Pith. sign in

REVIEW 4 major objections 5 minor 5 cited by

Visual Text Processing: A Comprehensive Review and Unified Evaluation

T0 review · 4 major / 5 minor · reviewed 2026-08-16 · deepseek-v4-flash

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

Pith's one-line read A single MLLM-based score now grades six visual text tasks.

desk verdict A genuinely useful survey and benchmark, but the human validation of VTPScore is statistically empty and the scale ambiguity in Table 7 needs fixing before the metric's claims hold. read the letter →

arxiv 2504.21682 v2 pith:VLX4NSKR submitted 2025-04-30 cs.CV

Yan Shu , Weichao Zeng , Fangmin Zhao , Zeyu Chen , Zhenhang Li , Xiaomeng Yang , Yu Zhou , Paolo Rota
show 4 more authors
Xiang Bai Lianwen Jin Xu-Cheng Yin Nicu Sebe
This is my paper · ORCID
classification cs.CV
keywords visualtextprocessingVTPBenchVTPScoremultimodallargelanguagemodelsevaluationmetricimagereconstructionmanipulationbenchmark
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 survey argues that visual text processing, from super-resolution and dewarping to removal, editing, and generation, can be reviewed and compared under one roof, and that current evaluations are too fragmented to tell which models actually work. To back that claim, the paper introduces VTPBench, a curated benchmark of 4,305 samples spanning six visual text tasks, and VTPScore, an evaluation metric that asks a multimodal large language model (GPT-4o) to rate each output on visual quality and visual text readability, each on a 0-5 scale, against a ground-truth reference. On more than 20 open-source models, VTPScore ranks current methods and, in the authors' reading, agrees with human ratings. The payoff if true is a standardized, human-aligned scale for comparing future visual text models, plus a clear statement that the state of the art still has substantial room to improve.

What carries the argument

The load-bearing object is VTPScore, a reference-based MLLM evaluation prompt protocol built on GPT-4o. For each of the six tasks, the paper writes a task-specific prompt with two rubrics, visual quality and visual text readability; the model must return JSON scores from 0 to 5 for each, and VTPScore is their sum. The second piece is VTPBench, a curated 4,305-sample benchmark drawn from existing datasets, which gives the metric a fixed test bed across super-resolution, dewarping, enhancement, removal, editing, and generation. What the machinery does is replace task-specific pixel metrics and ad hoc OCR checks with a single comparison that tracks human notions of quality and readability.

What would settle it

Give a sample of VTPBench (say 30-50 images per task) to at least 20 independent raters, compute per-sample mean human scores, and compare them with VTPScore via rank correlation and per-task scatter; if Spearman correlation falls below roughly 0.7 or the top-ranked model changes under bootstrap resampling, the claim that VTPScore is human-aligned and reliable fails. A simpler targeted probe is to hold ground truth constant and perturb images by small blur, color shift, or text corruption; if VTPScore does not order those perturbations the way raters do, the metric is not tracking readability or quality.

Watch

Extended reading notes

Core claim

The paper's central discovery is that a single multimodal-language-model-based metric, VTPScore, can serve as a fair and reliable evaluation standard across six visually distinct text tasks. VTPScore decomposes every result into two numbers: a visual quality score (clarity, geometric correctness, style consistency, or region-specific artifact control, depending on the task) and a visual text readability score (whether the text in the output matches the reference text), each from 0 to 5, with the total being their sum. The metric is reference-based: GPT-4o sees the predicted image and the ground-truth image under task-specific prompts and answers in structured JSON. On VTPBench, the benchmark assembled from existing datasets, the scores place LEMMA on top for super-resolution, UVDoc for dewarping, DocRes for enhancement, ViTEraser for removal, TextCtrl for editing, and AnyText for generation; the authors report that these rankings align with their human study and that even the best models leave large gaps on several tasks.

Load-bearing premise

VTPScore is only as reliable as GPT-4o's agreement with human judgment, and the paper's evidence for that agreement is a ten-rater study that reports only mean scores without correlation, variance, or inter-rater reliability.

Editorial extensions

If this is right

  • If VTPScore is accepted, future visual text models can be compared on one 0-10 scale across six tasks, ending the current situation where each task uses its own pixel metrics and OCR checks.
  • The reported rankings imply different bottlenecks per task: super-resolution and removal are comparatively mature, while enhancement, editing, and generation remain far from human-level output.
  • Because VTPScore is model-based and reference-based, it can be rerun on any new model without reimplementing per-task evaluation code, making benchmark updates cheap.
  • The finding that no single model dominates all six tasks strengthens the case for unified, generalist visual text models, as suggested in the paper's open-challenges section.

Reading between the lines

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

  • Beyond the paper: VTPScore's success depends on the specific choice of GPT-4o; a cheaper or open-weights MLLM might not reproduce the same ordering, so the metric should be re-validated per base model, and the prompt templates could be published as a regression suite.
  • Beyond the paper: the same prompt-rubric idea transfers to other fine-grained image-manipulation families, such as object removal, general image inpainting, and face editing, where reference-based MLLM scoring could replace FID and pixel metrics.
  • Beyond the paper: because the human study averaged only ten raters with no variance or agreement statistics, a public leaderboard with per-sample human scores and inter-rater reliability would be a direct, low-cost extension that makes the alignment claim testable.
Share X Bluesky LinkedIn Reddit HN

Signed reviews

No signed human review yet.

Editorial analysis

A structured set of objections, weighed in public.

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

Referee Report

4 major / 5 minor

Summary. The paper is a survey-and-benchmark submission on visual text processing. It organizes the field into text image reconstruction (super-resolution, dewarping, enhancement) and manipulation (removal, editing, generation), reviews methods through the lens of text features (structure, stroke, semantics, style) and learning paradigms, and introduces VTPBench, a 4,305-sample benchmark spanning six tasks, together with VTPScore, an MLLM-based metric that sums visual-quality and text-readability scores. The authors evaluate more than 20 models on VTPBench and report a small human study that they interpret as showing high consistency between VTPScore and human judgment. The survey portion is broad and generally accurate, but the evaluation contribution's central validity claim is not statistically established in the current manuscript.

Significance. If the evaluation claims were fully supported, VTPBench and VTPScore would fill a real need: current visual text processing evaluations are fragmented across task-specific datasets and metrics, and a unified protocol with a human-aligned metric would be valuable to the community. The paper also contributes a useful feature-based taxonomy and covers a wide and recent literature, which is itself a service. The prompt design for the six tasks is explicit, and the choice to evaluate more than 20 models with official weights is a concrete effort toward reproducibility. However, the headline contribution is the metric, and the metric's human alignment and benchmark construction are not yet validated with sufficient rigor; the central claim of 'fair and reliable evaluation' therefore remains unsubstantiated. The missing analyses are feasible within the manuscript's scope.

major comments (4)
  1. [Sec. 4.3, Table 7] The statement that VTPScore shows 'high consistency' with human evaluation is not supported by the reported data. Table 7 lists only mean HumanScore values; no correlation coefficient, per-sample comparison, standard deviation, or inter-rater agreement (ICC, Krippendorff's alpha, or similar) is given. With only three to five models per task, a pooled correlation across tasks would be confounded by task identity, and within-task correlations cannot be assessed from the table. The authors should report per-task and pooled within-task Spearman correlations between VTPScore and HumanScore, per-sample agreement, and inter-rater reliability; without these, the central claim of human-aligned evaluation remains unverified.
  2. [Sec. 4.2 and Sec. 4.3, Table 7] The scale of HumanScore is ambiguous and appears incompatible with VTPScore. VTPScore is defined as the sum of visual quality (0–5) and readability (0–5), giving a 0–10 range, while the human raters are described as rating both criteria on a 0–5 scale, which would give a 0–5 average. LEMMA's HumanScore of 8.20 cannot be a 0–5 average, while TSRN's HumanScore of 3.58 is implausibly low if it is a 0–10 sum given its VTPScore of 6.70. The paper must clarify whether HumanScore is a sum or an average, and rescale one set of scores so that VTPScore and HumanScore are compared in the same units.
  3. [Sec. 4.2, Data Construction] The VTPBench selection protocol is described only as 'carefully choose some representative data' and 'filter out extremely broken or severely damaged samples.' This is too vague to establish that the benchmark is unbiased or reproducible. The authors should specify the source datasets for each task, the exact filtering criteria (e.g., detection or OCR confidence thresholds, manual review rules), the number of samples removed per source, and the sampling procedure; ideally the filtered sample identifiers should be released.
  4. [Sec. 4.2 and Sec. 4.3] The human validation is in part circular: participants were instructed using the same visual-quality and readability criteria that are given to GPT-4o, so agreement between HumanScore and VTPScore could reflect shared task framing rather than the metric's accuracy. To validate VTPScore as a fair metric, the authors should also compare it against existing reference metrics (PSNR/SSIM, OCR accuracy, FID) and report per-sample agreement between GPT-4o and human ratings, not only model-level means.
minor comments (5)
  1. [Sec. 4.1] The sentence describing the Text Deblurring Dataset contains a duplicated typo: 'a cropped 300×300 patch.×300 pixel patch' should read 'a cropped 300×300 pixel patch.'
  2. [Sec. 3.6.2] The text contains typographical errors: 'trys to enhance controllability' should be 'tries to enhance controllability', and 'douple content and style learning' should be 'decouple content and style learning.'
  3. [Fig. 9] In the figure, the word 'Consturct' should be corrected to 'Construct.'
  4. [Table 4] The Pix2Pix row contains an extra numerical entry (10.2000) that is not aligned with the column headers; please realign the table.
  5. [Table 2] The column heading 'Easy Medium HardAverage↑' is missing spaces and the arrows appear inconsistently; please format the table for readability.

Circularity Check

0 steps flagged · score 0.0 of 10

No significant circularity: VTPScore is an MLLM-based metric defined directly from task prompts, and the human study is an independent, though statistically under-reported, validation.

full rationale

The paper's central contributions are a survey taxonomy, the VTPBench dataset collection, and the VTPScore evaluation metric. VTPScore is defined operationally as the sum of a GPT-4o visual quality score and a GPT-4o visual text readability score, each elicited by task-specific prompts (Section 4.2, Figure 9). There is no fitted parameter, no optimization against human scores, and no equation in which the claimed output is algebraically equivalent to an input. The metric is not derived from the benchmark; rather, it is applied to the benchmark, so no self-definitional reduction occurs. The human evaluation in Section 4.3 asks ten participants to rate the same criteria used in the VTPScore prompts. Although sharing a rubric can encourage agreement, the human scores are independently collected and averaged, and VTPScore is not fitted to them; this is a validation protocol, not a by-construction identity. The absence of correlation coefficients, variance, or inter-rater agreement, and the apparent scale inconsistency in HumanScore values (e.g., 8.20 for LEMMA if on a 0–5 average scale), are serious correctness and reporting concerns, but they are not circularity. The paper also evaluates the authors' own prior methods (TextCtrl [146], GlyphOnly [187]), but these are used only as benchmarked systems, not as evidence supporting the metric's validity or as load-bearing premises; no uniqueness theorem or ansatz is imported from self-citations. Therefore, no circular step can be exhibited with the required specificity.

Assumptions & free parameters 1 free parameters · 3 assumptions · 0 invented entities

The paper introduces no new physical or conceptual entities, forces, or dimensions. The only invented artifacts are the benchmark (VTPBench) and the metric (VTPScore), which are not 'entities' in the theory-loading sense. The free parameter is the equal weighting of the two sub-scores in VTPScore. The main axioms are domain assumptions about the reliability of reference-based MLLM evaluation and the adequacy of the human validation protocol.

free parameters (1)
  • VTPScore weighting = 1:1 sum of visual quality (0-5) and text readability (0-5)
    The overall VTPScore is defined as the sum of the two sub-scores with equal weights. This weighting is chosen by hand and is not fitted or justified by data; it directly affects the final score and ranking.
assumptions (3)
  • domain assumption Reference-based evaluation (providing ground truth to the evaluator) is more reliable than reference-free evaluation.
    Stated in Sec 4.2 Settings: 'Since reference-based evaluation is generally more reliable than reference-free evaluation, we simultaneously provide GPT-4o with both the predicted result and the corresponding ground truth label.' This is an unproved assumption that underpins VTPScore design.
  • domain assumption GPT-4o has strong enough low-level vision and OCR capabilities to accurately score visual quality and text readability.
    Invoked in Sec 4.2: GPT-4o is selected 'as it has demonstrated exceptional visual-language understanding and strong low-level vision perception capabilities.' The readability score relies on GPT-4o being able to recognize text in images without an external OCR module.
  • domain assumption Human evaluation with the same criteria is an appropriate gold standard for validating VTPScore.
    Used in Sec 4.3 Human Evaluation: ten participants rate the same benchmark using 'the same criteria as VTPScore'. The paper assumes that averaging these three-plus ratings per sample yields a reliable human preference signal.

how reviews work

0 comments
Cite this review

Pith. "Pith review of Visual Text Processing: A Comprehensive Review and Unified Evaluation." pith.science (2026). https://pith.science/paper/VLX4NSKR

@misc{pith2026250421682,
  author       = {Pith},
  title        = {Pith review of: Visual Text Processing: A Comprehensive Review and Unified Evaluation},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/VLX4NSKR}},
  note         = {Machine review of arXiv:2504.21682}
}
read the original abstract

Visual text is a crucial component in both document and scene images, conveying rich semantic information and attracting significant attention in the computer vision community. Beyond traditional tasks such as text detection and recognition, visual text processing has witnessed rapid advancements driven by the emergence of foundation models, including text image reconstruction and text image manipulation. Despite significant progress, challenges remain due to the unique properties that differentiate text from general objects. Effectively capturing and leveraging these distinct textual characteristics is essential for developing robust visual text processing models. In this survey, we present a comprehensive, multi-perspective analysis of recent advancements in visual text processing, focusing on two key questions: (1) What textual features are most suitable for different visual text processing tasks? (2) How can these distinctive text features be effectively incorporated into processing frameworks? Furthermore, we introduce VTPBench, a new benchmark that encompasses a broad range of visual text processing datasets. Leveraging the advanced visual quality assessment capabilities of multimodal large language models (MLLMs), we propose VTPScore, a novel evaluation metric designed to ensure fair and reliable evaluation. Our empirical study with more than 20 specific models reveals substantial room for improvement in the current techniques. Our aim is to establish this work as a fundamental resource that fosters future exploration and innovation in the dynamic field of visual text processing. The relevant repository is available at https://github.com/shuyansy/Visual-Text-Processing-survey.

Figures

Figures reproduced from arXiv: 2504.21682 by the authors.

Figure 1
Figure 1. Main structure of this survey. Initially, we introduce a hierarchical taxonomy, followed by related research areas. Subsequently, we conduct an [PITH_FULL_IMAGE:figures/full_fig_p003_1.png] view at source ↗
Figure 2
Figure 2. Text semantic and stroke features are utilized for [PITH_FULL_IMAGE:figures/full_fig_p005_2.png] view at source ↗
Figure 3
Figure 3. Text structure features are utilized for [PITH_FULL_IMAGE:figures/full_fig_p005_3.png] view at source ↗
Figures from the paper (6 more)
Figure 4
Figure 4. Figure 4: Text image enhancement methods leverage text stroke feature to perform classification or regression training paradigm. Two-stage Learning. Due to the challenges of annotating dewarping datasets and the lack of high-quality training data, many methods adopt a two-stage …
Figure 5
Figure 5. Figure 5: Text stroke features are utilized for scene text removal task with knowledge transfer, multi-task learning and progressive learning. the enhanced result from the preliminary coarse regression network and the target image. To ensure the preservation of text semantic fea…
Figure 6
Figure 6. Figure 6: Scene text editing methods mainly utilize the style feature to facilitate divide-and-conquer explicit transfer, implicit transfer, or inpainting￾based framework. Text stoke is also considered by template or text prompt representation. To improve readability, text seman…
Figure 7
Figure 7. Figure 7: Scene text generation methods mainly utilize the structure feature to realize geometry-aware text synthesis and generative layout control. To generate images resemble real-world data, the style feature is integrated into adversarial style learning. To render high-fidel…
Figure 8
Figure 8. Figure 8: Statistical overview of our VTPBench. Top: Task type of VTP￾Bench. Bottom: The number of samples and distribution of each task. (STE), OCR-related metrics are used to assess text rendering readability, while Frechet Inception Distance (FID) evaluates ´ the similarity b…
Figure 9
Figure 9. Figure 9: The overview of VTPBench evaluation and details of prompt design (visual quality and text readability) for visual text processing tasks. [PITH_FULL_IMAGE:figures/full_fig_p014_9.png]

Discussion (0). Continue with ORCID to comment.

Forward citations

Cited by 5 Pith papers

Reviewed papers in the Pith corpus that reference this work. Sorted by Pith novelty score. Full citation record

  1. Vid-SME: Membership Inference Attacks against Large Video Understanding Models

    cs.CV 2025-05 reject novelty 7.0 of 10

    Vid-SME computes Sharma-Mittal entropy differences between natural and reversed video frame sequences to infer training membership in video understanding LLMs, but its effectiveness is confounded by member/non-member ...

  2. SlerpFlow: Spherical Trajectory Correction for Rectified Flow Inversion

    cs.CV 2026-07 reject novelty 6.0 of 10

    SlerpFlow replaces Euclidean solver steps with spherical-linear-interpolation (slerp) direction correction for rectified-flow inversion, reporting improved FLUX reconstruction and editing on PIE-Bench.

  3. EpiAgent: An Agent-Centric System for Ancient Inscription Restoration

    cs.CV 2026-04 unverdicted novelty 6.0 of 10

    EpiAgent is a new agent-centric system that restores degraded ancient inscriptions with better quality and generalization than prior rigid AI methods by using an LLM planner to coordinate multimodal tools and iterativ...

  4. Uni-DocDiff: A Unified Document Restoration Model Based on Diffusion

    cs.CV 2025-08 conditional novelty 6.0 of 10

    A dual-stream diffusion model with a handcrafted prior pool and a prior fusion module unifies six document restoration tasks and matches task-specific specialists.

  5. Beyond Cropped Regions: New Benchmark and Corresponding Baseline for Chinese Scene Text Retrieval in Diverse Layouts

    cs.CV 2025-06 conditional novelty 6.0 of 10

    A full-image, mask-guided CLIP model with two-stage multi-granularity alignment training sets a new state of the art for Chinese scene text retrieval and introduces a diverse-layout benchmark.

Reference graph

Works this paper leans on

213 extracted references · 55 canonical work pages · cited by 5 Pith papers

  1. [1]

    UATVR: Uncertainty-adaptive text-video retrieval,

    B. Fang, W. Wu, C. Liu, Y. Zhou, Y. Song, W. Wang, X. Shu, X. Ji, and J. Wang, “UATVR: Uncertainty-adaptive text-video retrieval,” inICCV, 2023, pp. 13 723–13 733

    2023

  2. [2]

    Text detection from natural scene images: Towards a system for visually impaired persons,

    N. Ezaki, M. Bulacu, and L. Schomaker, “Text detection from natural scene images: Towards a system for visually impaired persons,” inICPR, vol. 2. IEEE, 2004, pp. 683–686

    2004

  3. [3]

    Resolv- ing sentiment discrepancy for multimodal sentiment detection via semantics completion and decomposition,

    D. Wu, D. Yang, H. Shen, C. Ma, and Y. Zhou, “Resolv- ing sentiment discrepancy for multimodal sentiment detection via semantics completion and decomposition,”arXiv preprint arXiv:2407.07026, 2024

    work page arXiv 2024

  4. [4]

    Document AI: Benchmarks, models and applications,

    L. Cui, Y. Xu, T. Lv, and F. Wei, “Document AI: Benchmarks, models and applications,”arXiv preprint arXiv:2111.08609, 2021

    arXiv 2021

  5. [5]

    Text detection and recognition in imagery: A survey,

    Q. Ye and D. Doermann, “Text detection and recognition in imagery: A survey,”TP AMI, vol. 37, no. 7, pp. 1480–1500, 2014

    2014

  6. [6]

    Scene text detection and recognition: Recent advances and future trends,

    Y. Zhu, C. Yao, and X. Bai, “Scene text detection and recognition: Recent advances and future trends,”Front. Comput. Sci., vol. 10, pp. 19–36, 2016

    2016

  7. [7]

    Text detection, tracking and recognition in video: A comprehensive survey,

    X.-C. Yin, Z.-Y. Zuo, S. Tian, and C.-L. Liu, “Text detection, tracking and recognition in video: A comprehensive survey,”TIP, vol. 25, no. 6, pp. 2752–2773, 2016

    2016

  8. [8]

    Scene text detection and recogni- tion with advances in deep learning: A survey,

    X. Liu, G. Meng, and C. Pan, “Scene text detection and recogni- tion with advances in deep learning: A survey,”IJDAR, vol. 22, pp. 143–162, 2019

    2019

Show all 213 references
  1. [9]

    Review of scene text detection and recognition,

    H. Lin, P . Yang, and F. Zhang, “Review of scene text detection and recognition,”Arch. Comput. Methods Eng., vol. 27, no. 2, pp. 433–454, 2020

    2020

  2. [10]

    Deep learning approaches to scene text detection: A comprehensive review,

    T. Khan, R. Sarkar, and A. F. Mollah, “Deep learning approaches to scene text detection: A comprehensive review,”ARTIF INTELL REV, vol. 54, pp. 3239–3298, 2021

    2021

  3. [11]

    Text recognition in the wild: A survey,

    X. Chen, L. Jin, Y. Zhu, C. Luo, and T. Wang, “Text recognition in the wild: A survey,”CSUR, vol. 54, no. 2, pp. 1–35, 2021

    2021

  4. [12]

    Scene text detection and recognition: The deep learning era,

    S. Long, X. He, and C. Yao, “Scene text detection and recognition: The deep learning era,”IJCV, vol. 129, no. 1, pp. 161–184, 2021

    2021

  5. [14]

    PreSTU: Pre-training for scene-text understand- ing,

    J. Kil, S. Changpinyo, X. Chen, H. Hu, S. Goodman, W.-L. Chao, and R. Soricut, “PreSTU: Pre-training for scene-text understand- ing,” inICCV, 2023, pp. 15 270–15 280

    2023

  6. [15]

    TPSNet: Reverse thinking of thin plate splines for arbitrary shape scene text representation,

    W. Wang, Y. Zhou, J. Lv, D. Wu, G. Zhao, N. Jiang, and W. Wang, “TPSNet: Reverse thinking of thin plate splines for arbitrary shape scene text representation,” inACM MM, 2022, pp. 5014– 5025

    2022

  7. [16]

    Partial scene text retrieval,

    H. Wang, M. Liao, Z. Xie, W. Liu, and X. Bai, “Partial scene text retrieval,”TP AMI, 2024

    2024

  8. [17]

    Selective concealment of characters for privacy protection,

    K. Inai, M. P ˚alsson, V . Frinken, Y. Feng, and S. Uchida, “Selective concealment of characters for privacy protection,” inCVPR, 2014, pp. 333–338

    2014

  9. [18]

    TranslatAR: A mobile augmented reality translator,

    V . Fragoso, S. Gauglitz, S. Zamora, J. Kleban, and M. Turk, “TranslatAR: A mobile augmented reality translator,” inWACV, 2011, pp. 497–502

    2011

  10. [19]

    Augmented reality meets computer vision: Efficient data generation for urban driving scenes,

    H. Abu Alhaija, S. K. Mustikovela, L. Mescheder, A. Geiger, and C. Rother, “Augmented reality meets computer vision: Efficient data generation for urban driving scenes,”IJCV, vol. 126, pp. 961–972, 2018

    2018

  11. [20]

    Denoising diffusion probabilistic models,

    J. Ho, A. Jain, and P . Abbeel, “Denoising diffusion probabilistic models,”NeurIPS, vol. 33, pp. 6840–6851, 2020

    2020

  12. [21]

    Diffusion models in vision: A survey,

    F.-A. Croitoru, V . Hondru, R. T. Ionescu, and M. Shah, “Diffusion models in vision: A survey,”TP AMI, 2023

    2023

  13. [22]

    Foundation models defining a new era in vision: A survey and outlook,

    M. Awais, M. Naseer, S. Khan, R. M. Anwer, H. Cholakkal, M. Shah, M.-H. Yang, and F. S. Khan, “Foundation models defining a new era in vision: A survey and outlook,”TP AMI, 2025

    2025

  14. [23]

    DocRes: A generalist model toward unifying document image restoration tasks,

    J. Zhang, D. Peng, C. Liu, P . Zhang, and L. Jin, “DocRes: A generalist model toward unifying document image restoration tasks,” inCVPR, 2024, pp. 15 654–15 664

    2024

  15. [24]

    TextDif- fuser: Diffusion models as text painters,

    J. Chen, Y. Huang, T. Lv, L. Cui, Q. Chen, and F. Wei, “TextDif- fuser: Diffusion models as text painters,” inNeurIPS, 2023, pp. 9353–9387

    2023

  16. [25]

    The surprisingly straightforward scene text removal method with gated attention and region of interest gen- eration: A comprehensive prominent model analysis,

    H. Lee and C. Choi, “The surprisingly straightforward scene text removal method with gated attention and region of interest gen- eration: A comprehensive prominent model analysis,” inECCV, 2022, pp. 457–472

    2022

  17. [26]

    A comprehensive study of multimodal large language models for image quality assessment,

    T. Wu, K. Ma, J. Liang, Y. Yang, and L. Zhang, “A comprehensive study of multimodal large language models for image quality assessment,” inECCV. Springer, 2024, pp. 143–160

    2024

  18. [27]

    Vi- sualcritic: Making LMMs perceive visual quality like humans,

    Z. Huang, Z. Zhang, Y. Lu, Z.-J. Zha, Z. Chen, and B. Guo, “Vi- sualcritic: Making LMMs perceive visual quality like humans,” arXiv preprint arXiv:2403.12806, 2024

    2024 arXiv

  19. [28]

    IC- DAR2015 competition on text image super-resolution,

    C. Peyrard, M. Baccouche, F. Mamalet, and C. Garcia, “IC- DAR2015 competition on text image super-resolution,”ICDAR, pp. 1201–1205, 2015

    2015

  20. [29]

    Selective super-resolution for scene text images,

    R. Nakao, B. K. Iwana, and S. Uchida, “Selective super-resolution for scene text images,”ICDAR, pp. 401–406, 2019

    2019

  21. [30]

    Photo-realistic single image super- resolution using a generative adversarial network,

    C. Ledig, L. Theis, F. Husz ´ar, J. Caballero, A. P . Aitken, A. Tejani, J. Totz, Z. Wang, and W. Shi, “Photo-realistic single image super- resolution using a generative adversarial network,”CVPR, pp. 105–114, 2016

    2016

  22. [31]

    Garden: Generative prior guided network for scene text image super-resolution,

    Y. Kong, W. Ma, L. Jin, and Y. Xue, “Garden: Generative prior guided network for scene text image super-resolution,” inIC- DAR. Springer, 2024, pp. 196–214

    2024

  23. [32]

    Scene text image super-resolution through multi-scale interaction of structural and semantic priors,

    Z. Zhu, L. Zhang, Y. Bai, Y. Wang, and P . Li, “Scene text image super-resolution through multi-scale interaction of structural and semantic priors,”TAI, 2024

    2024

  24. [33]

    Instruction-guided scene text recognition,

    Y. Du, Z. Chen, Y. Su, C. Jia, and Y.-G. Jiang, “Instruction-guided scene text recognition,”TP AMI, 2025

    2025

  25. [34]

    SEED: Semantics enhanced encoder-decoder framework for scene text recognition,

    Z. Qiao, Y. Zhou, D. Yang, Y. Zhou, and W. Wang, “SEED: Semantics enhanced encoder-decoder framework for scene text recognition,” inCVPR, 2020, pp. 13 528–13 537

    2020

  26. [35]

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

    B. Shi, X. Bai, and C. Yao, “An end-to-end trainable neural net- work for image-based sequence recognition and its application to scene text recognition,”TP AMI, vol. 39, no. 11, pp. 2298–2304, 2016

    2016

  27. [36]

    Turning a CLIP model into a scene text spotter,

    W. Yu, Y. Liu, X. Zhu, H. Cao, X. Sun, and X. Bai, “Turning a CLIP model into a scene text spotter,”TP AMI, 2024

    2024

  28. [37]

    Light-weight document image cleanup using perceptual loss,

    S. Dey and P . Jawanpuria, “Light-weight document image cleanup using perceptual loss,” inICDAR, 2021

    2021

  29. [38]

    The common fold: Utilizing the four-fold to dewarp printed documents from a single image,

    S. Das, G. Mishra, A. Sudharshana, and R. Shilkrot, “The common fold: Utilizing the four-fold to dewarp printed documents from a single image,”ACM SDE, 2017

    2017

  30. [39]

    A unified framework for document restoration using inpainting and shape- from-shading,

    L. Zhang, A. M. Yip, M. S. Brown, and C. L. Tan, “A unified framework for document restoration using inpainting and shape- from-shading,”PR, vol. 42, pp. 2961–2978, 2009

    2009

  31. [40]

    Metric rectification of curved document images,

    G. Meng, C. Pan, S. Xiang, J. Duan, and N. Zheng, “Metric rectification of curved document images,”TP AMI, vol. 34, no. 4, pp. 707–722, 2011

    2011

  32. [41]

    PIMNet: A parallel, iterative and mimicking network for scene text recognition,

    Z. Qiao, Y. Zhou, J. Wei, W. Wang, Y. Zhang, N. Jiang, H. Wang, and W. Wang, “PIMNet: A parallel, iterative and mimicking network for scene text recognition,” inACM MM, 2021, pp. 2046– 2055

    2021

  33. [42]

    PAN++: Towards efficient and accurate end-to-end spotting of arbitrarily-shaped text,

    W. Wang, E. Xie, X. Li, X. Liu, D. Liang, Z. Yang, T. Lu, and C. Shen, “PAN++: Towards efficient and accurate end-to-end spotting of arbitrarily-shaped text,”TP AMI, vol. 44, no. 9, pp. 5349–5367, 2021

    2021

  34. [43]

    SPTS v2: Single-point scene text spotting,

    Y. Liu, J. Zhang, D. Peng, M. Huang, X. Wang, J. Tang, C. Huang, D. Lin, C. Shen, X. Baiet al., “SPTS v2: Single-point scene text spotting,”TP AMI, vol. 45, no. 12, pp. 15 665–15 679, 2023

    2023

  35. [44]

    IPAD: Iterative, parallel, and diffusion-based network for scene text recognition,

    X. Yang, Z. Qiao, and Y. Zhou, “IPAD: Iterative, parallel, and diffusion-based network for scene text recognition,”IJCV, 2025

    2025

  36. [45]

    Linguistics-aware masked image modeling for self-supervised scene text recognition,

    Y. Zhang, C. Liu, J. Wei, X. Yang, Y. Zhou, C. Ma, and X. Ji, “Linguistics-aware masked image modeling for self-supervised scene text recognition,” inCVPR, 2025. 17

    2025

  37. [46]

    Divide rows and conquer cells: Towards structure recognition for large tables,

    H. Shen, X. Gao, J. Wei, L. Qiao, Y. Zhou, Q. Li, and Z. Cheng, “Divide rows and conquer cells: Towards structure recognition for large tables,” inIJCAI, 2023, pp. 1369–1377

    2023

  38. [47]

    BEDSR-Net: A deep shadow removal network from a single document image,

    Y.-H. Lin, W.-C. Chen, and Y.-Y. Chuang, “BEDSR-Net: A deep shadow removal network from a single document image,”CVPR, pp. 12 902–12 911, 2020

    2020

  39. [48]

    LP-IOANet: Efficient high resolution document shadow removal,

    K. Georgiadis, M. K. Yucel, E. Skartados, V . Dimaridou, A. Drosou, A. Sa `a-Garriga, and B. Manganelli, “LP-IOANet: Efficient high resolution document shadow removal,”ICASSP, pp. 1–5, 2023

    2023

  40. [49]

    Shadow removal of text document images using background estimation and adaptive text enhancement,

    W. Liu, B. Wang, J. Zheng, and W. Wang, “Shadow removal of text document images using background estimation and adaptive text enhancement,” inICASSP, 2023, pp. 1–5

    2023

  41. [50]

    High-fidelity document stain removal via a large-scale real-world dataset and a memory-augmented transformer,

    M. Li, H. Sun, Y. Lei, X. Zhang, Y. Dong, Y. Zhou, Z. Li, and X. Chen, “High-fidelity document stain removal via a large-scale real-world dataset and a memory-augmented transformer,”arXiv preprint arXiv:2410.22922, 2024

    2024 arXiv

  42. [51]

    DocDiff: Document enhancement via residual diffusion models,

    Z. Yang, B. Liu, Y. Xiong, L. Yi, G. Wu, X. Tang, Z. Liu, J. Zhou, and X. Zhang, “DocDiff: Document enhancement via residual diffusion models,”ACM MM, 2023

    2023

  43. [52]

    NAF-DPM: A nonlinear activation-free diffusion probabilistic model for document en- hancement,

    G. Cicchetti and D. Comminiello, “NAF-DPM: A nonlinear activation-free diffusion probabilistic model for document en- hancement,”arXiv preprint arXiv:2404.05669, 2024

    2024 arXiv

  44. [53]

    DocNLC: A document image enhancement framework with normalized and latent contrastive representation for multiple degradations,

    R. Wang, Y. Xue, and L. Jin, “DocNLC: A document image enhancement framework with normalized and latent contrastive representation for multiple degradations,” inAAAI, vol. 38, no. 6, 2024, pp. 5563–5571

    2024

  45. [54]

    DE-GAN: A conditional gen- erative adversarial network for document enhancement,

    M. A. Souibgui and Y. Kessentini, “DE-GAN: A conditional gen- erative adversarial network for document enhancement,”TP AMI, vol. 44, no. 3, pp. 1180–1191, 2020

    2020

  46. [55]

    DocTr: Document image transformer for geometric unwarping and illu- mination correction,

    H. Feng, Y. Wang, W. gang Zhou, J. Deng, and H. Li, “DocTr: Document image transformer for geometric unwarping and illu- mination correction,”ACM MM, 2021

    2021

  47. [56]

    Appearance en- hancement for camera-captured document images in the wild,

    J. Zhang, L. Liang, K. Ding, F. Guo, and L. Jin, “Appearance en- hancement for camera-captured document images in the wild,” TAI, vol. 5, no. 5, pp. 2319–2330, 2023

    2023

  48. [57]

    Doc- Stormer: Revitalizing multi-degraded colored document images to pristine PDF,

    C. Liu, J. Li, Y. Teng, C. Wang, N. Xu, J. Wu, and D. Tu, “Doc- Stormer: Revitalizing multi-degraded colored document images to pristine PDF,”arXiv preprint arXiv:2310.17910, 2023

    2023 arXiv

  49. [58]

    UDoc-GAN: Unpaired document illumination correction with background light prior,

    Y. Wang, W.-G. Zhou, Z. Lu, and H. Li, “UDoc-GAN: Unpaired document illumination correction with background light prior,” ACM MM, 2022

    2022

  50. [59]

    Predicting the original appearance of damaged historical documents,

    Z. Yang, D. Peng, Y. Shi, Y. Zhang, C. Liu, and L. Jin, “Predicting the original appearance of damaged historical documents,” in AAAI, vol. 39, no. 9, 2025, pp. 9382–9390

    2025

  51. [60]

    GDB: Gated convolutions- based document binarization,

    Z. Yang, B. Liu, Y. Xiong, and G. Wu, “GDB: Gated convolutions- based document binarization,”PR, vol. 146, p. 109989, 2024

    2024

  52. [61]

    A novel degraded document binarization model through vision transformer network,

    M. Yang and S. Xu, “A novel degraded document binarization model through vision transformer network,”Information Fusion, vol. 93, pp. 159–173, 2023

    2023

  53. [62]

    DeepOtsu: Document enhancement and binarization using iterative deep learning,

    S. He and L. Schomaker, “DeepOtsu: Document enhancement and binarization using iterative deep learning,”PR, vol. 91, pp. 379–390, 2019

    2019

  54. [63]

    DocBin- Former: A two-level transformer network for effective document image binarization,

    R. Biswas, S. K. Roy, N. Wang, U. Pal, and G.-B. Huang, “DocBin- Former: A two-level transformer network for effective document image binarization,”arXiv preprint arXiv:2312.03568, 2023

    2023 arXiv

  55. [64]

    DeepErase: Weakly supervised ink artifact removal in document text im- ages,

    W. R. Huang, Y. Qi, Q. Li, J. Degange, and Y. Llp, “DeepErase: Weakly supervised ink artifact removal in document text im- ages,”WACV, pp. 3511–3519, 2019

    2019

  56. [65]

    Mask-guided stamp erasure for real document image,

    X. Yang, D. Yang, Y. Zhou, Y. Guo, and W. Wang, “Mask-guided stamp erasure for real document image,”ICME, pp. 1631–1636, 2023

    2023

  57. [66]

    Scene text eraser,

    T. Nakamura, A. Zhu, K. Yanai, and S. Uchida, “Scene text eraser,” inICDAR, vol. 1, 2017, pp. 832–837

    2017

  58. [67]

    MTRNet: A generic scene text eraser,

    O. Tursun, R. Zeng, S. Denman, S. Sivapalan, S. Sridharan, and C. Fookes, “MTRNet: A generic scene text eraser,” inICDAR, 2019, pp. 39–44

    2019

  59. [68]

    MTRNet++: One-stage mask-based scene text eraser,

    O. Tursun, S. Denman, R. Zeng, S. Sivapalan, S. Sridharan, and C. Fookes, “MTRNet++: One-stage mask-based scene text eraser,”CVIU, vol. 201, p. 103066, 2020

    2020

  60. [69]

    Scene text magnifier,

    T. N. Nakamura, A. Zhu, and S. Uchida, “Scene text magnifier,” inICDAR, 2019, pp. 825–830

    2019

  61. [70]

    Selective style transfer for text,

    R. Gomez, A. F. Biten, L. Gomez, J. Gibert, D. Karatzas, and M. Rusi ˜nol, “Selective style transfer for text,” inICDAR, 2019, pp. 805–812

    2019

  62. [71]

    Editing text in the wild,

    L. Wu, C. Zhang, J. Liu, J. Han, J. Liu, E. Ding, and X. Bai, “Editing text in the wild,” inACM MM, 2019, pp. 1500–1508

    2019

  63. [72]

    SwapText: Image based texts transfer in scenes,

    Q. Yang, J. Huang, and W. Lin, “SwapText: Image based texts transfer in scenes,” inCVPR, 2020, pp. 14 700–14 709

    2020

  64. [73]

    Scene style text editing,

    T. Su, F. Yang, X. Zhou, D. Di, Z. Wang, and S. Li, “Scene style text editing,”arXiv preprint arXiv:2304.10097, 2023

    2023 arXiv

  65. [74]

    Improving diffusion models for scene text editing with dual encoders,

    J. Ji, G. Zhang, Z. Wang, B. Hou, Z. Zhang, B. Price, and S. Chang, “Improving diffusion models for scene text editing with dual encoders,”arXiv preprint arXiv:2304.05568, 2023

    2023 arXiv

  66. [75]

    AnyText2: Visual text genera- tion and editing with customizable attributes,

    Y. Tuo, Y. Geng, and L. Bo, “AnyText2: Visual text genera- tion and editing with customizable attributes,”arXiv preprint arXiv:2411.15245, 2024

    2024 arXiv

  67. [76]

    Scene text synthesis for efficient and effective deep network training,

    C. Zhang, F. Zhan, H. Zhu, and S. Lu, “Scene text synthesis for efficient and effective deep network training,”arXiv preprint arXiv:1901.09193, 2019

    1901 arXiv

  68. [77]

    eDiff-I: Text-to-image diffusion models with an ensemble of expert denoisers,

    Y. Balaji, S. Nah, X. Huang, A. Vahdat, J. Song, K. Kreis, M. Ait- tala, T. Aila, S. Laine, B. Catanzaroet al., “eDiff-I: Text-to-image diffusion models with an ensemble of expert denoisers,”arXiv preprint arXiv:2211.01324, 2022

    2022 arXiv

  69. [78]

    Dif- fusionPen: Towards controlling the style of handwritten text generation,

    K. Nikolaidou, G. Retsinas, G. Sfikas, and M. Liwicki, “Dif- fusionPen: Towards controlling the style of handwritten text generation,” inECCV. Springer, 2024, pp. 417–434

    2024

  70. [79]

    Font- Diffuser: One-shot font generation via denoising diffusion with multi-scale content aggregation and style contrastive learning,

    Z. Yang, D. Peng, Y. Kong, Y. Zhang, C. Yao, and L. Jin, “Font- Diffuser: One-shot font generation via denoising diffusion with multi-scale content aggregation and style contrastive learning,” inAAAI, vol. 38, no. 7, 2024, pp. 6603–6611

    2024

  71. [80]

    QT-Font: High-efficiency font synthesis via quadtree-based diffusion models,

    Y. Liu and Z. Lian, “QT-Font: High-efficiency font synthesis via quadtree-based diffusion models,” inSIGGRAPH, 2024, pp. 1–11

    2024

  72. [81]

    Geometric-aware control in diffusion model for handwritten chinese font generation,

    G. Yao, K. Zhao, C. Deng, N. Ding, T. Zhao, Y. Tao, and L. Peng, “Geometric-aware control in diffusion model for handwritten chinese font generation,” inICDAR. Springer, 2024, pp. 3–17

    2024

  73. [82]

    Hi-SAM: Marrying segment anything model for hierarchical text segmentation,

    M. Ye, J. Zhang, J. Liu, C. Liu, B. Yin, C. Liu, B. Du, and D. Tao, “Hi-SAM: Marrying segment anything model for hierarchical text segmentation,”TP AMI, 2024

    2024

  74. [83]

    Char-SAM: Turning segment anything model into scene text segmentation annotator with character-level visual prompts,

    E. Xie, J. Lyu, D. Wu, H. Shen, and Y. Zhou, “Char-SAM: Turning segment anything model into scene text segmentation annotator with character-level visual prompts,” inICASSP, 2025

    2025

  75. [84]

    Robust and accurate text stroke segmentation,

    S. Qin, P . Ren, S. Kim, and R. Manduchi, “Robust and accurate text stroke segmentation,” inWACV, 2018, pp. 242–250

    2018

  76. [85]

    Weak supervision for generating pixel–level annotations in scene text segmentation,

    S. Bonechi, M. Bianchini, F. Scarselli, and P . Andreini, “Weak supervision for generating pixel–level annotations in scene text segmentation,”PRL, vol. 138, pp. 1–7, 2020

    2020

  77. [86]

    Semi-supervised pixel-level scene text segmentation by mutu- ally guided network,

    C. Wang, S. Zhao, L. Zhu, K. Luo, Y. Guo, J. Wang, and S. Liu, “Semi-supervised pixel-level scene text segmentation by mutu- ally guided network,”TIP, vol. 30, pp. 8212–8221, 2021

    2021

  78. [87]

    Re- thinking text segmentation: A novel dataset and a text-specific refinement approach,

    X. Xu, Z. Zhang, Z. Wang, B. Price, Z. Wang, and H. Shi, “Re- thinking text segmentation: A novel dataset and a text-specific refinement approach,” inCVPR, 2021, pp. 12 045–12 055

    2021

  79. [88]

    Looking from a higher-level perspective: Attention and recognition enhanced multi-scale scene text segmentation,

    Y. Ren, J. Zhang, B. Chen, X. Zhang, and L. Jin, “Looking from a higher-level perspective: Attention and recognition enhanced multi-scale scene text segmentation,” inACCV, 2022, pp. 3138– 3154

    2022

  80. [89]

    Detecting tampered scene text in the wild,

    Y. Wang, H. Xie, M. Xing, J. Wang, S. Zhu, and Y. Zhang, “Detecting tampered scene text in the wild,” inECCV, 2022, pp. 215–232

    2022

  81. [90]

    Revisiting tampered scene text detection in the era of generative AI,

    C. Qu, Y. Zhong, F. Guo, and L. Jin, “Revisiting tampered scene text detection in the era of generative AI,” inAAAI, vol. 39, no. 1, 2025, pp. 694–702

    2025

  82. [91]

    Towards robust tampered text detection in document image: New dataset and new solution,

    C. Qu, C. Liu, Y. Liu, X. Chen, D. Peng, F. Guo, and L. Jin, “Towards robust tampered text detection in document image: New dataset and new solution,” inCVPR, 2023, pp. 5937–5946

    2023

  83. [92]

    Text prior guided scene text image super-resolution,

    J. Ma, S. Guo, and L. Zhang, “Text prior guided scene text image super-resolution,”TIP, vol. 32, pp. 1341–1353, 2021

    2021

  84. [93]

    A text attention network for spatial deformation robust scene text image super-resolution,

    J. Ma, Z. Liang, and L. Zhang, “A text attention network for spatial deformation robust scene text image super-resolution,” CVPR, pp. 5901–5910, 2022

    2022

  85. [94]

    C3- STISR: Scene text image super-resolution with triple clues,

    M. Zhao, M. Wang, F. Bai, B. Li, J. Wang, and S. Zhou, “C3- STISR: Scene text image super-resolution with triple clues,” in IJCAI, 2022

    2022

  86. [95]

    Towards robust scene text image super-resolution via explicit location enhancement,

    H. Guo, T. Dai, G. Meng, and S.-T. Xia, “Towards robust scene text image super-resolution via explicit location enhancement,” inIJCAI, 2023, pp. 782–790

    2023

  87. [96]

    Scene text image super-resolution based on text-conditional diffusion models,

    C. Noguchi, S. Fukuda, and M. Yamanaka, “Scene text image super-resolution based on text-conditional diffusion models,” in WACV, 2024, pp. 1485–1495

    2024

  88. [97]

    PEAN: A diffusion-based prior-enhanced attention network for scene text image super- resolution,

    Z. Zhao, H. Xue, P . Fang, and S. Zhu, “PEAN: A diffusion-based prior-enhanced attention network for scene text image super- resolution,” inACM MM, 2024, pp. 9769–9778. 18

    2024

  89. [98]

    Diffusion-based blind text image super-resolution,

    Y. Zhang, J. Zhang, H. Li, Z. Wang, L. Hou, D. Zou, and L. Bian, “Diffusion-based blind text image super-resolution,” inCVPR, 2024, pp. 25 827–25 836

    2024

  90. [99]

    DCDM: Diffusion-conditioned-diffusion model for scene text image super-resolution,

    S. Singh, P . Keserwani, M. Iwamura, and P . P . Roy, “DCDM: Diffusion-conditioned-diffusion model for scene text image super-resolution,” inECCV. Springer, 2025, pp. 303–320

    2025

  91. [100]

    TextSR: Content-aware text super-resolution guided by recogni- tion,

    W. Wang, E. Xie, P . Sun, W. Wang, L. Tian, C. Shen, and P . Luo, “TextSR: Content-aware text super-resolution guided by recogni- tion,”ArXiv, vol. abs/1909.07113, 2019

    1909 arXiv

  92. [101]

    PlugNet: Degradation aware scene text recognition supervised by a pluggable super-resolution unit,

    Y. Mou, L. Tan, H. Yang, J. Chen, L. Liu, R. Yan, and Y. Huang, “PlugNet: Degradation aware scene text recognition supervised by a pluggable super-resolution unit,” inECCV, 2020

    2020

  93. [102]

    Scene text telescope: Text-focused scene image super-resolution,

    J. Chen, B. Li, and X. Xue, “Scene text telescope: Text-focused scene image super-resolution,”CVPR, pp. 12 021–12 030, 2021

    2021

  94. [103]

    QT-TextSR: Enhancing scene text image super-resolution via efficient interaction with text recognition using a query-aware transformer,

    C. Liu, Q. Jiang, D. Peng, Y. Kong, J. Zhang, L. Xiong, J. Duan, C. Sun, and L. Jin, “QT-TextSR: Enhancing scene text image super-resolution via efficient interaction with text recognition using a query-aware transformer,”Neurocomputing, vol. 620, p. 129241, 2025

    2025

  95. [104]

    Text-attentional conditional gen- erative adversarial network for super-resolution of text images,

    Y. Wang, F. Su, and Y. Qian, “Text-attentional conditional gen- erative adversarial network for super-resolution of text images,” ICME, pp. 1024–1029, 2019

    2019

  96. [105]

    Improving scene text image super-resolution via dual prior modulation network,

    S. Zhu, Z. Zhao, P . Fang, and H. Xue, “Improving scene text image super-resolution via dual prior modulation network,” in AAAI, 2023

    2023

  97. [106]

    Learning generative structure prior for blind text image super-resolution,

    X. Li, W. Zuo, and C. C. Loy, “Learning generative structure prior for blind text image super-resolution,”CVPR, pp. 10 103–10 113, 2023

    2023

  98. [107]

    Analyzing and improving the image quality of StyleGAN,

    T. Karras, S. Laine, M. Aittala, J. Hellsten, J. Lehtinen, and T. Aila, “Analyzing and improving the image quality of StyleGAN,” CVPR, pp. 8107–8116, 2019

    2019

  99. [108]

    Text Gestalt: Stroke- aware scene text image super-resolution,

    J. Chen, H. Yu, J. Ma, B. Li, and X. Xue, “Text Gestalt: Stroke- aware scene text image super-resolution,” inAAAI, vol. 36, no. 1, 2022, pp. 285–293

    2022

  100. [109]

    Scene text image super-resolution in the wild,

    W. Wang, E. Xie, X. Liu, W. Wang, D. Liang, C. Shen, and X. Bai, “Scene text image super-resolution in the wild,” inECCV. Springer, 2020, pp. 650–666

    2020

  101. [110]

    Scene text image super-resolution via parallelly contextual attention network,

    C. Zhao, S. Feng, B. N. Zhao, Z. Ding, J. Wu, F. Shen, and H. T. Shen, “Scene text image super-resolution via parallelly contextual attention network,”ACM MM, 2021

    2021

  102. [112]

    Marior: Margin removal and iterative content rectification for document dewarp- ing in the wild,

    J. Zhang, C. Luo, L. Jin, F. Guo, and K. Ding, “Marior: Margin removal and iterative content rectification for document dewarp- ing in the wild,”ACM MM, 2022

    2022

  103. [113]

    Dewarping docu- ment image by displacement flow estimation with fully convolu- tional network,

    G.-W. Xie, F. Yin, X.-Y. Zhang, and C.-L. Liu, “Dewarping docu- ment image by displacement flow estimation with fully convolu- tional network,” inDAS, 2020

    2020

  104. [114]

    Geometric representation learning for document image rectification,

    H. Feng, W. gang Zhou, J. Deng, Y. Wang, and H. Li, “Geometric representation learning for document image rectification,” in ECCV, 2022

    2022

  105. [115]

    Deep unrestricted document image rectification,

    H. Feng, S. Liu, J. Deng, W. Zhou, and H. Li, “Deep unrestricted document image rectification,”TMM, vol. 26, pp. 6142–6154, 2023

    2023

  106. [116]

    Rethinking supervision in docu- ment unwarping: A self-consistent flow-free approach,

    S. Liu, H. Feng, and W. Zhou, “Rethinking supervision in docu- ment unwarping: A self-consistent flow-free approach,”TCSVT, 2023

    2023

  107. [117]

    Efficient joint rec- tification of photometric and geometric distortions in document images,

    H. Tang, J. Guo, T. Wang, Y. Yu, and C. Wang, “Efficient joint rec- tification of photometric and geometric distortions in document images,” inICASSP. IEEE, 2024, pp. 3690–3694

    2024

  108. [118]

    Revisiting document image dewarping by grid regularization,

    X. Jiang, R. Long, N. Xue, Z. Yang, C. Yao, and G. Xia, “Revisiting document image dewarping by grid regularization,”CVPR, pp. 4533–4542, 2022

    2022

  109. [119]

    Foreground and text-lines aware document image rectification,

    H. Li, X. Wu, Q. Chen, and Q. Xiang, “Foreground and text-lines aware document image rectification,” inICCV, 2023, pp. 19 574– 19 583

    2023

  110. [120]

    DewarpNet: Single-image document unwarping with stacked 3D and 2D regression networks,

    S. Das, K. Ma, Z. Shu, D. Samaras, and R. Shilkrot, “DewarpNet: Single-image document unwarping with stacked 3D and 2D regression networks,”ICCV, pp. 131–140, 2019

    2019

  111. [121]

    Document image rectifica- tion in complex scene using stacked siamese networks,

    Z. Xu, F. Yin, P . Yang, and C.-L. Liu, “Document image rectifica- tion in complex scene using stacked siamese networks,” inICPR. IEEE, 2022, pp. 1550–1556

    2022

  112. [122]

    UVDoc: Neural grid-based document unwarping,

    F. Verhoeven, T. Magne, and O. Sorkine-Hornung, “UVDoc: Neural grid-based document unwarping,” inSIGGRAPH Asia, 2023, pp. 1–11

    2023

  113. [123]

    Layout-aware single-image document flattening,

    P . Li, W. Quan, J. Guo, and D. Yan, “Layout-aware single-image document flattening,”ACM TOG, 2023

    2023

  114. [124]

    Automatic semantic content removal by learning to neglect,

    S. Qin, J. Wei, and R. Manduchi, “Automatic semantic content removal by learning to neglect,” inBMVC, 2018

    2018

  115. [125]

    Stroke-based scene text erasing using synthetic data for training,

    Z. Tang, T. Miyazaki, Y. Sugaya, and S. Omachi, “Stroke-based scene text erasing using synthetic data for training,”TIP, vol. 30, pp. 9306–9320, 2021

    2021

  116. [126]

    EnsNet: Ensconce text in the wild,

    S. Zhang, Y. Liu, L. Jin, Y. Huang, and S. Lai, “EnsNet: Ensconce text in the wild,” inAAAI, vol. 33, no. 01, 2019, pp. 801–808

    2019

  117. [127]

    Text region conditional generative adversarial network for text concealment in the wild,

    P . Keserwani and P . P . Roy, “Text region conditional generative adversarial network for text concealment in the wild,”TCSVT, vol. 32, no. 5, pp. 3152–3163, 2021

    2021

  118. [128]

    Don’t forget me: Accurate background recovery for text removal via modeling local-global context,

    C. Liu, L. Jin, Y. Liu, C. Luo, B. Chen, F. Guo, and K. Ding, “Don’t forget me: Accurate background recovery for text removal via modeling local-global context,” inECCV, 2022, pp. 409–426

    2022

  119. [129]

    EraseNet: End-to-end text removal in the wild,

    C. Liu, Y. Liu, L. Jin, S. Zhang, C. Luo, and Y. Wang, “EraseNet: End-to-end text removal in the wild,”TIP, vol. 29, pp. 8760–8775, 2020

    2020

  120. [130]

    PERT: A pro- gressively region-based network for scene text removal,

    Y. Wang, H. Xie, S. Fang, Y. Qu, and Y. Zhang, “PERT: A pro- gressively region-based network for scene text removal,”arXiv preprint arXiv:2106.13029, 2021

    2021 arXiv

  121. [131]

    PSSTRNet: Progressive segmentation- guided scene text removal network,

    G. Lyu and A. Zhu, “PSSTRNet: Progressive segmentation- guided scene text removal network,” inICME, 2022, pp. 1–6

    2022

  122. [132]

    Scene text removal via cascaded text stroke detection and erasing,

    X. Bian, C. Wang, W. Quan, J. Ye, X. Zhang, and D.-M. Yan, “Scene text removal via cascaded text stroke detection and erasing,” CVM, vol. 8, pp. 273–287, 2022

    2022

  123. [133]

    Progressive scene text erasing with self-supervision,

    X. Du, Z. Zhou, Y. Zheng, X. Wu, T. Ma, and C. Jin, “Progressive scene text erasing with self-supervision,”CVIU, vol. 233, p. 103712, 2023

    2023

  124. [134]

    Deep learning-based forgery attack on document images,

    L. Zhao, C. Chen, and J. Huang, “Deep learning-based forgery attack on document images,”TIP, vol. 30, pp. 7964–7979, 2021

    2021

  125. [135]

    Learning transferable visual models from natural language supervision,

    A. Radford, J. W. Kim, C. Hallacy, A. Ramesh, G. Goh, S. Agar- wal, G. Sastry, A. Askell, P . Mishkin, J. Clarket al., “Learning transferable visual models from natural language supervision,” inICML, 2021, pp. 8748–8763

    2021

  126. [136]

    Self-supervised cross-language scene text editing,

    F. Yang, T. Su, X. Zhou, D. Di, Z. Wang, and S. Li, “Self-supervised cross-language scene text editing,” inACM MM, 2023, pp. 4546– 4554

    2023

  127. [137]

    FAST: Font-agnostic scene text editing,

    A. Das, P . Roy, S. Bhattacharya, S. Ghosh, U. Pal, and M. Blu- menstein, “FAST: Font-agnostic scene text editing,”arXiv preprint arXiv:2308.02905, 2023

    2023 arXiv

  128. [138]

    Scene text transfer for cross-language,

    L. Zhang, X. Chen, Y. Xie, and Y. Lu, “Scene text transfer for cross-language,” inICIG, 2021, pp. 552–564

    2021

  129. [139]

    STEFANN: Scene text editor using font adaptive neural network,

    P . Roy, S. Bhattacharya, S. Ghosh, and U. Pal, “STEFANN: Scene text editor using font adaptive neural network,” inCVPR, 2020, pp. 13 228–13 237

    2020

  130. [140]

    DCfont: an end-to-end deep chinese font generation system,

    Y. Jiang, Z. Lian, Y. Tang, and J. Xiao, “DCfont: an end-to-end deep chinese font generation system,” inSIGGRAPH Asia, 2017, pp. 1–4

    2017

  131. [141]

    Separating style and content for generalized style transfer,

    Y. Zhang, Y. Zhang, and W. Cai, “Separating style and content for generalized style transfer,” inCVPR, 2018, pp. 8447–8455

    2018

  132. [142]

    Explor- ing stroke-level modifications for scene text editing,

    Y. Qu, Q. Tan, H. Xie, J. Xu, Y. Wang, and Y. Zhang, “Explor- ing stroke-level modifications for scene text editing,” inAAAI, vol. 37, no. 2, 2023, pp. 2119–2127

    2023

  133. [143]

    RewriteNet: Reliable scene text editing with implicit decompo- sition of text contents and styles,

    J. Lee, Y. Kim, S. Kim, M. Yim, S. Shin, G. Lee, and S. Park, “RewriteNet: Reliable scene text editing with implicit decompo- sition of text contents and styles,”CVPRW, 2021

    2021

  134. [144]

    TextStyleBrush: Transfer of text aesthetics from a single exam- ple,

    P . Krishnan, R. Kovvuri, G. Pang, B. Vassilev, and T. Hassner, “TextStyleBrush: Transfer of text aesthetics from a single exam- ple,”TP AMI, 2023

    2023

  135. [145]

    Letter embedding guidance diffusion model for scene text editing,

    C. Wang, L. Wu, X. Chen, X. Li, L. Meng, and X. Meng, “Letter embedding guidance diffusion model for scene text editing,” in ICME, 2023, pp. 588–593

    2023

  136. [146]

    Textctrl: Diffusion- based scene text editing with prior guidance control,

    W. Zeng, Y. Shu, Z. Li, D. Yang, and Y. Zhou, “Textctrl: Diffusion- based scene text editing with prior guidance control,”NeurIPS, 2024

    2024

  137. [147]

    On manipulating scene text in the wild with diffusion models,

    J. Santoso, C. Simonet al., “On manipulating scene text in the wild with diffusion models,” inWACV, 2024, pp. 5202–5211

    2024

  138. [148]

    Diffute: Universal text editing diffusion model,

    H. Chen, Z. Xu, Z. Gu, Y. Li, C. Meng, H. Zhu, W. Wang et al., “Diffute: Universal text editing diffusion model,”NeurIPS, vol. 36, pp. 63 062–63 074, 2023

    2023

  139. [149]

    Udifftext: A unified framework for high- quality text synthesis in arbitrary images via character-aware diffusion models,

    Y. Zhao and Z. Lian, “Udifftext: A unified framework for high- quality text synthesis in arbitrary images via character-aware diffusion models,” inECCV. Springer, 2024, pp. 217–233

    2024

  140. [150]

    Anytext: Multi- lingual visual text generation and editing,

    Y. Tuo, W. Xiang, J.-Y. He, Y. Geng, and X. Xie, “Anytext: Multi- lingual visual text generation and editing,” inICLR, 2023. 19

    2023

  141. [151]

    How control information influences multilingual text image generation and editing?

    B. Zhang, Z. Gao, Y. Qu, and H. Xie, “How control information influences multilingual text image generation and editing?” in NeurIPS, 2024

    2024

  142. [153]

    TextMaster: Universal controllable text edit,

    A. Wang, J. Wang, Z. Yan, W. Shang, R. Lin, and Z. Zhang, “TextMaster: Universal controllable text edit,”arXiv preprint arXiv:2410.09879, 2024

    2024

  143. [154]

    Ip-Adapter: Text compatible image prompt adapter for text-to-image diffusion models,

    H. Ye, J. Zhang, S. Liu, X. Han, and W. Yang, “Ip-Adapter: Text compatible image prompt adapter for text-to-image diffusion models,”arXiv preprint arXiv:2308.06721, 2023

    2023 arXiv

  144. [155]

    GlyphDraw: Learning to draw Chinese characters in image syn- thesis models coherently,

    J. Ma, M. Zhao, C. Chen, R. Wang, D. Niu, H. Lu, and X. Lin, “GlyphDraw: Learning to draw Chinese characters in image syn- thesis models coherently,”arXiv preprint arXiv:2303.17870, 2023

    2023 arXiv

  145. [156]

    Character-aware models improve visual text rendering,

    R. Liu, D. Garrette, C. Saharia, W. Chan, A. Roberts, S. Narang, I. Blok, R. Mical, M. Norouzi, and N. Constant, “Character-aware models improve visual text rendering,” inACL, 2023

    2023

  146. [157]

    Choose what you need: Disentangled representation learning for scene text recognition removal and editing,

    B. Zhang, H. Xie, Z. Gao, and Y. Wang, “Choose what you need: Disentangled representation learning for scene text recognition removal and editing,” inCVPR, 2024, pp. 28 358–28 368

    2024

  147. [158]

    Syn- thetic data and artificial neural networks for natural scene text recognition,

    M. Jaderberg, K. Simonyan, A. Vedaldi, and A. Zisserman, “Syn- thetic data and artificial neural networks for natural scene text recognition,” inNIPSW, 2014

    2014

  148. [159]

    SynthTIGER: Synthetic text image generator towards better text recognition models,

    M. Yim, Y. Kim, H.-C. Cho, and S. Park, “SynthTIGER: Synthetic text image generator towards better text recognition models,” in ICDAR, 2021, pp. 109–124

    2021

  149. [160]

    Synthetic data for text localisation in natural images,

    A. Gupta, A. Vedaldi, and A. Zisserman, “Synthetic data for text localisation in natural images,” inCVPR, 2016, pp. 2315–2324

    2016

  150. [161]

    Synth- Text3D: Synthesizing scene text images from 3D virtual worlds,

    M. Liao, B. Song, S. Long, M. He, C. Yao, and X. Bai, “Synth- Text3D: Synthesizing scene text images from 3D virtual worlds,” Sci. China Inf. Sci., vol. 63, pp. 1–14, 2020

    2020

  151. [162]

    UnrealText: Synthesizing realistic scene text images from the unreal world,

    S. Long and C. Yao, “UnrealText: Synthesizing realistic scene text images from the unreal world,”arXiv preprint arXiv:2003.10608, 2020

    2003 arXiv

  152. [163]

    Diffusion models beat GANs on image synthesis,

    P . Dhariwal and A. Nichol, “Diffusion models beat GANs on image synthesis,”NeurIPS, vol. 34, pp. 8780–8794, 2021

    2021

  153. [164]

    A diffusion model translator for efficient image-to-image translation,

    M. Xia, Y. Zhou, R. Yi, Y.-J. Liu, and W. Wang, “A diffusion model translator for efficient image-to-image translation,”TP AMI, 2024

    2024

  154. [165]

    Diffusion models in low-level vision: A survey,

    C. He, Y. Shen, C. Fang, F. Xiao, L. Tang, Y. Zhang, W. Zuo, Z. Guo, and X. Li, “Diffusion models in low-level vision: A survey,”TP AMI, 2025

    2025

  155. [166]

    Multimodal image synthesis and editing: The generative AI era,

    F. Zhan, Y. Yu, R. Wu, J. Zhang, S. Lu, L. Liu, A. Kortylewski, C. Theobalt, and E. Xing, “Multimodal image synthesis and editing: The generative AI era,”TP AMI, vol. 45, no. 12, pp. 15 098–15 119, 2023

    2023

  156. [167]

    Verisimilar image synthesis for accurate detection and recognition of texts in scenes,

    F. Zhan, S. Lu, and C. Xue, “Verisimilar image synthesis for accurate detection and recognition of texts in scenes,” inECCV, 2018, pp. 249–266

    2018

  157. [168]

    Glyph-ByT5: A customized text encoder for accurate visual text rendering,

    Z. Liu, W. Liang, Z. Liang, C. Luo, J. Li, G. Huang, and Y. Yuan, “Glyph-ByT5: A customized text encoder for accurate visual text rendering,” inECCV. Springer, 2024, pp. 361–377

    2024

  158. [169]

    TextDiffuser-2: Unleashing the power of language models for text rendering,

    J. Chen, Y. Huang, T. Lv, L. Cui, Q. Chen, and F. Wei, “TextDiffuser-2: Unleashing the power of language models for text rendering,” inECCV. Springer, 2024, pp. 386–402

    2024

  159. [170]

    Textflux: An ocr-free dit model for high-fidelity multilingual scene text synthesis,

    Y. Xie, J. Zhang, P . Chen, Z. Wang, W. Wang, L. Gao, P . Li, H. Sun, Q. Zhang, Q. Qiaoet al., “Textflux: An ocr-free dit model for high-fidelity multilingual scene text synthesis,”arXiv preprint arXiv:2505.17778, 2025

    2025

  160. [171]

    Textcrafter: Accurately rendering multiple texts in complex visual scenes,

    N. Du, Z. Chen, Z. Chen, S. Gao, X. Chen, Z. Jiang, J. Yang, and Y. Tai, “Textcrafter: Accurately rendering multiple texts in complex visual scenes,”arXiv preprint arXiv:2503.23461, 2025

    2025

  161. [172]

    Bizgen: Advancing article-level visual text render- ing for infographics generation,

    Y. Peng, S. Xiao, K. Wu, Q. Liao, B. Chen, K. Lin, D. Huang, J. Li, and Y. Yuan, “Bizgen: Advancing article-level visual text render- ing for infographics generation,”arXiv preprint arXiv:2503.20672, 2025

    2025 arXiv

  162. [173]

    Spatial fusion GAN for image synthesis,

    F. Zhan, H. Zhu, and S. Lu, “Spatial fusion GAN for image synthesis,” inCVPR, 2019, pp. 3653–3662

    2019

  163. [174]

    Learning to draw text in natural images with conditional adversarial networks

    S. Fang, H. Xie, J. Chen, J. Tan, and Y. Zhang, “Learning to draw text in natural images with conditional adversarial networks.” in IJCAI, 2019, pp. 715–722

    2019

  164. [175]

    ScrabbleGAN: Semi-supervised varying length handwritten text generation,

    S. Fogel, H. Averbuch-Elor, S. Cohen, S. Mazor, and R. Litman, “ScrabbleGAN: Semi-supervised varying length handwritten text generation,” inCVPR, 2020, pp. 4324–4333

    2020

  165. [176]

    Fonts: Text rendering with typography and style controls,

    W. Shi, Y. Song, D. Zhang, J. Liu, and X. Zou, “Fonts: Text rendering with typography and style controls,”arXiv preprint arXiv:2412.00136, 2024

    2024 arXiv

  166. [177]

    A benchmark for Chinese-English scene text image super-resolution,

    J. Ma, Z. Liang, W. Xiang, X. Yang, and L. Zhang, “A benchmark for Chinese-English scene text image super-resolution,” inICCV, 2023, pp. 19 452–19 461

    2023

  167. [178]

    DocUNet: Document image unwarping via a stacked U-Net,

    K. Ma, Z. Shu, X. Bai, J. Wang, and D. Samaras, “DocUNet: Document image unwarping via a stacked U-Net,”CVPR, pp. 4700–4709, 2018

    2018

  168. [179]

    DocReal: Robust document dewarping of real-life images via attention-enhanced control point prediction,

    F. Yu, Y. Xie, L. Wu, Y. Wen, G. Wang, S. Ren, X. Chen, J. Mao, and W. Li, “DocReal: Robust document dewarping of real-life images via attention-enhanced control point prediction,” inWACV, 2024, pp. 665–674

    2024

  169. [180]

    Document enhancement using visibility detection,

    N. Kligler, S. Katz, and A. Tal, “Document enhancement using visibility detection,”CVPR, pp. 2374–2382, 2018

    2018

  170. [181]

    ICFHR 2018 competition on handwritten document image binarization (H- DIBCO 2018),

    I. Pratikakis, K. Zagori, P . Kaddas, and B. Gatos, “ICFHR 2018 competition on handwritten document image binarization (H- DIBCO 2018),” inICFHR. IEEE, 2018, pp. 489–493

    2018

  171. [182]

    Local water-filling algorithm for shadow detection and removal of document images,

    B. Wang and C. P . Chen, “Local water-filling algorithm for shadow detection and removal of document images,”Sensors, vol. 20, no. 23, p. 6929, 2020

    2020

  172. [183]

    Document image shadow removal guided by color-aware back- ground,

    L. Zhang, Y. He, Q. Zhang, Z. Liu, X. Zhang, and C. Xiao, “Document image shadow removal guided by color-aware back- ground,” inCVPR, 2023, pp. 1818–1827

    2023

  173. [184]

    Self- supervised text erasing with controllable image synthesis,

    G. Jiang, S. Wang, T. Ge, Y. Jiang, Y. Wei, and D. Lian, “Self- supervised text erasing with controllable image synthesis,” in ACM MM, 2022, pp. 1973–1983

    2022

  174. [185]

    FETNet: Feature erasing and transferring network for scene text removal,

    G. Lyu, K. Liu, A. Zhu, S. Uchida, and B. K. Iwana, “FETNet: Feature erasing and transferring network for scene text removal,” PR, vol. 140, p. 109531, 2023

    2023

  175. [186]

    High-resolution image synthesis with latent diffusion models,

    R. Rombach, A. Blattmann, D. Lorenz, P . Esser, and B. Ommer, “High-resolution image synthesis with latent diffusion models,” inCVPR, 2022, pp. 10 684–10 695

    2022

  176. [187]

    First creating backgrounds then rendering texts: A new paradigm for visual text blending,

    Z. Li, Y. Shu, W. Zeng, D. Yang, and Y. Zhou, “First creating backgrounds then rendering texts: A new paradigm for visual text blending,” inECAI, 2024

    2024

  177. [188]

    GlyphControl: Glyph conditional control for visual text genera- tion,

    Y. Yang, D. Gui, Y. Yuan, W. Liang, H. Ding, H. Hu, and K. Chen, “GlyphControl: Glyph conditional control for visual text genera- tion,” inNeurIPS, 2023, pp. 44 050–44 066

    2023

  178. [189]

    Beyond flat text: Dual self- inherited guidance for visual text generation,

    M. Luo, Z. Xia, L. Chen, Z. Li, W. Zeng, J. Wang, W. Cheng, Y. Wang, Y. Zhou, and J. Yang, “Beyond flat text: Dual self- inherited guidance for visual text generation,”arXiv preprint arXiv:2501.05892, 2025

    2025 arXiv

  179. [190]

    Brush your text: Synthesize any scene text on images via diffusion model,

    L. Zhang, X. Chen, Y. Wang, Y. Lu, and Y. Qiao, “Brush your text: Synthesize any scene text on images via diffusion model,” inAAAI, vol. 38, no. 7, 2024, pp. 7215–7223

    2024

  180. [191]

    Visual text generation in the wild,

    Y. Zhu, J. Liu, F. Gao, W. Liu, X. Wang, P . Wang, F. Huang, C. Yao, and Z. Yang, “Visual text generation in the wild,” inECCV. Springer, 2024, pp. 89–106

    2024

  181. [192]

    Postermaker: Towards high-quality product poster generation with accurate text rendering,

    Y. Gao, Z. Lin, C. Liu, M. Zhou, T. Ge, B. Zheng, and H. Xie, “Postermaker: Towards high-quality product poster generation with accurate text rendering,”arXiv preprint arXiv:2504.06632, 2025

    2025 arXiv

  182. [193]

    Text image super- resolution guided by text structure and embedding priors,

    C. Huang, X. Peng, D. Liu, and Y. Lu, “Text image super- resolution guided by text structure and embedding priors,” TOMM, vol. 19, pp. 1 – 18, 2023

    2023

  183. [194]

    Multiscale structural similarity for image quality assessment,

    Z. Wang, E. Simoncelli, and A. Bovik, “Multiscale structural similarity for image quality assessment,” inACSSC, vol. 2, 2003, pp. 1398–1402 Vol.2

    2003

  184. [195]

    Multi-view document rectification using boundary,

    Y.-C. Tsoi and M. S. Brown, “Multi-view document rectification using boundary,” inCVPR, 2007, pp. 1–8

    2007

  185. [196]

    Learning from doc- uments in the wild to improve document unwarping,

    K. Ma, S. Das, Z. Shu, and D. Samaras, “Learning from doc- uments in the wild to improve document unwarping,”ACM SIGGRAPH, 2022

    2022

  186. [197]

    Convolutional neural networks for direct text deblurring,

    M. Hradi ˇs, J. Kotera, P . Zemcık, and F. ˇSroubek, “Convolutional neural networks for direct text deblurring,” inBMVC, vol. 10, no. 2, 2015

    2015

  187. [198]

    GANs trained by a two time-scale update rule converge to a local nash equilibrium,

    M. Heusel, H. Ramsauer, T. Unterthiner, B. Nessler, and S. Hochreiter, “GANs trained by a two time-scale update rule converge to a local nash equilibrium,”NeurIPS, vol. 30, 2017

    2017

  188. [199]

    LayoutTransformer: Layout generation and completion with self-attention,

    K. Gupta, J. Lazarow, A. Achille, L. S. Davis, V . Mahadevan, and A. Shrivastava, “LayoutTransformer: Layout generation and completion with self-attention,” inICCV, 2021, pp. 1004–1014

    2021

  189. [200]

    Geometric recti- fication of document images using adversarial gated unwarping network,

    X. Liu, G. Meng, B. Fan, S. Xiang, and C. Pan, “Geometric recti- fication of document images using adversarial gated unwarping network,”PR, vol. 108, p. 107576, 2020

    2020

  190. [201]

    End-to-end piece-wise unwarping of document images,

    S. Das, K. Y. Singh, J. Wu, E. Bas, V . Mahadevan, R. Bhotika, and D. Samaras, “End-to-end piece-wise unwarping of document images,”ICCV, pp. 4248–4257, 2021. 20

    2021

  191. [202]

    Document dewarp- ing with control points,

    G.-W. Xie, F. Yin, X.-Y. Zhang, and C.-L. Liu, “Document dewarp- ing with control points,” inICDAR. Springer, 2021, pp. 466–480

    2021

  192. [203]

    DocScanner: Robust document image rectification with progressive learning,

    H. Feng, W. gang Zhou, J. Deng, Q. Tian, and H. Li, “DocScanner: Robust document image rectification with progressive learning,” ArXiv, vol. abs/2110.14968, 2021

    2021 arXiv

  193. [204]

    Fourier document restoration for robust document dewarping and recognition,

    C. Xue, Z. Tian, F. Zhan, S. Lu, and S. Bai, “Fourier document restoration for robust document dewarping and recognition,” CVPR, pp. 4563–4572, 2022

    2022

  194. [205]

    Image-to-image translation with conditional adversarial networks,

    P . Isola, J.-Y. Zhu, T. Zhou, and A. A. Efros, “Image-to-image translation with conditional adversarial networks,” inCVPR, 2017, pp. 1125–1134

    2017

  195. [206]

    Multi-branch network with ensemble learning for text removal in the wild,

    Y. Hou, J. J. Chen, and Z. Wang, “Multi-branch network with ensemble learning for text removal in the wild,” inACCV, 2022, pp. 1333–1349

    2022

  196. [207]

    Modeling stroke mask for end-to-end text erasing,

    X. Du, Z. Zhou, Y. Zheng, T. Ma, X. Wu, and C. Jin, “Modeling stroke mask for end-to-end text erasing,” inWACV, 2023, pp. 6151–6159

    2023

  197. [208]

    What is the real need for scene text removal? Exploring the background integrity and erasure exhaustivity properties,

    Y. Wang, H. Xie, Z. Wang, Y. Qu, and Y. Zhang, “What is the real need for scene text removal? Exploring the background integrity and erasure exhaustivity properties,”TIP, 2023

    2023

  198. [209]

    Viteraser: Harnessing the power of vision transformers for scene text removal with segmim pretraining,

    D. Peng, C. Liu, Y. Liu, and L. Jin, “Viteraser: Harnessing the power of vision transformers for scene text removal with segmim pretraining,” inAAAI, 2024, pp. 4468–4477

    2024

  199. [210]

    LAION-400M: Open dataset of clip-filtered 400 million image- text pairs,

    C. Schuhmann, R. Vencu, R. Beaumont, R. Kaczmarczyk, C. Mullis, A. Katta, T. Coombes, J. Jitsev, and A. Komatsuzaki, “LAION-400M: Open dataset of clip-filtered 400 million image- text pairs,”arXiv preprint arXiv:2111.02114, 2021

    2021 arXiv

  200. [211]

    OpenAI, “Gpt-4o,” https://openai.com/index/hello-gpt-4o/, May 2024

    2024

  201. [212]

    Adding conditional control to text-to-image diffusion models,

    L. Zhang, A. Rao, and M. Agrawala, “Adding conditional control to text-to-image diffusion models,” inICCV, 2023, pp. 3836–3847

    2023

  202. [213]

    SnapFusion: Text-to-image diffusion model on mobile devices within two seconds,

    Y. Li, H. Wang, Q. Jin, J. Hu, P . Chemerys, Y. Fu, Y. Wang, S. Tulyakov, and J. Ren, “SnapFusion: Text-to-image diffusion model on mobile devices within two seconds,” inNeurIPS, 2023, pp. 20 662–20 678

    2023

  203. [214]

    STRIVE: Scene text replacement in videos,

    J. Subramanian, V . Chordia, E. Bart, S. Fang, K. Guan, R. Bala et al., “STRIVE: Scene text replacement in videos,” inICCV, 2021, pp. 14 549–14 558

    2021

  204. [215]

    Video-xl: Extra-long vision language model for hour- scale video understanding,

    Y. Shu, Z. Liu, P . Zhang, M. Qin, J. Zhou, Z. Liang, T. Huang, and B. Zhao, “Video-xl: Extra-long vision language model for hour- scale video understanding,”arXiv preprint arXiv:2409.14485, 2024

    2024 arXiv

  205. [216]

    UPOCR: Towards unified pixel-level ocr interface,

    D. Peng, Z. Yang, J. Zhang, C. Liu, Y. Shi, K. Ding, F. Guo, and L. Jin, “UPOCR: Towards unified pixel-level ocr interface,” in ICML, 2024

    2024

Pith tools

Reviewed August 16, 2026 · model on record in the stance chip above.