Pith. sign in

REVIEW 4 major objections 6 minor 39 references

DocSpiral: A Platform for Integrated Assistive Document Annotation through Human-in-the-Spiral

T0 review · 4 major / 6 minor · reviewed 2026-08-15 · deepseek-v4-flash

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

Pith's one-line read Human-in-the-spiral annotation cuts document labeling time by 41%.

desk verdict DocSpiral is a real system contribution with a plausible 'first' claim for its integrated workflow, but its headline numbers are thin and need a protocol before they are trustable. read the letter →

arxiv 2505.03214 v1 pith:DQMKS3K7 submitted 2025-05-06 cs.SE cs.AI

Qiang Sun , Sirui Li , Tingting Bi , Du Huynh , Mark Reynolds , Yuanyi Luo , Wei Liu This is my paper · ORCID
classification cs.SEcs.AI
keywords documentannotationhuman-in-the-looplayoutdetectionOCRverificationiterativemodeltrainingscannedprocessingstructureddataextractionassistive
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

The paper introduces DocSpiral, a web-based annotation platform for image-based documents such as scanned reports. Its central claim is that a "human-in-the-spiral" workflow, where humans verify and correct the outputs of baseline models and those corrections train progressively better models, reduces manual annotation time by at least 41% and improves model performance with each cycle. The motivation is that domain-specific documents like geological reports and medical records are often scanned images, and extracting structured data from them currently demands heavy human labeling. If the claim holds, DocSpiral offers a practical route to cheaper, faster creation of training data for document-understanding models in fields where such data are scarce.

What carries the argument

The central mechanism is the human-in-the-spiral loop: an iterative cycle in which uploaded documents are converted to PDF and processed by baseline models, human annotators review and correct the outputs through an interactive web interface with dynamic form generation, the verified annotations become training data, and retrained progressive models are fed back into the pipeline to reduce the human effort needed in the next round. This loop is what differentiates DocSpiral from one-shot annotation tools; the platform's other components, such as format normalization, customizable layout schemas, a metrics dashboard, and API endpoints, support the loop rather than replace it.

What would settle it

Reproduce the Faster-RCNN training with a strict held-out test set that is never seen during the three rounds; if mAP does not rise as in Table 2, or declines when test pages are re-randomized, the spiral's performance gain is an artifact. Similarly, run a counterbalanced annotation study with fresh annotators assigned to manual and assisted conditions; if the assisted condition does not show a significant per-page time reduction, the 41% claim is refuted.

Watch

Extended reading notes

Core claim

DocSpiral's central discovery is that a closed loop between human annotation and model training makes document annotation self-improving. Starting from baseline models for layout detection, OCR, and table, formula, and figure processing, the platform lets experts review, correct, and augment model outputs. The corrected data are then used to train or fine-tune progressive models, which in turn produce better initial outputs for the next round of human review. In a 90-page annotation experiment, this assistance cut per-page processing time from 28.4 seconds to 16.7 seconds, a 41% reduction, and up to 75% for low-quality scans. Separate training runs with a Faster-RCNN detector over three iterations, adding 100 pages each round, raised layout-detection mAP from 0.053 to 0.33, demonstrating the upward performance spiral.

Load-bearing premise

The reported 41% time savings and steady mAP gains depend on a controlled evaluation where the mAP reflects a held-out test set not contaminated by the incrementally added training pages, and where the manual-versus-assisted time comparison isolates annotator skill and learning effects; the paper does not spell out this protocol.

Editorial extensions

If this is right

  • Domain teams can bootstrap document extraction from scanned collections with limited initial labels, since baseline models provide a first pass and only corrections require human effort.
  • Each deployment round compounds: more annotated data yields better progressive models, which lower the cost of annotating the next batch.
  • The same platform can serve rule-based fixes, deep-learning fine-tuning, and LLM alignment, because the annotation schema and API endpoints are model-agnostic.
  • Open, freely accessible hosting lowers the entry barrier for geoscience, healthcare, and other document-intensive fields to adopt LLM-based extraction.
  • Dynamic form generation lets a single tool handle diverse output formats, including LaTeX, HTML, JSON, and descriptive text, for tables, formulas, and figures.

Reading between the lines

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

  • If the 41% figure generalizes, the cost model for creating domain-specific document datasets would shift from raw labeling effort to expert review capacity, making small expert teams viable for large collections.
  • The spiral design implies a concrete testable hypothesis: the marginal time saving per page should increase with each iteration and approach a floor set by the ceiling of baseline-model accuracy; tracking per-round review time would verify this.
  • A potential hidden risk is feedback-loop bias: if progressive models are trained on the same pages whose corrections seed them, the reported mAP gains could reflect memorization rather than true generalization, making the paper's omitted test-set protocol the first thing to check.
  • The architecture could generalize beyond documents: any image-to-structured-output task with a measurable model, such as medical imaging annotations, could adopt the same human-in-the-spiral loop.
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 / 6 minor

Summary. The paper introduces DocSpiral, a web-based annotation platform that integrates document format normalization (Anything2PDF), layout detection, OCR, table/figure/formula annotation, and model-development API endpoints into a single human-in-the-spiral workflow. The central idea is that human corrections on baseline model outputs generate training data that improve progressive models, which in turn reduce future manual effort. The authors report two quantitative results: a 41% overall annotation-time reduction (from 28.4s to 16.7s per page) in a 90-page experiment, and a Faster-RCNN layout-detection mAP progression from 0.053 to 0.33 across three training iterations, each adding 100 pages. The paper also provides a feature comparison table and a system-architecture description. The main contributions claimed are the comprehensive annotation interface, the assisted spiral improvement framework, and the open, deployable system.

Significance. If the human-in-the-spiral claims are substantiated, DocSpiral addresses a real pain point: curating training data for domain-specific, image-heavy document collections where existing tools stop at isolated annotation tasks. The platform design is constructive, and the paper's contribution as a systems paper is valuable: it unifies layout, OCR, table, figure, and formula annotation; supports custom schemas; and exposes RESTful APIs for model development, all in an openly accessible deployment. The software appears usable, and the freemium access lowers the barrier for researchers in geoscience and healthcare. However, the quantitative evidence is presently insufficient to support the headline claims, especially the mAP improvement, which is the only direct support for the 'spiral' feedback loop. The evaluation needs a clearly defined test protocol, controls, and uncertainty quantification before the contributions can be deemed valid.

major comments (4)
  1. [Section 4, Table 2] The claim of 'consistent performance gains across three iterations' is not supported without specifying the evaluation protocol. The manuscript states that each iteration adds 100 new pages of data to training, but it does not state how the 90-page experiment relates to the mAP evaluation, how the test set was constructed, or whether the test pages overlap with the incrementally added training pages. If the mAP is measured on pages that were used for training in a later iteration, the improvement from 0.053 to 0.33 is a data-leakage artifact rather than evidence of an upward spiral. Please provide the full protocol: the split between training and test pages at each iteration, the independence of the test set from all training data, the number of evaluation pages, and confidence intervals or error bars. In addition, a control condition that trains on the same number of randomly selected pages (rather than pages selected through the human-in-the-spiral loop) is needed to attribute the gains to the workflow rather than to mere additional data.
  2. [Section 4, timing experiment] The annotation-time comparison (28.4s vs. 16.7s per page, yielding 'at least 41%' reduction, and 75% for low-quality scans) is a single point estimate with no reported variance, number of annotators, or experimental design details. With only 90 pages, individual annotator skill and learning effects can easily shift the measured ratio. The manuscript does not describe whether the manual and assisted conditions were performed by the same annotators, whether tasks were counterbalanced, or whether the reported times include model-inference latency. Please report the number of annotators, the distribution of per-page times, a statistical significance test, and a clear definition of what constitutes 'assisted' versus 'manual' annotation, including how pre-filled predictions are scored.
  3. [Section 4, spiral mechanism] The 'human-in-the-spiral' claim is only partially tested by the mAP progression. The argument that the loop reduces manual intervention requires measuring the human correction effort per iteration (e.g., number of edits per page, time per page) as the progressive models improve. Table 2 shows model quality improving, but it does not show that the human workload decreases across iterations, which is the second half of the spiral. Without such a measurement, the abstract's statement that the platform makes models 'progressively require less manual intervention' is an extrapolation rather than an empirical result. Please add per-iteration measurements of annotation effort, or explicitly restrict the claim to model-performance gains.
  4. [Section 4, Table 2 and Section 3.3] The Faster-RCNN training setup is under-specified. The manuscript does not state the backbone, hyperparameters, or how the 100 pages per iteration were selected and annotated. It also does not provide the mAP definition (e.g., COCO-style AP@IoU=0.5 or AP@[0.5:0.95]) or the number of classes and class distribution in the evaluation. Because the default layout classes in the platform are few (content, title, figure, table, formula, footnote), a small test set with imbalanced classes could make the reported mAP values highly variable. Please provide the exact evaluation metric, the test-set construction, and the per-class breakdown, and consider reporting multiple random seeds or trial repetitions.
minor comments (6)
  1. [Abstract and Section 1] The claim 'at least 41% time reduction' is inconsistent with the single experimental point; 'at least' implies a lower bound across conditions, but the manuscript only reports one average. Please rephrase to 'a 41% reduction in our experiment' unless more data are provided.
  2. [Author affiliation] The affiliation string contains a typo: 'Melborune' should be 'Melbourne'.
  3. [Section 1] The word 'domain-specfic' is misspelled; should be 'domain-specific'.
  4. [Section 3.3] The phrase 'you can speficy your own baseline model' contains a typo: 'speficy' should be 'specify'.
  5. [Figure 5 caption] The caption reads 'Open evaluation eetrics'; it should be 'Open evaluation metrics'.
  6. [Table 1] The table claims DocSpiral supports 'Layout Ann.' etc., but the binary checkmarks do not indicate the depth or quality of support. A brief note in the caption on how the capabilities were verified would strengthen the comparison.

Circularity Check

0 steps flagged · score 0.0 of 10

No structural circularity found; the paper's claims are empirical measurements, and the only same-author citation is non-load-bearing.

full rationale

DocSpiral is a systems/empirical paper. The claimed reductions ('reduces annotation time by at least 41%') and the Faster-RCNN mAP progression in Table 2 are measurements, not analytical derivations, and the paper provides no equations or construction by which the outputs are forced to equal the inputs. The annotation-time comparison is an experiment; the mAP sequence is reported as a training experiment, but the paper does not state the evaluation protocol (e.g., held-out test set, confidence intervals, control condition). That under-specification is a correctness and reproducibility risk, not a demonstrated circularity: nothing in the quoted text shows that the mAP values are computed on the same pages added to training or otherwise constructed to guarantee improvement. The only same-author citation (Stewart and Liu 2020) appears in the introduction as an illustrative reference for Retrieval Augmented Generation and is not used to justify any load-bearing premise, uniqueness theorem, or ansatz. No self-definitional, fitted-input, imported-uniqueness, or renaming circularity is exhibited in the manuscript. Per the hard rule requiring a quoted reduction, no circular step can be substantiated.

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

No explicit free parameters are fitted in this paper; the quantitative claims are empirical measurements from a 90-page annotation experiment and three rounds of Faster-RCNN training. The lack of detailed training hyperparameters means the model training is not fully specified, but no numbers are chosen ad hoc to derive the headline results. The paper introduces no new physical or mathematical entities.

assumptions (3)
  • domain assumption The baseline models (DocLayout-YOLO, PaddleOCR, Pix2Text, vision LLM) function as described in their respective references.
    The platform's assistive value assumes these models provide usable pre-annotations; no integration tests or quality thresholds are reported.
  • domain assumption The 90-page evaluation dataset is representative of domain-specific image-based documents.
    The paper gives no information about page selection, document sources, or annotator background, so the 41% figure may not generalize.
  • domain assumption The Faster-RCNN mAP values are computed on a held-out test set.
    Section 4 and Table 2 do not state whether the 100-page increments are used for training and evaluation separately; without a held-out test, the mAP gains could be due to memorization.

how reviews work

0 comments
Cite this review

Pith. "Pith review of DocSpiral: A Platform for Integrated Assistive Document Annotation through Human-in-the-Spiral." pith.science (2026). https://pith.science/paper/DQMKS3K7

@misc{pith2026250503214,
  author       = {Pith},
  title        = {Pith review of: DocSpiral: A Platform for Integrated Assistive Document Annotation through Human-in-the-Spiral},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/DQMKS3K7}},
  note         = {Machine review of arXiv:2505.03214}
}
read the original abstract

Acquiring structured data from domain-specific, image-based documents such as scanned reports is crucial for many downstream tasks but remains challenging due to document variability. Many of these documents exist as images rather than as machine-readable text, which requires human annotation to train automated extraction systems. We present DocSpiral, the first Human-in-the-Spiral assistive document annotation platform, designed to address the challenge of extracting structured information from domain-specific, image-based document collections. Our spiral design establishes an iterative cycle in which human annotations train models that progressively require less manual intervention. DocSpiral integrates document format normalization, comprehensive annotation interfaces, evaluation metrics dashboard, and API endpoints for the development of AI / ML models into a unified workflow. Experiments demonstrate that our framework reduces annotation time by at least 41\% while showing consistent performance gains across three iterations during model training. By making this annotation platform freely accessible, we aim to lower barriers to AI/ML models development in document processing, facilitating the adoption of large language models in image-based, document-intensive fields such as geoscience and healthcare. The system is freely available at: https://app.ai4wa.com. The demonstration video is available: https://app.ai4wa.com/docs/docspiral/demo.

Figures

Figures reproduced from arXiv: 2505.03214 by the authors.

Figure 1
Figure 1. Our DocSpiral framework converts docu￾ments to PDF and processes them through iterative cy￾cles where human verification creates annotations that improve AI/ML models, reducing effort and enhancing performance within each iteration. Most existing document processing frame￾works (Faysse et al., 2025; Shen et al., 2021; Wang et al., 2024) rely on general purpose pipelines that convert raw documents into machine-readab… view at source ↗
Figure 3
Figure 3. Documents upload and management interface, [PITH_FULL_IMAGE:figures/full_fig_p004_3.png] view at source ↗
Figure 4
Figure 4. Layout annotation interface: user can click, [PITH_FULL_IMAGE:figures/full_fig_p005_4.png] view at source ↗
Figures from the paper (4 more)
Figure 6
Figure 6. Figure 6: Figure annotation interface in review mode with JSON viewer (left); Formula in review mode showing latex output in form (middle); Table in annotation mode with editable output field from html model using schema-generated form (right) [PITH_FULL_IMAGE:figures/full_fig_…
Figure 5
Figure 5. Figure 5: OCR verification and annotation interface [PITH_FULL_IMAGE:figures/full_fig_p005_5.png]
Figure 7
Figure 7. Figure 7: The settings interface allows configuring form [PITH_FULL_IMAGE:figures/full_fig_p005_7.png]
Figure 8
Figure 8. Figure 8: Table model performance dashboard display [PITH_FULL_IMAGE:figures/full_fig_p006_8.png]

Discussion (0). Continue with ORCID to comment.

Reference graph

Works this paper leans on

39 extracted references · 24 canonical work pages

  1. [1]

    online" 'onlinestring :=

    ENTRY address archivePrefix author booktitle chapter edition editor eid eprint eprinttype howpublished institution journal key month note number organization pages publisher school series title type volume year doi pubmed url lastchecked label extra.label sort.label short.list INTEGERS output.state before.all mid.sentence after.sentence after.block STRING...

  2. [2]

    write newline

    " write newline "" before.all 'output.state := FUNCTION n.dashify 't := "" t empty not t #1 #1 substring "-" = t #1 #2 substring "--" = not "--" * t #2 global.max substring 't := t #1 #1 substring "-" = "-" * t #2 global.max substring 't := while if t #1 #1 substring * t #2 global.max substring 't := if while FUNCTION word.in bbl.in capitalize " " * FUNCT...

  3. [3]

    ABBYY. 1993. https://pdf.abbyy.com/ Abbyy finereader pdf . Commercial document conversion and OCR software

    work page 1993

  4. [4]

    Iftakhar Ali Khandokar and Priya Deshpande. 2025. https://doi.org/10.1109/ACCESS.2024.3522141 Computer vision-based framework for data extraction from heterogeneous financial tables: A comprehensive approach to unlocking financial insights . IEEE Access, 13:17706--17723

    arXiv 2025

  5. [5]

    Justin Brooks. 2019. COCO Annotator . https://github.com/jsbroks/coco-annotator/

    work page 2019

  6. [6]

    Explosion AI . 2023. https://github.com/explosion/prodigy-pdf Prodigy pdf . PDF annotation plugin for Prodigy

    work page 2023

  7. [7]

    Wenqi Fan, Yujuan Ding, Liangbo Ning, Shijie Wang, Hengyun Li, Dawei Yin, Tat-Seng Chua, and Qing Li. 2024. https://doi.org/10.1145/3637528.3671470 A survey on rag meeting llms: Towards retrieval-augmented large language models . In Proceedings of the 30th ACM SIGKDD Conference on Knowledge Discovery and Data Mining, KDD '24, page 6491–6501, New York, NY,...

    arXiv 2024

  8. [8]

    Manuel Faysse, Hugues Sibille, Tony Wu, Bilel Omrani, Gautier Viaud, CELINE HUDELOT, and Pierre Colombo. 2025. https://openreview.net/forum?id=ogjBpZ8uSi Colpali: Efficient document retrieval with vision language models . In The Thirteenth International Conference on Learning Representations

    2025

Show all 39 references
  1. [9]

    gipplab . 2019. Annomathtex. https://github.com/gipplab/AnnoMathTeX. Accessed: 2025-03-19

    2019

  2. [10]

    Robert Heeg. 2023. Possibilities of unstructured data. https://shorturl.at/RaoHc. Accessed: 2025-03-28

    2023

  3. [11]

    Yupan Huang, Tengchao Lv, Lei Cui, Yutong Lu, and Furu Wei. 2022. https://arxiv.org/abs/2204.08387 Layoutlmv3: Pre-training for document ai with unified text and image masking . Preprint, arXiv:2204.08387

    2022 arXiv

  4. [12]

    Viet-Phi Huynh, Yoan Chabot, Thomas Labb \'e , Jixiong Liu, and Rapha \"e l Troncy. 2022. From Heuristics to Language Models: A Journey Through the Universe of Semantic Table Interpretation with DAGOBAH . In Semantic Web Challenge on Tabular Data to Knowledge Graph Matching (SemTab)

    2022

  5. [13]

    Christopher Kermorvant, Eva Bardou, Manon Blanco, and Bastien Abadie. 2024. https://arxiv.org/abs/2405.01071 Callico: a versatile open-source document image annotation platform . Preprint, arXiv:2405.01071

    2024 arXiv

  6. [14]

    Saeid Nahavandi. 2017. Trusted autonomy between humans and robots: Toward human-on-the-loop in robotics and autonomous systems. IEEE Systems, Man, and Cybernetics Magazine, 3(1):10--17

    2017

  7. [15]

    Mark Neumann, Zejiang Shen, and Sam Skjonsberg. 2021. https://doi.org/10.18653/v1/2021.acl-demo.31 PAWLS : PDF annotation with labels and structure . In Proceedings of the 59th Annual Meeting of the Association for Computational Linguistics and the 11th International Joint Con...

    2021 doi

  8. [16]

    Nick-Barney. 2025. What is unstructured data? https://www.techtarget.com/searchbusinessanalytics/definition/unstructured-data. Accessed: 2025-03-28

    2025

  9. [17]

    Alejandro Pe \ n a, Aythami Morales, Julian Fierrez, Javier Ortega-Garcia, I \ n igo Puente, Jorge Cordova, and Gonzalo Cordova. 2024. Continuous document layout analysis: Human-in-the-loop ai-based data curation, database, and evaluation in the domain of public affairs. Infor...

    2024

  10. [18]

    PFCCLab. 2020. https://github.com/PFCCLab/PPOCRLabel Ppocrlabel . Annotation tool for OCR tasks based on PaddleOCR

    2020

  11. [19]

    READ-COOP SCE . 2013. https://www.transkribus.org/ Transkribus . Platform for the automated recognition, transcription and searching of historical documents

    2013

  12. [20]

    Shaoqing Ren, Kaiming He, Ross Girshick, and Jian Sun. 2016. https://arxiv.org/abs/1506.01497 Faster r-cnn: Towards real-time object detection with region proposal networks . Preprint, arXiv:1506.01497

    2016 arXiv

  13. [21]

    Monica Riedler and Stefan Langer. 2024. https://arxiv.org/abs/2410.21943 Beyond text: Optimizing rag with multimodal inputs for industrial applications . Preprint, arXiv:2410.21943

    2024 arXiv

  14. [22]

    Farrell, Margaret J

    Angela Riganti, Terence R. Farrell, Margaret J. Ellis, Felicia Irimies, Colin D. Strickland, Sarah K. Martin, and Darren J. Wallace. 2015. https://doi.org/10.1016/j.grj.2015.02.015 125years of legacy data at the geological survey of western australia: Capture and delivery . Ge...

    2015 doi

  15. [23]

    Zejiang Shen, Ruochen Zhang, Melissa Dell, Benjamin Charles Germain Lee, Jacob Carlson, and Weining Li. 2021. Layoutparser: A unified toolkit for deep learning based document image analysis. arXiv preprint arXiv:2103.15348

    2021 arXiv

  16. [24]

    Hiroyuki Shindo, Yohei Munesada, and Yuji Matsumoto. 2018. https://aclanthology.org/L18-1175/ PDFA nno: a web-based linguistic annotation tool for PDF documents . In Proceedings of the Eleventh International Conference on Language Resources and Evaluation ( LREC 2018) , Miyaza...

    2018

  17. [25]

    Brandon Smock, Rohith Pesala, and Robin Abraham. 2023. Aligning benchmark datasets for table structure recognition. pages 371--386

    2023

  18. [26]

    Michael Stewart and Wei Liu. 2020. Seq2kg: An end-to-end neural model for domain agnostic knowledge graph (not text graph) construction from text. In Proceedings of the International Conference on Principles of Knowledge Representation and Reasoning, volume 17, pages 748--757

    2020

  19. [27]

    TagTog team . 2023. https://docs.tagtog.com/ Tagtog . Web application. Web-based text annotation platform for machine learning and AI with project management capabilities

    2023

  20. [28]

    Deep Search Team. 2024. https://doi.org/10.48550/arXiv.2408.09869 Docling technical report . Technical report

    2024 doi arXiv

  21. [29]

    Maxim Tkachenko, Mikhail Malyuk, Andrey Holmanyuk, and Nikolai Liubimov. 2020. https://github.com/HumanSignal/label-studio Label Studio : Data labeling software . Open source software available from https://github.com/HumanSignal/label-studio

    2020

  22. [30]

    Jianqiang Wan, Sibo Song, Wenwen Yu, Yuliang Liu, Wenqing Cheng, Fei Huang, Xiang Bai, Cong Yao, and Zhibo Yang. 2024. Omniparser: A unified framework for text spotting key information extraction and table recognition. In Proceedings of the IEEE/CVF Conference on Computer Visi...

    2024

  23. [31]

    Bin Wang, Chao Xu, Xiaomeng Zhao, Linke Ouyang, Fan Wu, Zhiyuan Zhao, Rui Xu, Kaiwen Liu, Yuan Qu, Fukai Shang, Bo Zhang, Liqun Wei, Zhihao Sui, Wei Li, Botian Shi, Yu Qiao, Dahua Lin, and Conghui He. 2024. https://arxiv.org/abs/2409.18839 Mineru: An open-source solution for p...

    2024 arXiv

  24. [32]

    Xingjiao Wu, Luwei Xiao, Yixuan Sun, Junhang Zhang, Tianlong Ma, and Liang He. 2022. A survey of human-in-the-loop for machine learning. Future Generation Computer Systems, 135:364--381

    2022

  25. [33]

    Renqiu Xia, Song Mao, Xiangchao Yan, Hongbin Zhou, Bo Zhang, Haoyang Peng, Jiahao Pi, Daocheng Fu, Wenjie Wu, Hancheng Ye, et al. 2024. Docgenome: An open large-scale scientific document benchmark for training and testing multi-modal large language models. arXiv preprint arXiv...

    2024

  26. [34]

    Renqiu Xia, Hongbin Zhou, Ziming Feng, Huanxi Liu, Boan Chen, Bo Zhang, and Junchi Yan. 2025. Latexnet: A specialized model for converting visual tables and equations to latex code. In ICASSP 2025-2025 IEEE International Conference on Acoustics, Speech and Signal Processing (I...

    2025

  27. [35]

    Qiu, and Lili Qiu

    Siyun Zhao, Yuqing Yang, Zilong Wang, Zhiyuan He, Luna K. Qiu, and Lili Qiu. 2024 a . https://arxiv.org/abs/2409.14924 Retrieval augmented generation (rag) and beyond: A comprehensive survey on how to make your llms use external data more wisely . Preprint, arXiv:2409.14924

    2024 arXiv

  28. [36]

    Weichao Zhao, Hao Feng, Qi Liu, Jingqun Tang, Shu Wei, Binghong Wu, Lei Liao, Yongjie Ye, Hao Liu, Wengang Zhou, Houqiang Li, and Can Huang. 2024 b . https://proceedings.neurips.cc/paper_files/paper/2024/file/0d97fe65d7a1dc12a05642d9fa4cd578-Paper-Conference.pdf Tabpedia: Towa...

    2024

  29. [37]

    Weichao Zhao, Hao Feng, Qi Liu, Jingqun Tang, Binghong Wu, Lei Liao, Shu Wei, Yongjie Ye, Hao Liu, Wengang Zhou, et al. 2024 c . Tabpedia: Towards comprehensive visual table understanding with concept synergy. Advances in Neural Information Processing Systems, 37:7185--7212

    2024

  30. [38]

    Zhiyuan Zhao, Hengrui Kang, Bin Wang, and Conghui He. 2024 d . https://arxiv.org/abs/2410.12628 Doclayout-yolo: Enhancing document layout analysis through diverse synthetic data and global-to-local adaptive perception . Preprint, arXiv:2410.12628

    2024 arXiv

  31. [39]

    Yinghao Zhu, Changyu Ren, Shiyun Xie, Shukai Liu, Hangyuan Ji, Zixiang Wang, Tao Sun, Long He, Zhoujun Li, Xi Zhu, and Chengwei Pan. 2024. https://arxiv.org/abs/2402.07016 Realm: Rag-driven enhancement of multimodal electronic health records analysis via large language models ...

    2024 arXiv

Pith tools

Reviewed August 15, 2026 · model on record in the stance chip above.