [54]

    Pushing the performance limit of scene text recognizer without human annotation

    Caiyuan Zheng, Hui Li, Seon-Min Rhee, Seungju Han, Jae-Joon Han, and Peng Wang. Pushing the performance limit of scene text recognizer without human annotation. In Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition, pages 14116–14125, 2022

    2022

  47. [55]

    Tps++: Attention- enhanced thin-plate spline for scene text recognition

    Tianlun Zheng, Zhineng Chen, Jinfeng Bai, Hongtao Xie, and Yu-Gang Jiang. Tps++: Attention- enhanced thin-plate spline for scene text recognition. arXiv preprint arXiv:2305.05322, 2023. 14 A Grandient of Loss Functions For Character Unidirectional Alignment (CUA) Loss, we have...

    2023 arXiv

Pith tools

Reviewed August 12, 2026 · model on record in the stance chip above.