Pith. sign in

REVIEW 5 major objections 6 minor 34 references

TopicVD: A Topic-Based Dataset of Video-Guided Multimodal Machine Translation for Documentaries

T0 review · 5 major / 6 minor · reviewed 2026-08-15 · deepseek-v4-flash

Pith's one-line read Documentary video clips, labeled by topic and position, consistently raise machine translation quality, lifting BLEU from 19.14 to 29.33.

desk verdict TopicVD is a genuinely useful new dataset; the BLEU claim is plausible but under-validated, and the MPNet filtering step deserves a closer look. read the letter →

arxiv 2505.05714 v1 pith:TVNRIUC5 submitted 2025-05-09 cs.CL

classification cs.CL
keywords documentarytranslationvideo-guidedmachinemultimodaltopic-baseddatasetcross-modalattentiondomainadaptationglobalcontextChinese-Englishparallelcorpus
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 constructs TopicVD, a Chinese-English parallel corpus of 122,930 subtitle pairs from 256 documentaries, organized into eight topic labels and annotated with each clip's position in its source documentary. It aims to show that video-guided multimodal machine translation works for documentaries, a setting where existing image- or short-clip datasets are too simple and where long-form context matters. To make this case, the authors propose a model with cross-modal bidirectional attention and report that adding video raises BLEU from 19.14 for text-only translation to 29.33, with further gains when nearby clips are used as context. The dataset is intended as a resource for studying domain adaptation in documentary translation, a task the paper argues has been missing from multimodal machine translation.

What carries the argument

The central object is TopicVD itself: 122,930 Chinese-English subtitle pairs from 256 documentaries, split into eight topics, with each pair carrying timestamps, a quality score, and its position in the full documentary. The model's load-bearing component is a cross-modal bidirectional attention module, which first uses text as a query to select relevant video frames and then computes an alignment matrix between text tokens and video regions so that each modality updates the other. Video features come from a pretrained video encoder and text from a transformer, and the position and topic metadata are what make the paper's domain and context experiments possible.

What would settle it

Train the text-only and video-guided models separately on the held-out low-score pairs that the cosine-similarity filter would remove, and check whether the video-guided model still beats text-only by the same margin; if the gap collapses on those pairs, the reported BLEU gain is an artifact of filtering toward easy, literal translations rather than evidence that video helps documentary translation.

Watch

Extended reading notes

Core claim

The paper's central claim is that video information consistently improves documentary translation when the data is organized by topic and clip position. With the proposed cross-modal attention model, multimodal translation reaches BLEU 29.33 on TopicVD while a text-only transformer scores 19.14, and the multimodal model also beats two frame-selection baselines by about 2.6 and 3.4 BLEU. The paper further claims that domain mismatch is costly: on History, Figure, Nature, and Technology test sets, models trained on same-topic data outperform equally sized out-of-domain models, though larger out-of-domain training pools can partially compensate. It also claims that supplying three or ten video clips from the same documentary as global context improves translation from 29.33 to 29.95 and 30.22, so both visual and contextual information help.

Load-bearing premise

The load-bearing premise is that the cosine-similarity filter in Section 3.2 removes only low-quality parallel pairs and not systematically hard or non-literal translations; if the filter keeps mostly easy pairs, the measured visual benefit could be an artifact of dataset construction.

Editorial extensions

If this is right

  • If the reported gains hold, TopicVD provides a benchmark where visual context changes translation output, allowing MMT models to be tested on long-form documentary content rather than isolated images.
  • The eight-topic split gives a controlled setup for measuring in-domain versus out-of-domain behavior in video-guided machine translation.
  • Positional metadata opens the door to models that use global documentary context, not just the single clip aligned with a sentence.
  • The out-of-domain BLEU drop indicates that documentary translation systems will need domain adaptation methods, and the augmentation results point to targeted in-domain data as one workable remedy.
  • The reported BLEU numbers give later work concrete reference points: 19.14 for text-only, 29.33 for the proposed multimodal model.

Reading between the lines

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

  • A testable extension is to train the same models on the pairs the cosine-similarity filter would remove: if the video advantage shrinks or disappears there, the headline gain is partly an artifact of keeping easy, literal pairs.
  • The preserved position information could let future models reason across scenes within a documentary, a capability the paper only begins to test with three-to-ten clip context windows.
  • The per-pair quality scores could be reused as difficulty labels, making it possible to measure video's contribution on hard versus easy translations instead of averaged BLEU.
  • The topic labels also enable retrieval-style context selection from same-topic documentaries, a domain-adaptation strategy the paper does not itself test.
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

5 major / 6 minor

Summary. The paper introduces TopicVD, a Chinese-English video-subtitle parallel corpus built from 256 documentaries (285 hours, 122,930 pairs), annotated with eight topic labels and per-clip position metadata. It also proposes a VMT/MMT model with a cross-modal bidirectional attention module and reports experiments on TopicVD: the proposed model reaches BLEU 29.33 versus 19.14 for a text-only NMT baseline and 25.73-26.71 for image-based frame-selection baselines (Table 4). Additional experiments address in-domain vs. out-of-domain training (Table 5), data augmentation (Table 6), and contextual video clips (Table 7). The central claim is that visual information consistently improves documentary translation and that topic/context information matters for this task.

Significance. TopicVD fills a real gap: existing VMT corpora such as How2, VATEX, BigVideo, EVA, and TAIL are not topic-organized and mostly use short clips, whereas documentaries offer long-form content with strong text-video correlation. The topic annotations and preserved clip positions are genuinely useful for domain-adaptation and context-aware VMT research, and the authors indicate that the dataset and implementations are publicly available. The reported 10-point BLEU gap, if it survives the filtering and statistical checks below, would be a substantial demonstration that video can help documentary translation. The proposed model is a modest combination of existing selective and bidirectional attention ideas, but it serves the dataset-validation purpose.

major comments (5)
  1. [3.2, 3.4] The MPNet-based quality filter is not actually specified. The manuscript states that the cosine-similarity score 'can be used to filter out low-quality pairs across different experimental settings' and that scores serve as a basis for 'similarity-based filtering experiments,' but no threshold, no fraction of removed pairs, and no score distribution are reported, and no such filtering experiment appears. Because MPNet embeddings reward literal cross-lingual paraphrase, a high-similarity cutoff would preferentially retain easy, literal pairs; the 10-point BLEU advantage of the proposed method in Table 4 could then be an artifact of dataset selection rather than evidence that visual information guides translation. Please report the filtering details and repeat the main comparison on the unfiltered corpus and on the low-similarity retained subset.
  2. [Tables 4-7] No variance, confidence intervals, or significance tests are reported, despite the text using 'significantly' in Sections 5.2 and 6.1. Since each table cell appears to be a single run, the reader cannot tell whether the 10.19-BLEU gap in Table 4 or the 0.6-0.9 gains in Table 7 are stable. Please report at least three random seeds with means and standard deviations, or bootstrap confidence intervals on test predictions, for all main comparisons.
  3. [6.2] The augmentation experiment is acknowledged in the manuscript to include 'a documentary closely related to the test set content' in the added Nature data. This makes the 20.16-to-30.60 improvement uninterpretable as evidence about domain adaptation: the gain may simply reflect near-test-content leakage. Either remove this condition or add a controlled augmentation set with no topical/test overlap; the current conclusion about targeted augmentation is not supported.
  4. [6.1] The 'Out-of-domain dataset (full)' condition is not comparable to the in-domain condition because it uses seven topics' worth of training data, so the higher BLEU scores (e.g., 24.76 vs. 17.45 for Figure) confound domain with data volume. The sampled out-of-domain condition is the only controlled comparison, and the prose should be revised to avoid claiming that out-of-domain (full) outperformance says anything about domain adaptation. The term 'significantly' should also be replaced with quantitative support once variance is reported.
  5. [Abstract, 5.2] The claim that visual information 'consistently improves' translation quality is supported only by one aggregate test-set BLEU comparison in Table 4. No per-topic text-only vs. multimodal comparison is reported, so the word 'consistently' overstates the evidence; the authors should either provide per-topic results or soften the claim.
minor comments (6)
  1. [5.1] The training setup lists two different learning rates ('learning rate of 0.005' and 'learning rate is set to 1e-4'); please clarify which value was used and whether 'learning rate' was changed during training.
  2. [4.2] The notation in the Bi-Attention equations reuses h_n for both the original and the updated text feature; please rename the updated feature (e.g., h'_n) to avoid ambiguity.
  3. [1] The topic list contains the typo 'Nagure'; it should be 'Nature'.
  4. [3.3] Reference [27] is cited for FFmpeg, but the reference is about image stream hashing and appears unrelated; please cite the FFmpeg documentation or a relevant reference instead.
  5. [3.2] The paper refers to 'a multilingual sentence embedding model (e.g., MPNet)' but does not specify which MPNet variant was used. If an English-only MPNet was used for Chinese-English similarity, the quality scores would be unreliable; please name the exact model and, ideally, report its correlation with human quality judgments.
  6. [6.3] The context experiment selects neighboring subtitle clips via text-text similarity, but the paper says the positional information is preserved; please clarify whether the stored position metadata is actually used in this experiment or only the textual similarity.

Circularity Check

0 steps flagged · score 0.0 of 10

No circularity: the dataset construction and model evaluation are self-contained, and the central BLEU comparisons are empirical rather than reducible to the paper's own definitions.

full rationale

TopicVD is an empirical dataset-and-model paper. The central claim that visual information improves documentary translation BLEU from 19.14 to 29.33 (Table 4) is obtained by training a text-only NMT, several IMT baselines, and the proposed VMT model on the same train/validation/test splits (Table 3) and reporting held-out test-set BLEU. No evaluation quantity is defined in terms of the model output, and the test documentaries are disjoint from the training documentaries, so there is no fit-and-predict-on-the-same-data step. The MPNet cosine 'quality estimation' score (Section 3.2) is an external pretrained similarity measure used to filter subtitle pairs before training; it is not fitted to the translation model, and the paper does not claim to predict that score. The reader's concern that the MPNet filter may select easy, literal pairs is a plausible dataset-bias hypothesis, but the paper gives no threshold, removal fraction, or score distribution that would let one exhibit a specific reduction from filter to result, so under the hard rules it is not circularity. The proposed cross-modal bidirectional attention is inspired by [13] and [24], and [24] includes two co-authors of this paper; however, that self-citation is used as prior architectural inspiration, and the proposed module is evaluated against external baselines, so it is not load-bearing. The paper itself flags two limitations: footnote 10 acknowledges that topic distribution is influenced by publicly available documentaries, and Section 6.2 admits that the augmented Nature set contains a documentary closely related to the test set, weakening that augmentation comparison. These are data-quality caveats, not circular derivations. No equation or reported result reduces to its own input, so the circularity score is 0.

Assumptions & free parameters 2 free parameters · 4 assumptions · 0 invented entities

The central empirical claims rest on the quality and alignment of the dataset, which depends on the filtering and alignment assumptions listed above. No new physical or conceptual entities are introduced.

free parameters (2)
  • SSIM threshold for frame deduplication = 0.5
    Used in the heuristic frame extraction baseline (Section 5.1) to eliminate redundant frames; chosen by hand, not optimized.
  • Quality estimation score cutoff
    Section 3.2 computes cosine similarity scores with MPNet but does not specify the exact threshold used to filter low-quality pairs in the released dataset; if applied, it is a free parameter affecting dataset composition.
assumptions (4)
  • domain assumption The BLEU metric is a valid proxy for translation quality in this setting.
    Used for all model comparisons in Section 5.1 and subsequent experiments.
  • domain assumption Video clips and subtitle texts are correctly aligned and the visual content is semantically relevant to the text.
    Dataset construction relies on Whisper verification and manual checks, as described in Section 3.3.
  • domain assumption The eight topic labels are accurate and mutually exclusive.
    Topics are assigned based on platform metadata and manual judgment, Section 3.1.
  • domain assumption MPNet-based cosine similarity between translations is a reliable indicator of subtitle pair quality.
    Used to filter low-quality pairs in Section 3.2; if the metric is not reliable, the dataset may be biased.

how reviews work

0 comments
Cite this review

Pith. "Pith review of TopicVD: A Topic-Based Dataset of Video-Guided Multimodal Machine Translation for Documentaries." pith.science (2026). https://pith.science/paper/TVNRIUC5

@misc{pith2026250505714,
  author       = {Pith},
  title        = {Pith review of: TopicVD: A Topic-Based Dataset of Video-Guided Multimodal Machine Translation for Documentaries},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/TVNRIUC5}},
  note         = {Machine review of arXiv:2505.05714}
}
read the original abstract

Most existing multimodal machine translation (MMT) datasets are predominantly composed of static images or short video clips, lacking extensive video data across diverse domains and topics. As a result, they fail to meet the demands of real-world MMT tasks, such as documentary translation. In this study, we developed TopicVD, a topic-based dataset for video-supported multimodal machine translation of documentaries, aiming to advance research in this field. We collected video-subtitle pairs from documentaries and categorized them into eight topics, such as economy and nature, to facilitate research on domain adaptation in video-guided MMT. Additionally, we preserved their contextual information to support research on leveraging the global context of documentaries in video-guided MMT. To better capture the shared semantics between text and video, we propose an MMT model based on a cross-modal bidirectional attention module. Extensive experiments on the TopicVD dataset demonstrate that visual information consistently improves the performance of the NMT model in documentary translation. However, the MMT model's performance significantly declines in out-of-domain scenarios, highlighting the need for effective domain adaptation methods. Additionally, experiments demonstrate that global context can effectively improve translation performance. % Dataset and our implementations are available at https://github.com/JinzeLv/TopicVD

Figures

Figures reproduced from arXiv: 2505.05714 by the authors.

Figure 1
Figure 1. Construction process of the TopicVD documentary translation dataset 3.1 Data Collection The collected data consists of documentary videos along with corresponding Chinese and English subtitle files. The specific collection steps are as follows: (1) We select highly played and highly rated documentaries from video platforms such as Douban Movie 5 , Tencent Video 6 and Youku Video 7 , ensuring that the dataset consist… view at source ↗
Figure 2
Figure 2. An overview of the proposed VMT model with cross-modal bidirectional atten￾tion It is important to note that, unlike encoding an image to obtain its semantic vector directly, encoding a video requires first extracting the most representa￾tive frames before encoding them. Therefore, we employ a single-head attention network as a selective attention mechanism to establish associations between the text and the extracte… view at source ↗
Figure 3
Figure 3. Example of correct translation by the proposed method [PITH_FULL_IMAGE:figures/full_fig_p011_3.png] view at source ↗

Discussion (0). Continue with ORCID to comment.

Reference graph

Works this paper leans on

34 extracted references · 12 canonical work pages

  1. [1]

    Com- putational linguistics 16(2), 79–85 (1990)

    Brown, P.F., Cocke, J., Della Pietra, S.A., Della Pietra, V.J., Jelinek, F., Lafferty, J., Mercer, R.L., Roossin, P.S.: A statistical approach to machine translation. Com- putational linguistics 16(2), 79–85 (1990)

  2. [2]

    Computational linguistics 19(2), 263–311 (1993)

    Brown, P.F., Della Pietra, S.A., Della Pietra, V.J., Mercer, R.L.: The mathematics of statistical machine translation: Parameter estimation. Computational linguistics 19(2), 263–311 (1993)

  3. [3]

    Caglayan, O., Madhyastha, P., Specia, L., Barrault, L.: Probing the need for visual contextinmultimodalmachinetranslation.arXivpreprintarXiv:1903.08678(2019)

  4. [4]

    In: proceedings of the IEEE Conference on Computer Vision and Pattern Recognition

    Carreira, J., Zisserman, A.: Quo vadis, action recognition? a new model and the kinetics dataset. In: proceedings of the IEEE Conference on Computer Vision and Pattern Recognition. pp. 6299–6308 (2017)

  5. [5]

    arXiv preprint arXiv:1710.07177 (2017)

    Elliott, D., Frank, S., Barrault, L., Bougares, F., Specia, L.: Findings of the second shared task on multimodal machine translation and multilingual image description. arXiv preprint arXiv:1710.07177 (2017)

  6. [6]

    arXiv preprint arXiv:1605.00459 (2016)

    Elliott, D., Frank, S., Sima’an, K., Specia, L.: Multi30k: Multilingual english- german image descriptions. arXiv preprint arXiv:1605.00459 (2016)

  7. [7]

    arXiv preprint arXiv:1705.04350 (2017)

    Elliott, D., Kádár, A.: Imagination improves multimodal translation. arXiv preprint arXiv:1705.04350 (2017)

  8. [8]

    Multimodal Pivots for Image Caption Translation

    Hitschler, J., Schamoni, S., Riezler, S.: Multimodal pivots for image caption trans- lation. arXiv preprint arXiv:1601.03916 (2016)

Show all 34 references
  1. [9]

    arXiv preprint arXiv:2308.12038 (2023)

    Hu, J., Yao, Y., Wang, C., Wang, S., Pan, Y., Chen, Q., Yu, T., Wu, H., Zhao, Y., Zhang, H., et al.: Large multilingual models pivot zero-shot multimodal learning across languages. arXiv preprint arXiv:2308.12038 (2023)

  2. [10]

    arXiv preprint arXiv:2305.18326 (2023)

    Kang, L., Huang, L., Peng, N., Zhu, P., Sun, Z., Cheng, S., Wang, M., Huang, D., Su, J.: Bigvideo: A large-scale video subtitle translation dataset for multimodal machine translation. arXiv preprint arXiv:2305.18326 (2023)

  3. [11]

    arXiv preprint arXiv:2305.17415 (2023)

    Lan, Z., Yu, J., Li, X., Zhang, W., Luan, J., Wang, B., Huang, D., Su, J.: Ex- ploring better text image translation with multimodal codebook. arXiv preprint arXiv:2305.17415 (2023)

  4. [12]

    arXiv preprint arXiv:2203.09173 (2022)

    Li, B., Lv, C., Zhou, Z., Zhou, T., Xiao, T., Ma, A., Zhu, J.: On vision features in multimodal machine translation. arXiv preprint arXiv:2203.09173 (2022)

  5. [13]

    arXiv preprint arXiv:2109.03415 (2021)

    Li, J., Ataman, D., Sennrich, R.: Vision matters when it should: Sanity checking multimodal machine translation models. arXiv preprint arXiv:2109.03415 (2021)

  6. [14]

    arXiv preprint arXiv:2310.20201 (2023)

    Li, Y., Shimizu, S., Chu, C., Kurohashi, S., Li, W.: Video-helpful multimodal machine translation. arXiv preprint arXiv:2310.20201 (2023)

  7. [15]

    arXiv preprint arXiv:2201.08054 (2022)

    Li, Y., Shimizu, S., Gu, W., Chu, C., Kurohashi, S.: Visa: An ambiguous subtitles dataset for visual scene-aware machine translation. arXiv preprint arXiv:2201.08054 (2022)

  8. [16]

    In: Proceed- ingsofthe28thACMinternationalconferenceonmultimedia.pp.1320–1329(2020)

    Lin, H., Meng, F., Su, J., Yin, Y., Yang, Z., Ge, Y., Zhou, J., Luo, J.: Dynamic context-guided capsule network for multimodal machine translation. In: Proceed- ingsofthe28thACMinternationalconferenceonmultimedia.pp.1320–1329(2020)

  9. [17]

    In: Computer vision– ECCV 2014: 13th European conference, zurich, Switzerland, September 6-12, 2014, proceedings, part v 13

    Lin, T.Y., Maire, M., Belongie, S., Hays, J., Perona, P., Ramanan, D., Dollár, P., Zitnick, C.L.: Microsoft coco: Common objects in context. In: Computer vision– ECCV 2014: 13th European conference, zurich, Switzerland, September 6-12, 2014, proceedings, part v 13. pp. 740–755...

  10. [18]

    arXiv preprint arXiv:1908.03265 (2019) Title Suppressed Due to Excessive Length 15

    Liu, L., Jiang, H., He, P., Chen, W., Liu, X., Gao, J., Han, J.: On the variance of the adaptive learning rate and beyond. arXiv preprint arXiv:1908.03265 (2019) Title Suppressed Due to Excessive Length 15

  11. [19]

    In: Proceedings of the 40th annual meeting of the Association for Computational Linguistics

    Papineni, K., Roukos, S., Ward, T., Zhu, W.J.: Bleu: a method for automatic evaluation of machine translation. In: Proceedings of the 40th annual meeting of the Association for Computational Linguistics. pp. 311–318 (2002)

  12. [20]

    In: Proceedings of the IEEE international conference on computer vision

    Plummer, B.A., Wang, L., Cervantes, C.M., Caicedo, J.C., Hockenmaier, J., Lazeb- nik, S.: Flickr30k entities: Collecting region-to-phrase correspondences for richer image-to-sentence models. In: Proceedings of the IEEE international conference on computer vision. pp. 2641–2649 (2015)

  13. [21]

    arXiv preprint arXiv:1811.00347 (2018)

    Sanabria, R., Caglayan, O., Palaskar, S., Elliott, D., Barrault, L., Specia, L., Metze, F.: How2: a large-scale dataset for multimodal language understanding. arXiv preprint arXiv:1811.00347 (2018)

  14. [22]

    Advances in neural information processing systems 33, 16857–16867 (2020)

    Song, K., Tan, X., Qin, T., Lu, J., Liu, T.Y.: Mpnet: Masked and permuted pre- training for language understanding. Advances in neural information processing systems 33, 16857–16867 (2020)

  15. [23]

    Information Sciences 554, 47–60 (2021)

    Su, J., Chen, J., Jiang, H., Zhou, C., Lin, H., Ge, Y., Wu, Q., Lai, Y.: Multi-modal neural machine translation with deep semantic interactions. Information Sciences 554, 47–60 (2021)

  16. [24]

    arXiv preprint arXiv:2208.00767 (2022)

    Tang, Z., Zhang, X., Long, Z., Fu, X.: Multimodal neural machine translation with search engine based image retrieval. arXiv preprint arXiv:2208.00767 (2022)

  17. [25]

    In: Proceedings of the 3rd Workshop on Advances in Language and Vision Research (ALVR)

    Teramen, A., Ohtsuka, T., Kondo, R., Kajiwara, T., Ninomiya, T.: English-to- japanese multimodal machine translation based on image-text matching of lecture videos. In: Proceedings of the 3rd Workshop on Advances in Language and Vision Research (ALVR). pp. 86–91 (2024)

  18. [26]

    Advances in neural information pro- cessing systems 30 (2017)

    Vaswani,A.,Shazeer,N.,Parmar,N.,Uszkoreit,J.,Jones,L.,Gomez,A.N.,Kaiser, Ł., Polosukhin, I.: Attention is all you need. Advances in neural information pro- cessing systems 30 (2017)

  19. [27]

    Journal of Forensic Sciences69(2), 515–528 (2024)

    Wales, G.S.: Validation of image stream hashing: A forensic method for content verification. Journal of Forensic Sciences69(2), 515–528 (2024)

  20. [28]

    In: Proceedings of the IEEE/CVF international conference on computer vision

    Wang, X., Wu, J., Chen, J., Li, L., Wang, Y.F., Wang, W.Y.: Vatex: A large-scale, high-quality multilingual dataset for video-and-language research. In: Proceedings of the IEEE/CVF international conference on computer vision. pp. 4581–4591 (2019)

  21. [29]

    IEEE transactions on image processing 13(4), 600–612 (2004)

    Wang, Z., Bovik, A.C., Sheikh, H.R., Simoncelli, E.P.: Image quality assessment: from error visibility to structural similarity. IEEE transactions on image processing 13(4), 600–612 (2004)

  22. [30]

    arXiv preprint arXiv:2105.14462 (2021)

    Wu, Z., Kong, L., Bi, W., Li, X., Kao, B.: Good for misconceived reasons: An em- pirical revisiting on the need for visual context in multimodal machine translation. arXiv preprint arXiv:2105.14462 (2021)

  23. [31]

    Journal of Information Processing30, 388–396 (2022)

    Yang,Z.,Hirasawa,T.,Komachi,M.,Okazaki,N.:Whyvideosdonotguidetransla- tions in video-guided machine translation? an empirical evaluation of video-guided machine translation dataset. Journal of Information Processing30, 388–396 (2022)

  24. [32]

    In: Proceedings of the 58th annual meeting of the association for computational linguistics

    Yao, S., Wan, X.: Multimodal transformer for multimodal machine translation. In: Proceedings of the 58th annual meeting of the association for computational linguistics. pp. 4346–4350 (2020)

  25. [33]

    arXiv preprint arXiv:2007.08742 (2020)

    Yin, Y., Meng, F., Su, J., Zhou, C., Yang, Z., Zhou, J., Luo, J.: A novel graph- based multi-modal fusion encoder for neural machine translation. arXiv preprint arXiv:2007.08742 (2020)

  26. [34]

    arXiv preprint arXiv:2212.10313 (2022)

    Zhu, Y., Sun, Z., Cheng, S., Huang, L., Wu, L., Wang, M.: Beyond triplet: Leveraging the most data for multimodal machine translation. arXiv preprint arXiv:2212.10313 (2022)

Pith tools

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