Pith. sign in

REVIEW 4 major objections 5 minor 2 cited by

MIT-10M: A Large Scale Parallel Corpus of Multilingual Image Translation

T0 review · 4 major / 5 minor · reviewed 2026-08-11 · deepseek-v4-flash

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

Pith's one-line read The MIT-10M dataset provides 840,855 real-world images and 10,931,115 image–text pairs in 14 languages, claimed to be the largest corpus of its kind.

desk verdict A genuinely large and useful image-translation dataset, with real but fixable evaluation problems: machine-generated references, a too-small human check, and headline numbers that don't match the tables. read the letter →

arxiv 2412.07147 v2 pith:QRCHW3XY submitted 2024-12-10 cs.CV cs.AI

Bo Li , Shaolin Zhu , Lijie Wen This is my paper · ORCID
classification cs.CVcs.AI
keywords imagetranslationmultilingualdatasetvision-languagemodelsfine-tuningreal-worldcorpuse-commerceimagesOCRmachineevaluation
verification ladder T0 review T1 audit T2 compute T3 formal

The pith

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

The reading

This paper introduces MIT-10M, a multilingual image translation corpus built from 840,855 high-resolution images crawled from real e-commerce and general web pages, with 10,931,115 image–text pairs spanning 14 languages and 28 categories. The authors claim this is the largest real-world image translation dataset to date, and they support the claim with a three-stage pipeline that filters images, recognizes embedded text with GPT-4o, translates it with GPT-4, and keeps translations only when they agree with Google Translate at a spaCy cosine similarity above 0.8. The corpus includes easy, medium, and hard splits based on text length and bounding-box count, plus images at three resolutions. On the MIT-10M test set, seven state-of-the-art models score low on BLEU, chrF++, and METEOR, which the authors take as evidence that the benchmark reflects real difficulty; fine-tuning Qwen2-VL on the training set raises BLEU by 230%, chrF++ by 88%, and METEOR by 130%.

What carries the argument

The central object is the MIT-10M corpus and its three-stage construction pipeline. Web pages are crawled and deduplicated; EasyOCR, langid, and langdetect prefilter images that contain text; GPT-4o performs precise OCR; GPT-4 translates the recognized text into 13 languages; and Google Gemini 1.5 Pro provides a second translation whose spaCy cosine similarity must exceed 0.8 for a pair to be kept. Difficulty labels come from bounding-box count and token length, and each image is released at original, 768-pixel, and 500-pixel widths. The dataset is both the training resource and the evaluation benchmark, so the pipeline's filtering decisions determine what the reported gains measure.

What would settle it

Sample a random set of MIT-10M reference translations, have professional translators rate them blind for adequacy and fluency, and compare the scores with the claimed 99.4% accuracy from the 10 human-checked batches; if accuracy is materially lower, or if the fine-tuned model's lead shrinks when evaluated against human-translated references instead of GPT-4/Google-style references, the central claim would need revision.

Watch

Extended reading notes

Core claim

The paper constructs MIT-10M, a corpus of 840,855 high-resolution images crawled from real e-commerce web pages in eight source languages, paired with machine translations into 13 target languages plus the original, for a total of 10,931,115 image–text pairs across 14 languages and 28 product categories. Each sample carries metadata — bounding-box count, token length, category, difficulty level, and three image sizes — and the pipeline filters translations by asking GPT-4 and Google Translate to agree, requiring a spaCy cosine similarity above 0.8, with 10 of 10,000 batches human-checked at 99.4% accuracy. On the MIT-10M test set, seven end-to-end and cascade models score low on BLEU, chrF++, and METEOR, which the authors read as evidence that the benchmark captures real-world difficulty; fine-tuning Qwen2-VL on the MIT-10M training set raises BLEU by 230%, chrF++ by 88%, and METEOR by 130%, and a controlled comparison against IIMT at similar data size shows MIT-10M yields higher BLEU on four language pairs.

Load-bearing premise

The reference translations generated by GPT-4 and kept only when Google Translate's translation has spaCy cosine similarity above 0.8 are accurate enough to serve as gold labels for both training and evaluation.

Editorial extensions

If this is right

  • With 14 languages and 28 categories, MIT-10M lets image translation models be trained and compared across a much wider language and domain spread than previous real-world datasets.
  • Because the test set includes easy, medium, and hard splits, models can be analyzed for how performance degrades as text length and bounding-box count grow; the paper reports average METEOR falling from 18.3 on easy to 14.0 on hard.
  • Fine-tuning on MIT-10M improves Qwen2-VL's BLEU from 14.6 to 35.9, chrF++ from 29.0 to 56.6, and METEOR from 21.2 to 52.0 at full training data, and even 10% of the data beats the base model.
  • At controlled data size, fine-tuning on MIT-10M outperforms fine-tuning on IIMT on DE-EN, EN-DE, EN-FR, and FR-EN BLEU.

Reading between the lines

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

  • If GPT-4's style is systematically embedded in the reference labels, part of the fine-tuning gain may reflect imitation of GPT-4 rather than general translation quality; evaluating against human-curated references would separate these effects.
  • The 0.8 cosine-similarity cutoff between two machine translators may systematically discard hard, idiomatic, or low-resource text, so the corpus could under-represent the most difficult real-world translation cases.
  • The same pipeline could be applied to additional domains such as social media screenshots, signage, and documents, or extended to other multimodal generation tasks that need aligned image–text data.
  • The per-sample metadata could support controlled studies of how vision-language models use resolution and layout cues, since each image is released at three sizes.
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 introduces MIT-10M, a claimed large-scale real-world multilingual image translation dataset with 840,855 images and 10,931,115 image-text pairs across 14 languages, 28 categories, and three difficulty levels. The construction pipeline crawls images from multilingual e-commerce and general websites, applies OCR and language identification, produces translations with GPT-4, and retains translations whose spaCy-based semantic similarity to Google Translate exceeds 0.8. The authors evaluate seven image translation models on the resulting test set and fine-tune Qwen2-VL on the training set, reporting large BLEU, chrF++, and METEOR gains, as well as a fine-tuning comparison against the IIMT dataset.

Significance. If the dataset quality claims withstand scrutiny, MIT-10M would be a valuable community resource: it is an order of magnitude larger than existing real-world image translation datasets, it is publicly released, and it provides multi-resolution images, per-category metadata, and difficulty annotations. The paper also ships a broad comparison of current MLLMs on a realistic, multilingual image translation benchmark. However, the central quantitative claims currently rest on self-referential reference labels and on evaluation evidence that is too thin to support the headline numbers, so the contribution needs substantial additional validation before the dataset can be recommended for benchmarking.

major comments (4)
  1. [§3.3 and §5.4] The reference translations used for both training and evaluation are produced by GPT-4 and retained only when spaCy cosine similarity to Google Translate exceeds 0.8. Because the test references come from the same pipeline, the reported fine-tuning gains partly measure how well a model imitates the GPT-4/Google-Translate label distribution rather than how well it translates into human-quality target text. The manuscript's own Limitations section concedes that 'the inherent complexity of multilingual data may lead to inaccuracies that could affect the reliability of the dataset.' To support the central claim, the authors should validate references against independent human translations on a held-out sample, ideally stratified by language and difficulty, and report automatic metrics computed against those human references.
  2. [§3.3] The human evaluation covers only 10 of 10,000 batches (0.1%), and the protocol is not described: it is unclear what 'translation accuracy' means, whether judgments are made per sentence or per batch, how many annotators participated, or how the 99.4% figure is computed. No confidence interval or inter-annotator agreement is reported. A sample of this size cannot validate a 10.9M-pair corpus spanning 13 target languages, especially when the same pipeline is used to generate the evaluation references. The authors should report the complete protocol and a substantially larger, stratified human evaluation.
  3. [Abstract, §1, and §5.4/Figure 6] The reported improvements are inconsistent. The abstract and introduction state that BLEU, chrF++, and METEOR increased by 230%, 88%, and 130%, respectively, but Figure 6 shows BLEU moving from 14.6 to 35.9 (+146%), chrF++ from 29.0 to 56.6 (+95%), and METEOR from 21.2 to 52.0 (+145%). The claim in the abstract that performance 'has tripled' is likewise not supported by the numbers in §5.4. The authors should correct the headline percentages in all locations and clarify whether the abstract refers to a different training configuration or metric.
  4. [§3.3, §4.3, and §5.4] The quantitative fine-tuning and evaluation results lack uncertainty estimates and are reported as single runs with no error bars or significance tests. This is especially problematic given the degree of label noise expected from the machine-translation pipeline and the strong language imbalance documented in §4.3 (English is about 49% of the data and Chinese about 60% of the English count) and in Table 3 (e.g., Digital Music has 44 images and Books 211). The claim that fine-tuning with MIT-10M 'tripled' performance should therefore be supported by multiple random seeds, confidence intervals, and per-language or per-difficulty breakdowns.
minor comments (5)
  1. [§3.1 and §4.4] The resolution criterion is stated inconsistently: §3.1 says images with resolution above 800x800 pixels are kept, while §4.4 says MIT-10M selects images with resolution above 1000x1000 pixels. Please clarify which threshold was actually used.
  2. [Throughout] There are several typographical errors: the §3.3 heading reads 'translationn', Figure 6 has 'BLUE' instead of 'BLEU', Figure 2 uses 'SRC_TEXT(ZN)' for Chinese, and the appendix shows 'SRC_TOEKN_LEN' instead of 'SRC_TOKEN_LEN'.
  3. [§4.2 and §5.3] The difficulty terminology is inconsistent: §4.2 defines 'Easy', 'Medium', and 'Hard', while §5.3 and Table 5 use 'Easy', 'Medium', and 'Difficult'. Please standardize.
  4. [Appendix A.1] The JSON schema in the prompt contains 'en' twice, which is not a valid dictionary schema and may confuse readers trying to reproduce the OCR/translation prompt.
  5. [§5.4] The sentence describing Figure 6 says 'BLEU increases from 14.6 for the base model (base@10%)', but the figure legend labels the first model 'Base'; this should be clarified to avoid ambiguity about which data subset is being referred to.

Circularity Check

1 steps flagged · score 6.0 of 10

Fine-tuning gains are measured against GPT-4-generated references, so the headline improvements partly measure imitation of the label generator rather than independent translation quality.

  1. fitted input called prediction [See Section 3.3 (Multilingual text translationn) and Section 5.4 (Fine-tune with MIT-10M), Figure 6.]
    "To translate the text into other languages, we first use GPT-4 ... We use the tool spacy to convert the results of GPT-4 and Google Translate into word vectors and calculate their semantic similarity. For text pairs with a similarity score below 0.8, we filter out those with significant differences and keep translations with a semantic similarity above 0.8. ... the BLEU increases from 14.6 for the base model (base@10%) to 35.9."

    The training targets for the fine-tuned model are GPT-4's translations, and the test-set references used in BLEU, chrF++ and METEOR come from the same GPT-4 pipeline, filtered only by similarity to Google Translate. A model fine-tuned to reproduce GPT-4's output distribution will therefore score higher against GPT-4-generated references by construction, even if human-judged translation quality does not improve. The 10-batch human spot check (99.4% accuracy) is too small and its protocol is undescribed to break this dependence, and the paper's own Limitations section concedes that translation inaccuracies may remain. The fine-tuning gains are thus partly a measure of imitation of the label generator, not an independent validation of translation quality.

full rationale

The dataset construction pipeline (Section 3) is not circular in itself: it collects real-world images, applies OCR, translates with GPT-4, and filters by spaCy similarity to Google Translate. The core circularity is in the fine-tuning claim (Section 5.4 and Figure 6). The model is trained on GPT-4-produced target translations and then evaluated against test references produced by the same GPT-4 pipeline (Section 3.3); BLEU, chrF++ and METEOR therefore reward reproducing GPT-4's translation style. The human evaluation (10 of 10,000 batches, 99.4% accuracy) is too small and its protocol is undescribed to establish that the references are human-quality, and the Limitations section explicitly concedes that 'the inherent complexity of multilingual data may lead to inaccuracies that could affect the reliability of the dataset.' Separately, the abstract's 230%/88%/130% gains do not match Figure 6's +146%/+95%/+145%, a correctness inconsistency that is not itself a circularity but compounds the difficulty of interpreting the headline result. The self-citations in the reference list (Zhu et al., 2023; Zhu et al., 2024) are related-work background and are not load-bearing for the dataset or the fine-tuning claim. Overall, the central quantitative claim is partially circular because the benchmark's reference distribution is the same as the fine-tuning label distribution; score 6.

Assumptions & free parameters 4 free parameters · 4 assumptions · 0 invented entities

The dataset construction depends on several hand-chosen thresholds (similarity, difficulty, resolution, text length) and on assumptions that OCR and machine translation outputs are accurate enough to serve as ground truth. No new entities are introduced.

free parameters (4)
  • semantic similarity threshold = 0.8 (cosine similarity)
    Translations with spaCy cosine similarity below 0.8 between GPT-4 and Google Translate are discarded (Section 3.3). This hand-chosen threshold controls which samples enter the dataset and can bias it toward easy, agreeable translations.
  • difficulty cutoffs = Easy: boxes <= 2 and tokens <= 16; Hard: boxes > 5 or tokens > 25; Medium: otherwise
    Hand-chosen thresholds define the three difficulty levels used in evaluation and analysis (Section 4.2).
  • minimum image resolution = 800x800 pixels (later stated as 1000x1000 width)
    Images below the resolution cutoff are removed during collection (Section 3.1). The paper is inconsistent about the exact value.
  • maximum text length = 450 tokens or 60 words
    Images with longer recognized text are removed in cleaning (Section 3.2). This hand-set cut affects the distribution of the dataset.
assumptions (4)
  • domain assumption OCR tools (EasyOCR, GPT-4o) accurately transcribe text in product images.
    Section 3.2 relies on OCR output to define source text for translation and evaluation; errors propagate to all translations.
  • domain assumption Semantic similarity between word vectors reflects translation equivalence.
    In Section 3.3, a spaCy cosine similarity above 0.8 is treated as evidence of correct translation, but this is not validated against human judgments.
  • domain assumption GPT-4 and Google Translate agreement indicates correct translation.
    The validation pipeline keeps only translations where two machine systems agree (Section 3.3). Agreement is used as a proxy for gold-standard quality without independent verification.
  • ad hoc to paper Human evaluation on 10 out of 10,000 batches is representative.
    Section 3.3 extrapolates 99.4% accuracy from an unspecified small sample, with no description of batch size or selection method.

how reviews work

0 comments
Cite this review

Pith. "Pith review of MIT-10M: A Large Scale Parallel Corpus of Multilingual Image Translation." pith.science (2026). https://pith.science/paper/QRCHW3XY

@misc{pith2026241207147,
  author       = {Pith},
  title        = {Pith review of: MIT-10M: A Large Scale Parallel Corpus of Multilingual Image Translation},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/QRCHW3XY}},
  note         = {Machine review of arXiv:2412.07147}
}
read the original abstract

Image Translation (IT) holds immense potential across diverse domains, enabling the translation of textual content within images into various languages. However, existing datasets often suffer from limitations in scale, diversity, and quality, hindering the development and evaluation of IT models. To address this issue, we introduce MIT-10M, a large-scale parallel corpus of multilingual image translation with over 10M image-text pairs derived from real-world data, which has undergone extensive data cleaning and multilingual translation validation. It contains 840K images in three sizes, 28 categories, tasks with three levels of difficulty and 14 languages image-text pairs, which is a considerable improvement on existing datasets. We conduct extensive experiments to evaluate and train models on MIT-10M. The experimental results clearly indicate that our dataset has higher adaptability when it comes to evaluating the performance of the models in tackling challenging and complex image translation tasks in the real world. Moreover, the performance of the model fine-tuned with MIT-10M has tripled compared to the baseline model, further confirming its superiority.

Figures

Figures reproduced from arXiv: 2412.07147 by the authors.

Figure 1
Figure 1. Categories and languages of MIT-10M. It in [PITH_FULL_IMAGE:figures/full_fig_p001_1.png] view at source ↗
Figure 2
Figure 2. Examples of MIT-10M dataset. Each image contains the original text and the corresponding language. [PITH_FULL_IMAGE:figures/full_fig_p002_2.png] view at source ↗
Figure 3
Figure 3. Overview of MIT-10M dataset construction pipeline. [PITH_FULL_IMAGE:figures/full_fig_p003_3.png] view at source ↗
Figures from the paper (5 more)
Figure 4
Figure 4. Figure 4: Distribution of the number of bounding boxes [PITH_FULL_IMAGE:figures/full_fig_p005_4.png]
Figure 5
Figure 5. Figure 5: Example of images with different resolutions. [PITH_FULL_IMAGE:figures/full_fig_p006_5.png]
Figure 6
Figure 6. Figure 6: The performance comparison of the 5 model in the metrics: BLEU, chrF++, and METEOR. The 5 models [PITH_FULL_IMAGE:figures/full_fig_p009_6.png]
Figure 7
Figure 7. Figure 7: The detail field description for the MIT-10M dataset. [PITH_FULL_IMAGE:figures/full_fig_p013_7.png]
Figure 8
Figure 8. Figure 8: The heatmap of various models in image translation tasks. [PITH_FULL_IMAGE:figures/full_fig_p014_8.png]

Discussion (0). Continue with ORCID to comment.

Forward citations

Cited by 2 Pith papers

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

  1. Rethinking Multilingual Vision-Language Translation: Dataset, Evaluation, and Adaptation

    cs.CV 2025-06 conditional novelty 6.0 of 10

    The authors introduce AibTrans, a multilingual image-text translation benchmark, show that common translation metrics mislead on dense images, and find that balanced multilingual fine-tuning preserves generalization b...

  2. MT$^{3}$: Scaling MLLM-based Text Image Machine Translation via Multi-Task Reinforcement Learning

    cs.CL 2025-05 conditional novelty 6.0 of 10

    A 7B multimodal model trained with multi-task reinforcement learning beats much larger models on image-text translation benchmarks, though some out-of-distribution claims are contradicted by the paper's own tables.

Reference graph

Works this paper leans on

41 extracted references · 21 canonical work pages · cited by 2 Pith papers

  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]

    Josh Achiam, Steven Adler, Sandhini Agarwal, Lama Ahmad, Ilge Akkaya, Florencia Leoni Aleman, Diogo Almeida, Janko Altenschmidt, Sam Altman, Shyamal Anadkat, et al. 2023. Gpt-4 technical report. arXiv preprint arXiv:2303.08774

    arXiv 2023

  4. [4]

    Haithem Afli and Andy Way. 2016. https://aclanthology.org/W16-4015 Integrating optical character recognition and machine translation of historical documents . In Proceedings of the Workshop on Language Technology Resources and Tools for Digital Humanities ( LT 4 DH ) , pages 109--116, Osaka, Japan. The COLING 2016 Organizing Committee

    work page 2016

  5. [5]

    Jinze Bai, Shuai Bai, Shusheng Yang, Shijie Wang, Sinan Tan, Peng Wang, Junyang Lin, Chang Zhou, and Jingren Zhou. 2023. Qwen-vl: A frontier large vision-language model with versatile abilities. arXiv preprint arXiv:2308.12966

    arXiv 2023

  6. [6]

    Satanjeev Banerjee and Alon Lavie. 2005. https://aclanthology.org/W05-0909 METEOR : An automatic metric for MT evaluation with improved correlation with human judgments . In Proceedings of the ACL Workshop on Intrinsic and Extrinsic Evaluation Measures for Machine Translation and/or Summarization , pages 65--72, Ann Arbor, Michigan. Association for Comput...

    2005

  7. [7]

    Andong Chen, Lianzhang Lou, Kehai Chen, Xuefeng Bai, Yang Xiang, Muyun Yang, Tiejun Zhao, and Min Zhang. 2024 a . https://doi.org/10.18653/v1/2024.acl-short.64 DUAL - REFLECT : Enhancing large language models for reflective translation through dual learning feedback mechanisms . In Proceedings of the 62nd Annual Meeting of the Association for Computationa...

    work page doi:10.18653/v1/2024.acl-short.64 2024

  8. [8]

    Kehai Chen, Masao Utiyama, Eiichiro Sumita, Rui Wang, and Min Zhang. 2022. https://doi.org/10.18653/v1/2022.findings-acl.235 Synchronous refinement for neural machine translation . In Findings of the Association for Computational Linguistics: ACL 2022, pages 2986--2996, Dublin, Ireland. Association for Computational Linguistics

    work page doi:10.18653/v1/2022.findings-acl.235 2022

Show all 41 references
  1. [9]

    Zhe Chen, Jiannan Wu, Wenhai Wang, Weijie Su, Guo Chen, Sen Xing, Muyan Zhong, Qinglong Zhang, Xizhou Zhu, Lewei Lu, et al. 2024 b . Internvl: Scaling up vision foundation models and aligning for generic visual-linguistic tasks. In Proceedings of the IEEE/CVF Conference on Com...

    2024

  2. [10]

    Zhuo Chen, Fei Yin, Xu-Yao Zhang, Qing Yang, and Chena-Lin Liu. 2021. https://doi.org/10.1109/ICPR48806.2021.9412281 Cross-lingual text image recognition via multi-task sequence to sequence learning . In 2020 25th International Conference on Pattern Recognition (ICPR), pages 3...

    2021

  3. [11]

    Team Gemini. 2024. https://arxiv.org/abs/2403.05530 Gemini 1.5: Unlocking multimodal understanding across millions of tokens of context . Preprint, arXiv:2403.05530

    2024 arXiv

  4. [12]

    Ryota Hinami, Shonosuke Ishiwatari, Kazuhiko Yasuda, and Yusuke Matsui. 2021. Towards fully automated manga translation. In Proceedings of the AAAI Conference on Artificial Intelligence, volume 35, pages 12998--13008

    2021

  5. [13]

    Wenyi Hong, Weihan Wang, Ming Ding, Wenmeng Yu, Qingsong Lv, Yan Wang, Yean Cheng, Shiyu Huang, Junhui Ji, Zhao Xue, et al. 2024. Cogvlm2: Visual language models for image and video understanding. arXiv preprint arXiv:2408.16500

    2024 arXiv

  6. [14]

    Puneet Jain, Orhan Firat, Qi Ge, and Sihang Liang. 2021. Image translation network. Github.com

    2021

  7. [15]

    Gant Laborde. 2024. https://github.com/GantMan/nsfw_model Deep nn for nsfw detection

    2024

  8. [16]

    Zhibin Lan, Liqiang Niu, Fandong Meng, Jie Zhou, Min Zhang, and Jinsong Su. 2024. https://aclanthology.org/2024.findings-acl.325 Translatotron- V (ison): An end-to-end model for in-image machine translation . In Findings of the Association for Computational Linguistics ACL 202...

    2024

  9. [17]

    Zhibin Lan, Jiawei Yu, Xiang Li, Wen Zhang, Jian Luan, Bin Wang, Degen Huang, and Jinsong Su. 2023. Exploring better text image translation with multimodal codebook. arXiv preprint arXiv:2305.17415

    2023 arXiv

  10. [18]

    Richardson Leonard. 2004. https://www.crummy.com/software/BeautifulSoup/ Beautiful soup: a library designed for screen-scraping html and xml. crummy.com

    2004

  11. [19]

    Junnan Li, Dongxu Li, Silvio Savarese, and Steven Hoi. 2023. Blip-2: Bootstrapping language-image pre-training with frozen image encoders and large language models. In International conference on machine learning, pages 19730--19742. PMLR

    2023

  12. [20]

    Yupu Liang, Yaping Zhang, Cong Ma, Zhiyang Zhang, Yang Zhao, Lu Xiang, Chengqing Zong, and Yu Zhou. 2024. https://doi.org/10.18653/v1/2024.naacl-long.392 Document image machine translation with dynamic multi-pre-trained models assembling . In Proceedings of the 2024 Conference...

    2024 doi

  13. [21]

    Haotian Liu, Chunyuan Li, Yuheng Li, Bo Li, Yuanhan Zhang, Sheng Shen, and Yong Jae Lee. 2024. https://llava-vl.github.io/blog/2024-01-30-llava-next/ Llava-next: Improved reasoning, ocr, and world knowledge

    2024

  14. [22]

    Haoyu Lu, Wen Liu, Bo Zhang, Bingxuan Wang, Kai Dong, Bo Liu, Jingxiang Sun, Tongzheng Ren, Zhuoshu Li, Yaofeng Sun, et al. 2024. Deepseek-vl: towards real-world vision-language understanding. arXiv preprint arXiv:2403.05525

    2024 arXiv

  15. [23]

    Cong Ma, Yaping Zhang, Mei Tu, Xu Han, Linghui Wu, Yang Zhao, and Yu Zhou. 2022. Improving end-to-end text image translation from the auxiliary text translation task. 2022 26th International Conference on Pattern Recognition (ICPR), pages 1664--1670

    2022

  16. [24]

    Cong Ma, Yaping Zhang, Mei Tu, Yang Zhao, Yu Zhou, and Chengqing Zong. 2023 a . E2timt: Efficient and effective modal adapter for text image machine translation. In International Conference on Document Analysis and Recognition, pages 70--88. Springer

    2023

  17. [25]

    Cong Ma, Yaping Zhang, Mei Tu, Yang Zhao, Yu Zhou, and Chengqing Zong. 2023 b . Multi-teacher knowledge distillation for end-to-end text image machine translation. In International Conference on Document Analysis and Recognition, pages 484--501. Springer

    2023

  18. [26]

    Elman Mansimov, Mitchell Stern, Mia Chen, Orhan Firat, Jakob Uszkoreit, and Puneet Jain. 2020. https://doi.org/10.18653/v1/2020.nlpbt-1.8 Towards end-to-end in-image neural machine translation . In Proceedings of the First International Workshop on Natural Language Processing ...

    2020 doi

  19. [27]

    Liqiang Niu, Fandong Meng, and Jie Zhou. 2024. https://aclanthology.org/2024.lrec-main.1474 UMTIT : Unifying recognition, translation, and generation for multimodal text image translation . In Proceedings of the 2024 Joint International Conference on Computational Linguistics,...

    2024

  20. [28]

    NLLB Team , Marta R. Costa-jussà, James Cross, Onur Çelebi, Maha Elbayad, Kenneth Heafield, Kevin Heffernan, Elahe Kalbassi, Janice Lam, Daniel Licht, Jean Maillard, Anna Sun, Skyler Wang, Guillaume Wenzek, Al Youngblood, Bapi Akula, Loic Barrault, Gabriel Mejia-Gonzalez, Pran...

    2022

  21. [29]

    OpenAI. 2024. https://openai.com/index/hellogpt-4o/ Hello gpt-4o . openai.com

    2024

  22. [30]

    Kishore Papineni, Salim Roukos, Todd Ward, and Wei-Jing Zhu. 2002. https://doi.org/10.3115/1073083.1073135 Bleu: a method for automatic evaluation of machine translation . In Proceedings of the 40th Annual Meeting on Association for Computational Linguistics, ACL '02, page 311...

    2002

  23. [31]

    Maja Popovi \'c . 2017. https://doi.org/10.18653/v1/W17-4770 chr F ++: words helping character n-grams . In Proceedings of the Second Conference on Machine Translation, pages 612--618, Copenhagen, Denmark. Association for Computational Linguistics

    2017 doi

  24. [32]

    K Chandra Shekar, Maria Anisha Cross, and Vignesh Vasudevan. 2021. Optical character recognition and neural machine translation using deep learning techniques. In Innovations in Computer Science and Engineering: Proceedings of 8th ICICSE, pages 277--283. Springer

    2021

  25. [33]

    Tonghua Su, Shuchen Liu, and Shengjie Zhou. 2021. Rtnet: An end-to-end method for handwritten text image translation. In Document Analysis and Recognition--ICDAR 2021: 16th International Conference, Lausanne, Switzerland, September 5--10, 2021, Proceedings, Part II 16, pages 9...

    2021

  26. [34]

    Yasuhiko Watanabe, Yoshihiro Okada, Yeun-Bae Kim, and Tetsuya Takeda. 1998. Translation camera. In Proceedings. Fourteenth International Conference on Pattern Recognition (Cat. No. 98EX170), volume 1, pages 613--617. IEEE

    1998

  27. [35]

    Jie Yang, Xilin Chen, Jing Zhang, Ying Zhang, and Alex Waibel. 2002. Automatic detection and translation of text from natural scenes. In 2002 IEEE International conference on acoustics, speech, and signal processing, volume 2, pages II--2101. IEEE

    2002

  28. [36]

    Yuan Yao, Tianyu Yu, Ao Zhang, Chongyi Wang, Junbo Cui, Hongji Zhu, Tianchi Cai, Haoyu Li, Weilin Zhao, Zhihui He, et al. 2024. Minicpm-v: A gpt-4v level mllm on your phone. arXiv preprint arXiv:2408.01800

    2024 arXiv

  29. [37]

    Yaping Zhang, Shuai Nie, Wenju Liu, Xing Xu, Dongxiang Zhang, and Heng Tao Shen. 2019. Sequence-to-sequence domain adaptation network for robust text image recognition. In Proceedings of the IEEE/CVF conference on computer vision and pattern recognition, pages 2740--2749

    2019

  30. [38]

    Yang Zhao, Lu Xiang, Junnan Zhu, Jiajun Zhang, Yu Zhou, and Chengqing Zong. 2020. https://doi.org/10.18653/v1/2020.coling-main.397 Knowledge graph enhanced neural machine translation via multi-task learning on sub-entity granularity . In Proceedings of the 28th International C...

    2020 doi

  31. [39]

    Meizhi Zhong, Lemao Liu, Kehai Chen, Mingming Yang, and Min Zhang. 2024. https://doi.org/10.18653/v1/2024.acl-long.727 Context consistency between training and inference in simultaneous machine translation . In Proceedings of the 62nd Annual Meeting of the Association for Comp...

    2024 doi

  32. [40]

    Shaolin Zhu, Shangjie Li, Yikun Lei, and Deyi Xiong. 2023. https://doi.org/10.18653/v1/2023.acl-long.751 PEIT : Bridging the modality gap with pre-trained models for end-to-end image translation . In Proceedings of the 61st Annual Meeting of the Association for Computational L...

    2023 doi

  33. [41]

    Shaolin Zhu, Leiyu Pan, Bo Li, and Deyi Xiong. 2024. https://doi.org/10.18653/v1/2024.acl-long.656 LAND e RMT : Dectecting and routing language-aware neurons for selectively finetuning LLM s to machine translation . In Proceedings of the 62nd Annual Meeting of the Association ...

    2024 doi

Pith tools

Reviewed August 11, 2026 · model on record in the stance chip above.