Pith. sign in

REVIEW 4 major objections 4 minor 52 references

PRIM: Towards Practical In-Image Multilingual Machine Translation

T0 review · 4 major / 4 minor · reviewed 2026-08-05 · deepseek-v4-flash

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

Pith's one-line read The paper introduces PRIM, a real-world multilingual in-image translation benchmark, and VisTrans, an end-to-end model that separates text and background to translate photos without damaging the scene.

desk verdict PRIM is a genuinely useful new real-world IIMT benchmark and VisTrans has credible visual-quality gains, but the abstract overclaims translation quality and the missing DebackX baseline must be addressed before the comparative claims are verifiable. read the letter →

arxiv 2509.05146 v1 pith:4NSB36X4 submitted 2025-09-05 cs.CL cs.CV

Yanzhi Tian , Zeming Liu , Zhengyang Liu , Chong Feng , Xin Li , Heyan Huang , Yuhang Guo This is my paper · ORCID
classification cs.CLcs.CV
keywords in-imagemachinetranslationmultilingualimage-to-imagereal-worldbenchmarkvisualtextgenerationcodebookquantizationtwo-stagetrainingimage
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 argues that in-image machine translation can move from simplified synthetic setups to practical, real-world conditions. To make that case, it introduces PRIM, a new benchmark of real-world photographed one-line text images with complex backgrounds, varied fonts, diverse text positions, and five English-to-X translation directions. It also proposes VisTrans, an end-to-end model that processes visual text and background information separately, then combines them when generating the translated image. The authors report that VisTrans outperforms earlier end-to-end and most cascade baselines on both translation quality and visual fidelity, while a strong pretrained cascade still leads on translation quality.

What carries the argument

The central mechanism is VisTrans's separation of visual text and background into two representation streams. The text stream runs the source image through an MT Encoder, generates subword target text with a 1-Pass Decoder using a language tag, converts it to character-level representations via a Subword-to-Char (S2C) decoder, and predicts codebook codes with a 2-Pass Decoder; the background stream produces a background representation that is added to the text-code representation before image decoding. The codebook stores visual text patches, and the S2C decoder bridges the granularity gap between subwords and image codes. Two-stage training with auxiliary tasks—background reconstruction, te

What would settle it

Train or fine-tune VisTrans on PRIM's real source images paired with their annotated target images, then compare BLEU and FID against the synthetic-trained model; if the real-trained model does not improve, or if synthetic-trained performance collapses on held-out fonts, backgrounds, and text lengths outside the training renderer, the transferability premise is falsified. A second check: measure background preservation directly by computing FID on background-only regions to confirm that adding Hback preserves the scene rather than regenerating it.

Watch

Extended reading notes

Core claim

VisTrans is an end-to-end model that translates text inside a photo while preserving the surrounding scene. It encodes the source image through separate vision-transformer streams: one captures the visual appearance of the text, the other captures the background. A codebook quantizes target text-image patches, and the model generates the target image by autoregressively predicting codebook codes from the translated text, then adding the preserved background. The model is trained in two stages, first teaching these branches to reconstruct text-images and backgrounds, then teaching the translation-to-code path. On PRIM, VisTrans reports an average BLEU of 11.3 and FID of 28.8, outperforming th

Load-bearing premise

The load-bearing premise is that a model trained on synthetically rendered text images—Arial target fonts on video-frame backgrounds—will transfer to the real-world photographs in PRIM; if that simulation-to-real gap is large, the reported quality and multilingual gains will not survive deployment.

Editorial extensions

If this is right

  • If VisTrans's design holds, end-to-end in-image translation no longer requires the OCR-translate-render cascade, avoiding the background damage and incomplete-text-rendering failure modes the paper documents.
  • Multilingual support is achieved without separate models per language pair: a language tag steers the same decoder across five translation directions, suggesting easy extension to more languages.
  • Because the codebook stores visual text patches, the model can re-render translated text at a fitting font size, avoiding the truncation that limits cascade models on long text.
  • Reported results on the IIMT30k dataset indicate the architecture transfers to multi-line and more complex text layouts when trained appropriately, not just PRIM's one-line images.
  • PRIM provides a reusable real-world benchmark with two reference translations per direction, making future IIMT systems comparable on both translation quality and visual fidelity.

Reading between the lines

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

  • The two-stage recipe—translate to text first, then render text into the image—could transfer beyond one-line IIMT to document image translation, signboard translation, or any text-bearing image where background preservation matters.
  • If the synthetic-to-real transfer is confirmed, the training pipeline can be scaled cheaply to many more languages and scripts by rendering parallel text from existing corpora, without new human annotation.
  • The S2C decoder ablation suggests subword-to-patch granularity mismatch is a central bottleneck, so finer intermediate representations such as characters, strokes, or adaptive group sizes are a natural next lever for quality.
  • PRIM's two-reference design (GPT-4 and Google Translate) lets future work separate translation accuracy from rendering fidelity, for example by scoring OCR round-trip agreement between the two references.
Share X Bluesky LinkedIn Reddit HN

Editorial analysis

A structured set of objections, weighed in public.

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

Referee Report

4 major / 4 minor

Summary. The paper introduces PRIM, a benchmark for practical in-image multilingual machine translation (IIMMT) consisting of real-world captured source images with one-line text, five translation directions (En-De, En-Fr, En-Cs, En-Ru, En-Ro), and 340 images per direction with two reference target images per source image. The authors also propose VisTrans, an end-to-end model that separates background information from visual-text information using a two-stage training procedure, a codebook, and a subword-to-character decoder. Experiments on PRIM compare VisTrans with pre-trained and non-pre-trained cascade systems, with prior end-to-end IIMT models, and with the golden reference. The paper claims that VisTrans achieves better translation quality and visual effect than other models. The visual-quality claim is substantially supported by the reported FID (VisTrans 28.8 vs. 69.1 for TranslatotronV and >100 for cascades), but the translation-quality claim is not: Table 1 shows VisTrans average BLEU 11.3 vs. 23.0 for EasyOCR-NLLB-Render and average COMET 47.0 vs. 48.0 for PEIT-Render.

Significance. If the results hold, the paper makes two valuable contributions. PRIM is, to my knowledge, the first publicly available multilingual IIMT benchmark using real-world source images, and the authors provide code and data. VisTrans is a serious end-to-end attempt at practical IIMT, and its large FID improvement over cascades and TranslatotronV is a meaningful result, as is its demonstrated ability to handle multi-line synthetic IIMT30k data. The paper also provides a large synthetic training set with multilingual directions. However, the headline claim of better translation quality is not supported by the paper's own table, and the absence of the recent DebackX baseline from the main PRIM comparison, together with the lack of significance testing, leaves the comparative claim under-supported. The synthetic-to-real transfer of the training procedure is also not validated. The central empirical claims are partially supported and the issues are fixable, but the paper needs revision before the claims can be accepted as stated.

major comments (4)
  1. [Abstract; §7 (Conclusion)] The claim that VisTrans achieves "better translation quality ... compared to other models" is contradicted by Table 1. VisTrans has average BLEU 11.3 vs. 23.0 for EasyOCR-NLLB-Render and average COMET 47.0 vs. 48.0 for PEIT-Render and 62.7 for EasyOCR-NLLB-Render. VisTrans is the best end-to-end model in the table, but it is not better in translation quality than all models. Please revise the abstract and conclusion to state the qualified finding, e.g., "best visual quality and competitive translation quality among end-to-end models," or report the full comparison without the blanket superiority claim.
  2. [Table 1 / §5.3] DebackX is identified as a recent end-to-end IIMT model in Figure 1 and is compared against VisTrans on IIMT30k in Table 4, but it is absent from the PRIM evaluation in Table 1. Since the paper's central comparative claim is that VisTrans is better than other models, omitting the strongest recent end-to-end IIMT baseline from the main benchmark makes that claim unverifiable. Please add DebackX to the PRIM evaluation, or clearly justify its exclusion and relax the comparative claim accordingly.
  3. [Table 1 / §5.1] No confidence intervals, standard deviations, or significance tests are reported for any metric. With only 340 images per direction, the BLEU difference between VisTrans (11.3) and PEIT-Render (10.4) is about 0.9, and the COMET difference is -1.0 (47.0 vs. 48.0); these differences are plausibly within noise. Report bootstrap confidence intervals, paired significance tests, or per-direction variance to support the relative ranking of systems.
  4. [§3 Training Set] The training set is entirely synthetic: source text is rendered with TRDG, target text uses Arial, and backgrounds are extracted from video frames. PRIM, by contrast, contains real-world captured advertising and subtitle images. No experiment measures this simulation-to-real transfer: there is no training-on-real-data comparison, no per-domain breakdown, and no analysis of how the domain gap affects BLEU/COMET/FID. The "practical" claim therefore rests on an unvalidated transfer assumption. The Limitations section acknowledges computational cost and codebook simplicity but not this risk, which should be addressed empirically or explicitly listed as a limitation.
minor comments (4)
  1. [Abstract; §3; §7] Several language errors should be fixed: "convince the lack" should be "compensate for the lack" or "address the lack"; in §7 "retrains" should be "retains"; and "5translation" is missing a space.
  2. [Table 1] The table formatting is poor in places, e.g., "10.4 /45.114.0" and "12.8 /53.59.9 / 46.8". Please ensure values and slashes are clearly separated.
  3. [Figure 1] The caption uses "ML." and "RB." without defining them. Spell out "multilingual" and "real-world background" in the caption or legend.
  4. [§6.1] The "Speedup" column in Table 3 is not defined precisely. State what is measured (e.g., inference-only time, batch size, hardware) and whether the "None" row means removing the S2C decoder entirely and generating codes directly from the 1-Pass representation.

Circularity Check

0 steps flagged · score 0.0 of 10

No circular derivation: VisTrans is an empirical system evaluated on a held-out reference benchmark, and no prediction reduces to a fitted input or self-citation by construction.

full rationale

The paper's derivation chain is not circular. VisTrans is trained with explicit two-stage losses: image reconstruction and VQ losses (Eqs. 1-6) in stage 1, then cross-entropy losses for char-level and code-sequence generation (Eq. 7) in stage 2. These losses are standard supervised objectives on synthetic training images derived from MTed and rendered backgrounds; PRIM is a separately constructed, held-out real-world benchmark. The target-code sequence is generated from text representations and looked up in a codebook trained on target text-images, but this is an architectural design, not an equation that reduces to its own input. Evaluation uses EasyOCR on generated images with BLEU/COMET against reference texts and FID against reference images; the model is not fitted to these metrics or to EasyOCR outputs. The only self-citations (Tian et al. 2023, 2025) appear in related work or as the IIMT30k robustness benchmark; they are not load-bearing for the central PRIM claim and do not invoke a uniqueness theorem or forbid alternatives. The abstract's claim of 'better translation quality' is overstated because Table 1 shows EasyOCR-NLLB-Render achieves higher BLEU/COMET (23.0/62.7 vs 11.3/47.0), and no significance tests are reported; however, this is an empirical-validity/statistical concern, not circularity. The Limitations section honestly notes computational cost and lack of advanced codebook/decoder exploration, which further does not create a circular step. The core derivation is self-contained against external benchmarks, so the circularity score is 0.

Assumptions & free parameters 4 free parameters · 5 assumptions · 0 invented entities

The paper is an empirical systems paper, so the ledger mostly records domain assumptions about data representativeness and evaluation validity rather than mathematical axioms. The main unverified premise is that synthetic training data transfers to real-world PRIM images; the evaluation then assumes EasyOCR and FID are acceptable proxies for translation and visual quality.

free parameters (4)
  • S2C decoder group size K = 2
    Chosen based on validation set to balance BLEU and speed; ablation (Table 3) shows K affects translation quality, so the central result depends on this manual choice.
  • perceptual loss weight lambda_p = 0.1
    Set manually in Equations 1 and 2; not ablated in the paper.
  • label smoothing coefficient = 0.1
    Applied to all cross-entropy losses; chosen manually.
  • codebook size and dimension = 8192 and 32
    Architecture hyperparameters set by hand; the codebook is central to the visual text representation.
assumptions (5)
  • domain assumption Source images in PRIM, taken from Ma et al. (2024) and Li et al. (2025a), are representative of real-world one-line text images.
    The dataset is built by cropping and filtering these existing TIT images; if they are not representative of practical conditions, the PRIM benchmark would not measure what the paper claims.
  • domain assumption Synthetic training images rendered from MTed texts on extracted video-frame backgrounds approximate real-world test images well enough for transfer.
    The model is trained only on synthetic data; no real-source training images are used. See Section 3, Training Set.
  • domain assumption GPT-4-turbo and Google Translate provide reference translations of sufficient quality.
    Target texts are machine-generated; quality is checked only by reference-free COMET, not by human evaluation.
  • domain assumption EasyOCR correctly recognizes the generated multilingual images often enough that BLEU and COMET reflect translation quality.
    Section 5.1 uses EasyOCR for all systems; OCR errors are acknowledged with a Golden reference but no human evaluation is reported.
  • domain assumption FID is a valid proxy for visual quality on 32x512 images.
    FID typically uses Inception features designed for 299x299 inputs; the paper does not describe resizing or standardization, so the FID numbers may be sensitive to implementation.

how reviews work

0 comments
Cite this review

Pith. "Pith review of PRIM: Towards Practical In-Image Multilingual Machine Translation." pith.science (2026). https://pith.science/paper/4NSB36X4

@misc{pith2026250905146,
  author       = {Pith},
  title        = {Pith review of: PRIM: Towards Practical In-Image Multilingual Machine Translation},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/4NSB36X4}},
  note         = {Machine review of arXiv:2509.05146}
}
read the original abstract

In-Image Machine Translation (IIMT) aims to translate images containing texts from one language to another. Current research of end-to-end IIMT mainly conducts on synthetic data, with simple background, single font, fixed text position, and bilingual translation, which can not fully reflect real world, causing a significant gap between the research and practical conditions. To facilitate research of IIMT in real-world scenarios, we explore Practical In-Image Multilingual Machine Translation (IIMMT). In order to convince the lack of publicly available data, we annotate the PRIM dataset, which contains real-world captured one-line text images with complex background, various fonts, diverse text positions, and supports multilingual translation directions. We propose an end-to-end model VisTrans to handle the challenge of practical conditions in PRIM, which processes visual text and background information in the image separately, ensuring the capability of multilingual translation while improving the visual quality. Experimental results indicate the VisTrans achieves a better translation quality and visual effect compared to other models. The code and dataset are available at: https://github.com/BITHLP/PRIM.

Figures

Figures reproduced from arXiv: 2509.05146 by the authors.

Figure 1
Figure 1. Comparison between PRIM to publicly avail [PITH_FULL_IMAGE:figures/full_fig_p001_1.png] view at source ↗
Figure 2
Figure 2. Annotation procedure of PRIM. The first step is inpainting the text region of the source image manually, [PITH_FULL_IMAGE:figures/full_fig_p003_2.png] view at source ↗
Figure 3
Figure 3. Architecture of VisTrans. The trapezoid represents Vision Transformer (ViT) ( [PITH_FULL_IMAGE:figures/full_fig_p004_3.png] view at source ↗
Figures from the paper (5 more)
Figure 4
Figure 4. Figure 4: The two stages training of VisTrans. The modules and arrows with dashed lines represent auxiliary [PITH_FULL_IMAGE:figures/full_fig_p005_4.png]
Figure 5
Figure 5. Figure 5: Samples from PRIM, which includes source images, source texts, multilingual target images, and [PITH_FULL_IMAGE:figures/full_fig_p014_5.png]
Figure 6
Figure 6. Figure 6: Samples from our training set. Due to the auxiliary training tasks in the training process, the training [PITH_FULL_IMAGE:figures/full_fig_p014_6.png]
Figure 7
Figure 7. Figure 7: Multilingual output images of VisTrans on PRIM. Our model maintains the integrity of the image [PITH_FULL_IMAGE:figures/full_fig_p015_7.png]
Figure 8
Figure 8. Figure 8: Outputs from AnyTrans (Qian et al., 2024). The prompt is the translation result of recognized text by PPOCR, based on Qwen-1.5 7B. We find that AnyTrans is not well-suited for our test set, which is limited by the text editing capability of lengthy text in the image […

Discussion (0). Continue with ORCID to comment.

Reference graph

Works this paper leans on

52 extracted references · 33 canonical work pages

  1. [1]

    Jinze Bai, Shuai Bai, Yunfei Chu, Zeyu Cui, Kai Dang, Xiaodong Deng, Yang Fan, Wenbin Ge, Yu Han, Fei Huang, Binyuan Hui, Luo Ji, Mei Li, Junyang Lin, Runji Lin, Dayiheng Liu, Gao Liu, Chengqiang Lu, Keming Lu, Jianxin Ma, Rui Men, Xingzhang Ren, Xuancheng Ren, Chuanqi Tan, Sinan Tan, Jianhong Tu, Peng Wang, Shijie Wang, Wei Wang, Shengguang Wu, Benfeng X...

    arXiv 2023

  2. [2]

    Shuai Bai, Keqin Chen, Xuejing Liu, Jialin Wang, Wenbin Ge, Sibo Song, Kai Dang, Peng Wang, Shijie Wang, Jun Tang, Humen Zhong, Yuanzhi Zhu, Mingkun Yang, Zhaohai Li, Jianqiang Wan, Pengfei Wang, Wei Ding, Zheren Fu, Yiheng Xu, Jiabo Ye, Xi Zhang, Tianbao Xie, Zesen Cheng, Hang Zhang, Zhibo Yang, Haiyang Xu, and Junyang Lin. 2025. https://arxiv.org/abs/25...

    arXiv 2025

  3. [3]

    Darwin Bautista and Rowel Atienza. 2022. https://doi.org/10.1007/978-3-031-19815-1_11 Scene text recognition with permuted autoregressive sequence models . In European Conference on Computer Vision, pages 178--196, Cham. Springer Nature Switzerland

    doi:10.1007/978-3-031-19815-1_11 2022

  4. [4]

    Andong Chen, Yuchen Song, Kehai Chen, Muyun Yang, Tiejun Zhao, and Min Zhang. 2025. https://arxiv.org/abs/2412.12627 Make imagination clearer! stable diffusion-based visual imagination for multimodal machine translation . Preprint, arXiv:2412.12627

    work page Pith review arXiv 2025

  5. [5]

    Alexey Dosovitskiy, Lucas Beyer, Alexander Kolesnikov, Dirk Weissenborn, Xiaohua Zhai, Thomas Unterthiner, Mostafa Dehghani, Matthias Minderer, Georg Heigold, Sylvain Gelly, Jakob Uszkoreit, and Neil Houlsby. 2021. https://openreview.net/forum?id=YicbFdNTTy An image is worth 16x16 words: Transformers for image recognition at scale . In International Confe...

    work page 2021

  6. [6]

    Kevin Duh. 2018. The multitarget ted talks task. http://www.cs.jhu.edu/ kevinduh/a/multitarget-tedtalks/

    work page 2018

  7. [7]

    Patrick Esser, Robin Rombach, and Bjorn Ommer. 2021. Taming transformers for high-resolution image synthesis. In Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR), pages 12873--12883

    work page 2021

  8. [8]

    Qingkai Fang and Yang Feng. 2022. https://doi.org/10.18653/v1/2022.acl-long.390 Neural machine translation with phrase-level universal visual representations . In Proceedings of the 60th Annual Meeting of the Association for Computational Linguistics (Volume 1: Long Papers), pages 5687--5698, Dublin, Ireland. Association for Computational Linguistics

    work page doi:10.18653/v1/2022.acl-long.390 2022

Show all 52 references
  1. [9]

    Qingkai Fang, Shaolei Zhang, Zhengrui Ma, Min Zhang, and Yang Feng. 2024. https://doi.org/10.18653/v1/2024.acl-long.392 Can we achieve high-quality direct speech-to-speech translation without parallel speech data? In Proceedings of the 62nd Annual Meeting of the Association fo...

    2024 doi

  2. [10]

    Qingkai Fang, Yan Zhou, and Yang Feng. 2023. https://proceedings.neurips.cc/paper_files/paper/2023/file/e5b1c0d4866f72393c522c8a00eed4eb-Paper-Conference.pdf Daspeech: Directed acyclic transformer for fast and high-quality speech-to-speech translation . In Advances in Neural I...

    2023

  3. [11]

    Alex Graves, Santiago Fern\' a ndez, Faustino Gomez, and J\" u rgen Schmidhuber. 2006. https://doi.org/10.1145/1143844.1143891 Connectionist temporal classification: labelling unsegmented sequence data with recurrent neural networks . In Proceedings of the 23rd International C...

    2006

  4. [12]

    Sylvain Gugger, Lysandre Debut, Thomas Wolf, Philipp Schmid, Zachary Mueller, Sourab Mangrulkar, Marc Sun, and Benjamin Bossan. 2022. Accelerate: Training and inference at scale made simple, efficient and adaptable. https://github.com/huggingface/accelerate

    2022

  5. [13]

    Hirofumi Inaguma, Sravya Popuri, Ilia Kulikov, Peng-Jen Chen, Changhan Wang, Yu-An Chung, Yun Tang, Ann Lee, Shinji Watanabe, and Juan Pino. 2023. https://doi.org/10.18653/v1/2023.acl-long.872 U nit Y : Two-pass direct speech-to-speech translation with discrete units . In Proc...

    2023 doi

  6. [14]

    Max Jaderberg, Karen Simonyan, Andrea Vedaldi, and Andrew Zisserman. 2014. https://arxiv.org/abs/1406.2227 Synthetic data and artificial neural networks for natural scene text recognition . Preprint, arXiv:1406.2227

    2014 arXiv

  7. [15]

    Ye Jia, Michelle Tadmor Ramanovich, Tal Remez, and Roi Pomerantz. 2022 a . https://proceedings.mlr.press/v162/jia22b.html Translatotron 2: High-quality direct speech-to-speech translation with voice preservation . In Proceedings of the 39th International Conference on Machine ...

    2022

  8. [16]

    Ye Jia, Michelle Tadmor Ramanovich, Quan Wang, and Heiga Zen. 2022 b . https://aclanthology.org/2022.lrec-1.720/ CVSS corpus and massively multilingual speech-to-speech translation . In Proceedings of the Thirteenth Language Resources and Evaluation Conference, pages 6691--670...

    2022

  9. [17]

    Melvin Johnson, Mike Schuster, Quoc V Le, Maxim Krikun, Yonghui Wu, Zhifeng Chen, Nikhil Thorat, Fernanda Vi \'e gas, Martin Wattenberg, Greg Corrado, et al. 2017. Google’s multilingual neural machine translation system: Enabling zero-shot translation. Transactions of the Asso...

    2017

  10. [18]

    Zhibin Lan, Liqiang Niu, Fandong Meng, Jie Zhou, Min Zhang, and Jinsong Su. 2024. https://doi.org/10.18653/v1/2024.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...

    2024 doi

  11. [19]

    Zhibin Lan, Jiawei Yu, Xiang Li, Wen Zhang, Jian Luan, Bin Wang, Degen Huang, and Jinsong Su. 2023. https://doi.org/10.18653/v1/2023.acl-long.192 Exploring better text image translation with multimodal codebook . In Proceedings of the 61st Annual Meeting of the Association for...

    2023 doi

  12. [20]

    Bo Li, Shaolin Zhu, and Lijie Wen. 2025 a . https://aclanthology.org/2025.coling-main.346/ MIT -10 M : A large scale parallel corpus of multilingual image translation . In Proceedings of the 31st International Conference on Computational Linguistics, pages 5154--5167, Abu Dhab...

    2025

  13. [21]

    Silin Li, Yuhang Guo, Jiashu Yao, Zeming Liu, and Haifeng Wang. 2025 b . https://doi.org/10.18653/v1/2025.acl-long.597 homebench : Evaluating LLM s in smart homes with valid and invalid instructions across single and multiple devices . In Proceedings of the 63rd Annual Meeting...

    2025 doi

  14. [22]

    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

  15. [23]

    Yupu Liang, Yaping Zhang, Zhiyang Zhang, Zhiyuan Chen, Yang Zhao, Lu Xiang, Chengqing Zong, and Yu Zhou. 2025 a . https://doi.org/10.18653/v1/2025.findings-acl.1213 Improving MLLM ' s document image machine translation via synchronously self-reviewing its OCR proficiency . In ...

    2025 doi

  16. [24]

    Yupu Liang, Yaping Zhang, Zhiyang Zhang, Yang Zhao, Lu Xiang, Chengqing Zong, and Yu Zhou. 2025 b . https://doi.org/10.18653/v1/2025.acl-long.606 Single-to-mix modality alignment with multimodal large language model for document image machine translation . In Proceedings of th...

    2025 doi

  17. [25]

    Ilya Loshchilov and Frank Hutter. 2019. https://openreview.net/forum?id=Bkg6RiCqY7 Decoupled weight decay regularization . In International Conference on Learning Representations

    2019

  18. [26]

    Yuheng Lu, Qian Yu, Hongru Wang, Zeming Liu, Wei Su, Yanping Liu, Yuhang Guo, Maocheng Liang, Yunhong Wang, and Haifeng Wang. 2025. https://doi.org/10.18653/v1/2025.findings-acl.645 T rans B ench: Breaking barriers for transferable graphical user interface agents in dynamic di...

    2025 doi

  19. [27]

    Cong Ma, Xu Han, Linghui Wu, Yaping Zhang, Yang Zhao, Yu Zhou, and Chengqing Zong. 2024. https://doi.org/10.1109/TASLP.2023.3324540 Modal contrastive learning based end-to-end text image machine translation . IEEE/ACM Transactions on Audio, Speech, and Language Processing, 32:...

    2024

  20. [28]

    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

  21. [29]

    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

  22. [30]

    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 arXiv

  23. [31]

    Kishore Papineni, Salim Roukos, Todd Ward, and Wei-Jing Zhu. 2002. https://doi.org/10.3115/1073083.1073135 B leu: a method for automatic evaluation of machine translation . In Proceedings of the 40th Annual Meeting of the Association for Computational Linguistics, pages 311--3...

    2002

  24. [32]

    William Peebles and Saining Xie. 2023. Scalable diffusion models with transformers. In Proceedings of the IEEE/CVF International Conference on Computer Vision (ICCV), pages 4195--4205

    2023

  25. [33]

    Wong, Xiaoshuai Sun, and Rongrong Ji

    Zhipeng Qian, Pei Zhang, Baosong Yang, Kai Fan, Yiwei Ma, Derek F. Wong, Xiaoshuai Sun, and Rongrong Ji. 2024. https://doi.org/10.18653/v1/2024.findings-emnlp.137 A ny T rans: Translate A ny T ext in the image with large scale models . In Findings of the Association for Comput...

    2024 doi

  26. [34]

    Ricardo Rei, Craig Stewart, Ana C Farinha, and Alon Lavie. 2020. https://doi.org/10.18653/v1/2020.emnlp-main.213 COMET : A neural framework for MT evaluation . In Proceedings of the 2020 Conference on Empirical Methods in Natural Language Processing (EMNLP), pages 2685--2702, ...

    2020 doi

  27. [35]

    Robin Rombach, Andreas Blattmann, Dominik Lorenz, Patrick Esser, and Bj\"orn Ommer. 2022. High-resolution image synthesis with latent diffusion models. In Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR), pages 10684--10695

    2022

  28. [36]

    Yanzhi Tian, Xiang Li, Zeming Liu, Yuhang Guo, and Bin Wang. 2023. https://doi.org/10.18653/v1/2023.findings-emnlp.1004 In-image neural machine translation with segmented pixel sequence-to-sequence model . In Findings of the Association for Computational Linguistics: EMNLP 202...

    2023 doi

  29. [37]

    Yanzhi Tian, Zeming Liu, Zhengyang Liu, and Yuhang Guo. 2025. https://aclanthology.org/2025.findings-acl.6/ Exploring in-image machine translation with real-world background . In Findings of the Association for Computational Linguistics: ACL 2025, pages 124--137, Vienna, Austr...

    2025

  30. [38]

    Yuxiang Tuo, Wangmeng Xiang, Jun-Yan He, Yifeng Geng, and Xuansong Xie. 2024. https://openreview.net/forum?id=ezBH9WE9s2 Anytext: Multilingual visual text generation and editing . In The Twelfth International Conference on Learning Representations

    2024

  31. [39]

    Ashish Vaswani, Noam Shazeer, Niki Parmar, Jakob Uszkoreit, Llion Jones, Aidan N Gomez, ukasz Kaiser, and Illia Polosukhin. 2017. https://proceedings.neurips.cc/paper_files/paper/2017/file/3f5ee243547dee91fbd053c1c4a845aa-Paper.pdf Attention is all you need . In Advances in Ne...

    2017

  32. [40]

    Chunqi Wang, Ji Zhang, and Haiqing Chen. 2018. https://doi.org/10.18653/v1/D18-1044 Semi-autoregressive neural machine translation . In Proceedings of the 2018 Conference on Empirical Methods in Natural Language Processing, pages 479--488, Brussels, Belgium. Association for Co...

    2018 doi

  33. [41]

    Jiashu Yao, Heyan Huang, Zeming Liu, and Yuhang Guo. 2024. https://doi.org/10.18653/v1/2024.findings-acl.481 Deterministic reversible data augmentation for neural machine translation . In Findings of the Association for Computational Linguistics: ACL 2024, pages 8075--8089, Ba...

    2024 doi

  34. [42]

    Donglei Yu, Yang Zhao, Jie Zhu, Yangyifan Xu, Yu Zhou, and Chengqing Zong. 2025. https://openreview.net/forum?id=XBF63bHDZw Simul PL : Aligning human preferences in simultaneous machine translation . In The Thirteenth International Conference on Learning Representations

    2025

  35. [43]

    Jiahui Yu, Xin Li, Jing Yu Koh, Han Zhang, Ruoming Pang, James Qin, Alexander Ku, Yuanzhong Xu, Jason Baldridge, and Yonghui Wu. 2022. https://openreview.net/forum?id=pfNyExj7z2 Vector-quantized image modeling with improved VQGAN . In International Conference on Learning Repre...

    2022

  36. [44]

    Li Zeng, Zeming Liu, Chong Feng, Heyan Huang, and Yuhang Guo. 2025. https://doi.org/10.18653/v1/2025.findings-acl.1012 D oc ME dit: Towards document-level model editing . In Findings of the Association for Computational Linguistics: ACL 2025, pages 19725--19743, Vienna, Austri...

    2025 doi

  37. [45]

    Lingjun Zhang, Xinyuan Chen, Yaohui Wang, Yue Lu, and Yu Qiao. 2024 a . https://doi.org/10.1609/aaai.v38i7.28550 Brush your text: Synthesize any scene text on images via diffusion model . Proceedings of the AAAI Conference on Artificial Intelligence, 38(7):7215--7223

    2024 doi

  38. [46]

    Efros, Eli Shechtman, and Oliver Wang

    Richard Zhang, Phillip Isola, Alexei A. Efros, Eli Shechtman, and Oliver Wang. 2018. The unreasonable effectiveness of deep features as a perceptual metric. In Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition (CVPR)

    2018

  39. [47]

    Shaolei Zhang, Qingkai Fang, Shoutao Guo, Zhengrui Ma, Min Zhang, and Yang Feng. 2024 b . https://doi.org/10.18653/v1/2024.acl-long.485 S tream S peech: Simultaneous speech-to-speech translation with multi-task learning . In Proceedings of the 62nd Annual Meeting of the Associ...

    2024 doi

  40. [48]

    Zhiyang Zhang, Yaping Zhang, Yupu Liang, Zhiyuan Chen, Lu Xiang, Yang Zhao, Yu Zhou, and Chengqing Zong. 2025 a . https://doi.org/10.18653/v1/2025.findings-acl.372 A query-response framework for whole-page complex-layout document image translation with relevant regional concen...

    2025 doi

  41. [49]

    Zhiyang Zhang, Yaping Zhang, Yupu Liang, Cong Ma, Lu Xiang, Yang Zhao, Yu Zhou, and Chengqing Zong. 2025 b . https://doi.org/10.1109/TPAMI.2025.3530998 Understand layout and translate text: Unified feature-conductive end-to-end document image translation . IEEE Transactions on...

    2025

  42. [50]

    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

  43. [51]

    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...

  44. [52]

    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 gl...

Pith tools

Reviewed August 5, 2026 · model on record in the stance chip above.