Pith. sign in

REVIEW 4 major objections 4 minor 1 cited by

Improvement in Sign Language Translation Using Text CTC Alignment

T0 review · 4 major / 4 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 A text-side CTC loss lifts sign language translation by 7 BLEU

desk verdict Useful SLT recipe, but the paper's headline attribution of the gain to text CTC does not survive its own ablations; the final pipeline changes too much at once. read the letter →

arxiv 2412.09014 v4 pith:ZB6GSKYI submitted 2024-12-12 cs.CL

Sihan Tan , Taro Miyazaki , Nabeela Khan , Kazuhiro Nakadai This is my paper · ORCID
classification cs.CL
keywords signlanguagetranslationconnectionisttemporalclassificationtextCTCalignmentjointCTC/attentiondecodinghierarchicalencodingtransferlearninggloss-freePHOENIX14T
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 claims that applying Connectionist Temporal Classification (CTC) directly to the target spoken text — not only to glosses — improves end-to-end sign language translation. The author argues that a hierarchical encoder, where a gloss-oriented CTC adjusts sequence lengths and a text-oriented CTC (TxtCTC) reorders representations, can handle the non-monotonic mapping between sign video and spoken text. Combined with a joint CTC/attention decoder and two-stage transfer learning, the method outperforms the pure-attention Sign Language Transformer baseline by 6.97 BLEU on PHOENIX14T and 6.50 BLEU on CSL-Daily development sets. If the claim holds, CTC's reordering capability, previously exploited in speech and text translation, extends to sign languages and offers a gloss-free route for SLT.

What carries the argument

The load-bearing component is TxtCTC, a Connectionist Temporal Classification loss computed against the target spoken text at the output of the text-oriented encoder. CTC is the standard alignment loss that marginalizes over all monotonic frame-to-label paths using a blank label; here it is applied to the text side of sign language translation, where the mapping is non-monotonic. The paper's mechanism is a hierarchical encoder: GlsEnc first compresses sign representations under gloss-CTC supervision (length adjustment), then TxtEnc reorders the resulting representations under TxtCTC supervision (reordering), and a joint decoder combines attention scores with normalized TxtCTC scores in beam search. Transfer learning adds a warm-start stage in which multiple pre-trained sign embeddings are paired with augmented spoken text, followed by fine-tuning on a single embedding and the original corpus.

What would settle it

Inspect the TxtCTC alignment posteriors on PHOENIX14T development utterances where the gloss order and spoken-text order differ; if the most probable alignment path is nearly monotonic while the BLEU gain over pure attention persists, then the latent-reordering mechanism the paper appeals to is not what carries the improvement.

Watch

Extended reading notes

Core claim

The central claim is that CTC is not limited to monotonic gloss alignment in sign language translation: a text-side CTC loss, trained against the spoken text at the output of an additional encoder stage, can perform the latent reordering needed to map sign video to natural-language word order. The model therefore uses two stacked encoder blocks — GlsEnc, supervised by gloss CTC for length adjustment, and TxtEnc, supervised by TxtCTC for reordering — followed by an attention decoder whose beam search scores hypotheses with a weighted combination of attention and TxtCTC likelihoods. On RWTH-PHOENIX-Weather 2014 T the full pipeline reaches 27.93 BLEU on the development set, 6.97 points above the pure-attention baseline, and on CSL-Daily it reaches 22.04 BLEU, 6.50 points higher; these results are comparable to or better than reported state-of-the-art end-to-end systems. The authors further show that TxtCTC alone, without any gloss supervision, already improves joint-decoding performance, which they offer as evidence toward gloss-free sign language translation.

Load-bearing premise

The load-bearing premise is that CTC's monotonic alignment constraint can be overcome by the network's latent reordering, so that a CTC loss computed against the spoken text actually teaches the encoder to reorder sign representations; if that reordering never happens, the measured gains could be due to the extra parameters, the joint decoder, or the transfer-learning setup rather than to text CTC alignment.

Editorial extensions

If this is right

  • TxtCTC alone, with no gloss supervision, reaches 25.63 BLEU on PHOENIX14T and 15.79 BLEU on CSL-Daily in joint decoding, so gloss-free sign language translation can keep the alignment benefits of CTC.
  • Hierarchical encoding and joint decoding are complementary: combining them improves over either alone by 0.53 BLEU on PHOENIX14T and 0.90 BLEU on CSL-Daily.
  • Warm-start transfer learning converts the performance drop from combining multiple sign embeddings into a gain (+0.98 BLEU on PHOENIX14T and +3.00 BLEU on CSL-Daily), and fine-tuning adds a further gain.
  • The final system outperforms the pure-attention Sign Language Transformer baseline by 6.97 BLEU on PHOENIX14T and 6.50 BLEU on CSL-Daily, producing translations that are more faithful and detailed in the paper's case study.

Reading between the lines

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

  • A direct way to test the reordering story is to read out TxtCTC's most probable alignment paths on utterances where gloss order and text order differ; if the translations still improve while the alignments stay monotonic, the gain is not coming from latent reordering.
  • Because the pre-trained sign embeddings (SMKD, Corrnet, CNN+LSTM+HMM) were themselves trained with gloss supervision, the current pipeline is not yet fully gloss-free; replacing them with self-supervised video encoders would be the decisive test of the paper's gloss-free claim.
  • The transfer-learning gains may owe as much to back-translated augmented text as to the combination of embeddings; ablating augmented text from the warm-start stage while keeping multiple embeddings would separate the two contributions.
  • The same hierarchical TxtCTC recipe should transfer to other weakly-supervised video-to-text problems, such as instructional video captioning, where the source order and target order differ systematically.
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 proposes a joint CTC/attention framework for end-to-end sign language translation, with a hierarchical encoder consisting of a gloss-oriented encoder (GlsEnc) trained with GlsCTC and a text-oriented encoder (TxtEnc) trained with a newly introduced TxtCTC loss, along with joint CTC/attention decoding and a two-stage transfer-learning procedure using multiple pre-trained sign embeddings. Experiments on PHOENIX14T and CSL-Daily report that the method outperforms a pure-attention baseline by 6.97 BLEU on PHOENIX14T and 6.50 BLEU on CSL-Daily (Table 2, rows 1→10 and 2→10), and the paper claims results comparable to or better than published state-of-the-art (Tables 3 and 4).

Significance. If the attribution to text-side CTC were established, this would be a useful contribution to gloss-free SLT, showing that CTC's reordering capability can be applied beyond monotonic gloss alignment. The paper provides code and detailed hyperparameters, and the ablation study covers several components on two benchmarks, which is a strength. However, the current evidence does not isolate the contribution of TxtCTC: the largest contiguous gains in Table 2 come from the transfer-learning stages, the isolated CTC gains are small and sometimes negative on CSL-Daily, and no variance or significance information is reported. The contribution is therefore interesting but needs stronger controlled evidence before the title-level claim is supported.

major comments (4)
  1. [Table 2 and §5.1] The headline gains of +6.97 (PHOENIX14T) and +6.50 (CSL-Daily) are not controlled comparisons. Moving from ID1/ID2 to ID10 changes the sign embedding, adds GlsCTC and TxtCTC losses, adds joint decoding, and adds multi-embedding warm-start training followed by fine-tuning; on CSL-Daily no pure-attention baseline row is reported at all, since ID2 already includes the SMKD embedding replacement. The isolated TxtCTC contributions are small and mixed (ID4−ID2: +0.68 on PHOENIX14T and −0.18 on CSL-Daily), while the largest contiguous gains come from the transfer-learning rows (ID9−ID7: +0.70 and +1.12; ID10−ID9: +0.86 and +1.75). The central claim that text-side CTC alignment is responsible for the improvement therefore needs a matched ablation in which TxtCTC is removed while the added TxtEnc parameters, the extra loss, and the transfer-learning schedule are held fixed.
  2. [§5 and Tables 2–4] All BLEU differences are reported as single runs with no error bars or significance tests. In the key contrasts, the differences are small (e.g., ID4 vs ID2 is +0.68 on PHOENIX14T and −0.18 on CSL-Daily; ID6 vs ID2 is +1.31 and +0.25), so the contributions of TxtCTC and joint decoding cannot be distinguished from optimization noise. Please report multiple seeds or another variance estimate and, where feasible, significance tests for the main ablations.
  3. [Table 3] The pure-attention baseline used as the reference point reproduces 20.96 BLEU on the PHOENIX14T dev set, while the published Joint-SLRT result listed in the same table is 22.38 BLEU. Because the headline 'outperforms pure-attention baseline' is computed against this weaker reimplementation, the authors should either tune the baseline to match the published configuration, explain why the reimplementation is lower, or state the comparison relative to the published number.
  4. [§3.2, Eq. (8), and §5.3] The premise that 'the neural network architecture enables latent reordering' is asserted but never directly verified. No analysis of the learned TxtCTC alignments, no diagnostic of monotonicity or reordering, and no control with an alternative non-CTC reordering mechanism is provided. Given the confounds in Table 2, adding such an analysis is necessary to support the paper's central claim that text CTC alignment is the source of the improvement.
minor comments (4)
  1. [Table 1] The CSL-Daily training-sentence count is printed as 184,401; if the standard split is 18,401, this appears to be a typo and should be corrected.
  2. [Table 6 and Table 1] The Table 6 header contains 'Hyerparameters' and 'papramenters', and Table 1 uses 'V ocab.'; these typos should be fixed.
  3. [Eq. (2) and Algorithm 1] Equation (2) writes 'yl:l−1'; this should be 'y1:l−1' or 'y<l', and the notation in Algorithm 1 for accumulated hypothesis scores should be defined more explicitly.
  4. [Table 8] The case-study translations are presented without marking that the English glosses are approximate; adding a note would prevent readers from misreading the English approximations as literal translations.

Circularity Check

0 steps flagged · score 1.0 of 10

No significant circularity: reported gains are held-out BLEU differences, and no equation or fitted parameter reduces by construction to the claimed prediction.

full rationale

Walking the paper's derivation chain: the central claim is an empirical improvement on held-out dev/test BLEU (Table 2, ID 1 to 10; Tables 3-4). The training objective in Eq. 9 is a weighted sum of GlsCTC, TxtCTC, and MLE losses, with weights lambda1, lambda2, lambda3 tuned on development sets, which is standard practice and not circular. Joint decoding (Algorithm 1) combines TxtCTC and attention scores, but this is a model-combination decoding procedure, not a quantity defined as its own output. No equation in the paper defines a predicted value in terms of the fitted value it is supposed to predict; the reported BLEU numbers are out-of-sample evaluations on standard benchmarks. The 'latent reordering' statement in Section 3.2 is an architectural assumption rather than a result derived from itself, and the paper does not rely on it as a tautological proof of improvement. The ablation in Table 2 changes several components simultaneously, so attributing the final +6.97 BLEU gain specifically to TxtCTC is under-supported; however, a confounded ablation is a validity concern, not circularity under the stated rules. The only self-citation, Tan et al. (2024), appears in a related-work enumeration and is not load-bearing; it does not supply the paper's premise or uniqueness claim. Therefore, no circular step is exhibited, and the paper is self-contained against external benchmarks.

Assumptions & free parameters 6 free parameters · 4 assumptions · 0 invented entities

The central result is an empirical method comparison, not a mathematical derivation. It relies on standard CTC and attention machinery, plus several domain assumptions about latent reordering and transferability of pretrained sign embeddings. The only hand-tuned numbers are loss weights and decoding parameters, which are calibrated on development sets. No new particles, forces, or theoretical entities are introduced.

free parameters (6)
  • λ1 (GlsCTC loss weight) = 1.0 (PHOENIX14T), 5.0 (CSL-Daily)
    Set per dataset in Table 6 and used in the total loss LSLT = λ1 LGlsCTC + λ2 LTxtCTC + λ3 LMLE. Tuned on development BLEU.
  • λ2 (TxtCTC loss weight) = 1.0 (PHOENIX14T), 2.0 (CSL-Daily)
    Set per dataset in Table 6 and used in the total loss. Tuned on development BLEU.
  • λ3 (MLE loss weight) = 1.0 (both datasets)
    Used in the total loss in Table 6. Held at 1.0 rather than tuned.
  • CTC score weight in joint decoding = 0.3
    Used in Algorithm 1 through PBeam = α_hat_CTC + α_hat_Attn. Reported in Table 7 without a search range.
  • Length penalty β = 0.6
    Used in Algorithm 1 to adjust candidate scores during beam search. Reported in Table 7.
  • Beam size Nbeam = searched over [1..10], final value not reported
    Used in Algorithm 1 for candidate selection. Table 7 lists the grid but not the selected beam size, making the exact decoding setup under-specified.
assumptions (4)
  • standard math CTC marginalizes over monotonic alignments with a blank label and can be combined with attention for decoding.
    Invoked in §2.1 as the foundation of the method; follows Graves et al. (2006) and Yan et al. (2023). Not derived in this paper.
  • domain assumption Neural network layers can perform latent reordering so that CTC, despite its monotonicity, can align source frames to non-monotonic target text.
    Stated in §3.2: 'the neural network architecture enables latent reordering to handle the non-monotonic nature of text alignment.' This is the load-bearing assumption behind TxtCTC and is not proven independently.
  • domain assumption Frozen pretrained sign embeddings (CNN+LSTM+HMM, SMKD, Corrnet) provide transferable visual representations for SLT.
    Used in §3.1 and §4. The whole pipeline depends on these externally pretrained models being representative enough of sign video, and the paper does not test the method without them.
  • domain assumption Back-translated and normalized parallel text stays close enough to the original distribution to be useful for warm-start training.
    Used in §3.4 and §4 for data augmentation. No distribution shift analysis is provided, and the augmented corpus differs substantially in size and vocabulary from the original training data (Table 1).

how reviews work

0 comments
Cite this review

Pith. "Pith review of Improvement in Sign Language Translation Using Text CTC Alignment." pith.science (2026). https://pith.science/paper/ZB6GSKYI

@misc{pith2026241209014,
  author       = {Pith},
  title        = {Pith review of: Improvement in Sign Language Translation Using Text CTC Alignment},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/ZB6GSKYI}},
  note         = {Machine review of arXiv:2412.09014}
}
read the original abstract

Current sign language translation (SLT) approaches often rely on gloss-based supervision with Connectionist Temporal Classification (CTC), limiting their ability to handle non-monotonic alignments between sign language video and spoken text. In this work, we propose a novel method combining joint CTC/Attention and transfer learning. The joint CTC/Attention introduces hierarchical encoding and integrates CTC with the attention mechanism during decoding, effectively managing both monotonic and non-monotonic alignments. Meanwhile, transfer learning helps bridge the modality gap between vision and language in SLT. Experimental results on two widely adopted benchmarks, RWTH-PHOENIX-Weather 2014 T and CSL-Daily, show that our method achieves results comparable to state-of-the-art and outperforms the pure-attention baseline. Additionally, this work opens a new door for future research into gloss-free SLT using text-based CTC alignment.

Figures

Figures reproduced from arXiv: 2412.09014 by the authors.

Figure 1
Figure 1. An overview of sign language recognition and [PITH_FULL_IMAGE:figures/full_fig_p001_1.png] view at source ↗
Figure 2
Figure 2. Overview of the proposed method: Stage 1: warm-start training using multiple sign embeddings with a hierarchical encoder and joint decoding. Stage 2: fine-tuning the model with a single sign embedding. 1: Warm-start training, where the parent model is trained using sign videos with multiple SEs (SE1, SE2, ..., SENSE ). Each SE extracts differ￾ent representations (F1, F2, ..., FNSE ) from the sign video (V), which ar… view at source ↗
Figure 3
Figure 3. The shared SLT encoders (left): sign repre [PITH_FULL_IMAGE:figures/full_fig_p004_3.png] view at source ↗

Discussion (0). Continue with ORCID to comment.

Forward citations

Cited by 1 Pith paper

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

  1. Developing Lightweight DNN Models With Limited Data For Real-Time Sign Language Recognition

    cs.CV 2025-06 conditional novelty 4.0 of 10

    A 7.2 MB branched DNN, fed with MediaPipe landmarks encoded as 947 ASL parameter features, classifies 343 isolated American Sign Language signs with 92% video-level accuracy and sub-10 ms latency on edge devices.

Reference graph

Works this paper leans on

49 extracted references · 17 canonical work pages · cited by 1 Pith paper

  1. [1]

    Daniel Andor, Chris Alberti, David Weiss, Aliaksei Severyn, Alessandro Presta, Kuzman Ganchev, Slav Petrov, and Michael Collins. 2016. https://doi.org/10.18653/v1/P16-1231 Globally normalized transition-based neural networks . In Proceedings of the 54th Annual Meeting of the Association for Computational Linguistics (Volume 1: Long Papers), pages 2442--24...

    work page doi:10.18653/v1/p16-1231 2016

  2. [2]

    Penny Boyes Braem and Rachel Sutton-Spence. 2001. The Hands Are The Head of The Mouth. The Mouth as Articulator in Sign Languages. Hamburg: Signum Press

    work page 2001

  3. [3]

    Necati Cihan Camgoz, Simon Hadfield, Oscar Koller, Hermann Ney, and Richard Bowden. 2018. Neural sign language translation. In Proceedings of the IEEE conference on computer vision and pattern recognition (CVPR), pages 7784--7793

    2018

  4. [4]

    Necati Cihan Camgoz, Oscar Koller, Simon Hadfield, and Richard Bowden. 2020. Sign Language Transformers : Joint end-to-end sign language recognition and translation. In Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR), pages 10023--10033

    2020

  5. [5]

    Yutong Chen, Fangyun Wei, Xiao Sun, Zhirong Wu, and Stephen Lin. 2022. A simple multi-modality transfer learning baseline for sign language translation. In Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR), pages 5120--5130

    work page 2022

  6. [6]

    Shun-Po Chuang, Yung-Sung Chuang, Chih-Chiang Chang, and Hung-yi Lee. 2021. https://doi.org/10.18653/v1/2021.findings-acl.92 Investigating the reordering capability in CTC -based non-autoregressive end-to-end speech translation . In Findings of the Association for Computational Linguistics: ACL-IJCNLP 2021, pages 1068--1077, Online. Association for Comput...

    doi:10.18653/v1/2021.findings-acl.92 2021

  7. [7]

    Runpeng Cui, Hu Liu, and Changshui Zhang. 2017. Recurrent convolutional neural networks for continuous sign language recognition by staged optimization. In Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR), pages 7361--7369

    work page 2017

  8. [8]

    Biao Fu, Peigen Ye, Liang Zhang, Pei Yu, Cong Hu, Xiaodong Shi, and Yidong Chen. 2023. https://doi.org/10.1109/ICASSP49357.2023.10095466 A token-level contrastive framework for sign language translation . In ICASSP 2023 - 2023 IEEE International Conference on Acoustics, Speech and Signal Processing (ICASSP), pages 1--5

    arXiv 2023

Show all 49 references
  1. [9]

    Jia Gong, Lin Geng Foo, Yixuan He, Hossein Rahmani, and Jun Liu. 2024. Llms are good sign language translators. In Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR), pages 18362--18372

    2024

  2. [10]

    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 International Conference on Machine Lea...

    2006

  3. [11]

    Barry Haddow, Rachel Bawden, Antonio Valerio Miceli Barone, Jindřich Helcl, and Alexandra Birch. 2022. https://doi.org/10.1162/coli_a_00446 Survey of Low-Resource Machine Translation . Computational Linguistics, 48(3):673--732

    2022 doi

  4. [12]

    Awni Hannun. 2017. Sequence modeling with CTC . Distill, 2(11):e8

    2017

  5. [13]

    Aiming Hao, Yuecong Min, and Xilin Chen. 2021. https://doi.org/10.1109/ICCV48922.2021.01111 Self-mutual distillation learning for continuous sign language recognition . In 2021 IEEE/CVF International Conference on Computer Vision (ICCV), pages 11283--11292

    2021

  6. [14]

    Adi Haviv, Lior Vassertail, and Omer Levy. 2021. https://doi.org/10.18653/v1/2021.naacl-main.209 Can latent alignments improve autoregressive machine translation? In Proceedings of the 2021 Conference of the North American Chapter of the Association for Computational Linguisti...

    2021 doi

  7. [15]

    Yosuke Higuchi, Keita Karube, Tetsuji Ogawa, and Tetsunori Kobayashi. 2022. Hierarchical conditional end-to-end ASR with CTC and multi-granular subword units. In ICASSP 2022-2022 IEEE International Conference on Acoustics, Speech and Signal Processing (ICASSP), pages 7797--7801. IEEE

    2022

  8. [16]

    Lianyu Hu, Liqing Gao, Zekang Liu, and Wei Feng. 2023. Continuous sign language recognition with correlation network. In Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR), pages 2529--2539

    2023

  9. [17]

    Tao Jin, Zhou Zhao, Meng Zhang, and Xingshan Zeng. 2022. Prior knowledge and memory enriched transformer for sign language translation. In Findings of the Association for Computational Linguistics: ACL 2022, pages 3766--3775

    2022

  10. [18]

    Oscar Koller, Necati Cihan Camgoz, Hermann Ney, and Richard Bowden. 2020. https://doi.org/10.1109/TPAMI.2019.2911077 Weakly supervised learning with multi-stream cnn-lstm-hmms to discover sequential parallelism in sign language videos . IEEE Transactions on Pattern Analysis an...

    2020

  11. [19]

    Kalpesh Krishna, Shubham Toshniwal, and Karen Livescu. 2018. Hierarchical multitask learning for CTC -based speech recognition. arXiv preprint arXiv:1807.06234

    2018 arXiv

  12. [20]

    Jind r ich Libovick \'y and Jind r ich Helcl. 2018. https://doi.org/10.18653/v1/D18-1336 End-to-end non-autoregressive neural machine translation with connectionist temporal classification . In Proceedings of the 2018 Conference on Empirical Methods in Natural Language Process...

    2018 doi

  13. [21]

    Chin-Yew Lin. 2004. ROUGE : A package for automatic evaluation of summaries. In Text summarization branches out, pages 74--81

    2004

  14. [22]

    Yuecong Min, Aiming Hao, Xiujuan Chai, and Xilin Chen. 2021. Visual alignment constraint for continuous sign language recognition. In Proceedings of the IEEE/CVF international conference on computer vision (ICCV), pages 11542--11551

    2021

  15. [23]

    Amit Moryossef, Kayo Yin, Graham Neubig, and Yoav Goldberg. 2021. Data augmentation for sign language gloss translation. arXiv preprint arXiv:2105.07476

    2021 arXiv

  16. [24]

    Mathias M \"u ller, Zifan Jiang, Amit Moryossef, Annette Rios, and Sarah Ebling. 2023. https://doi.org/10.18653/v1/2023.acl-short.60 Considerations for meaningful sign language machine translation based on glosses . In Proceedings of the 61st Annual Meeting of the Association ...

    2023 doi

  17. [25]

    Graham Neubig and Junjie Hu. 2018. https://doi.org/10.18653/v1/D18-1103 Rapid adaptation of neural machine translation to new languages . In Proceedings of the 2018 Conference on Empirical Methods in Natural Language Processing, pages 875--880, Brussels, Belgium. Association f...

    2018 doi

  18. [26]

    Jumon Nozaki and Tatsuya Komatsu. 2021. Relaxing the conditional independence assumption of CTC -based ASR by conditioning on intermediate predictions. arXiv preprint arXiv:2104.02724

    2021 arXiv

  19. [27]

    Kishore Papineni, Salim Roukos, Todd Ward, and Wei-Jing Zhu. 2002. BLEU : A method for automatic evaluation of machine translation. In Proceedings of the 40th annual meeting of the Association for Computational Linguistics, pages 311--318

    2002

  20. [28]

    Yifan Peng, Yui Sudo, Muhammad Shakeel, and Shinji Watanabe. 2024. https://aclanthology.org/2024.acl-long.549 OWSM - CTC : An open encoder-only speech foundation model for speech recognition, translation, and language identification . In Proceedings of the 62nd Annual Meeting ...

    2024

  21. [29]

    Razieh Rastgoo, Kourosh Kiani, Sergio Escalera, and Mohammad Sabokrou. 2021. Sign language production: A review. In Proceedings of the IEEE/CVF conference on computer vision and pattern recognition (CVPR), pages 3451--3461

    2021

  22. [30]

    Chitwan Saharia, William Chan, Saurabh Saxena, and Mohammad Norouzi. 2020. https://doi.org/10.18653/v1/2020.emnlp-main.83 Non-autoregressive machine translation with latent alignments . In Proceedings of the 2020 Conference on Empirical Methods in Natural Language Processing, ...

    2020 doi

  23. [31]

    Marcelo Sandoval-Castaneda, Yanhong Li, Diane Brentari, Karen Livescu, and Gregory Shakhnarovich. 2023. Self-supervised video transformers for isolated sign language recognition. arXiv preprint arXiv:2309.02450

    2023 arXiv

  24. [32]

    Rachel Sutton-Spence and Bencie Woll. 1999. The Linguistics of British Sign Language: An Introduction. Cambridge University Press

    1999

  25. [33]

    Sihan Tan, Taro Miyazaki, Katsutoshi Itoyama, and Kazuhiro Nakadai. 2024. https://aclanthology.org/2024.signlang-1.41 SEDA : Simple and effective data augmentation for sign language understanding . In Proceedings of the LREC-COLING 2024 11th Workshop on the Representation and ...

    2024

  26. [34]

    J \"o rg Tiedemann and Santhosh Thottingal. 2020. OPUS-MT — B uilding open translation services for the W orld. In Proceedings of the 22nd Annual Conferenec of the European Association for Machine Translation (EAMT), Lisbon, Portugal

    2020

  27. [35]

    Dave Uthus, Garrett Tanzer, and Manfred Georg. 2023. https://proceedings.neurips.cc/paper_files/paper/2023/file/5c61452daca5f0c260e683b317d13a3f-Paper-Datasets_and_Benchmarks.pdf YouTube-ASL : A large-scale, open-domain american sign language-english parallel corpus . In Advan...

    2023

  28. [36]

    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

  29. [37]

    Ryan Wong, Necati Cihan Camgoz, and Richard Bowden. 2024. https://openreview.net/forum?id=LqaEEs3UxU Sign2 GPT : Leveraging large language models for gloss-free sign language translation . In The Twelfth International Conference on Learning Representations

    2024

  30. [38]

    Brian Yan, Siddharth Dalmia, Yosuke Higuchi, Graham Neubig, Florian Metze, Alan W Black, and Shinji Watanabe. 2023. https://doi.org/10.18653/v1/2023.eacl-main.119 CTC alignments improve autoregressive translation . In Proceedings of the 17th Conference of the European Chapter ...

    2023 doi

  31. [39]

    Jinhui Ye, Wenxiang Jiao, Xing Wang, Zhaopeng Tu, and Hui Xiong. 2023. https://doi.org/10.18653/v1/2023.findings-emnlp.904 Cross-modality data augmentation for end-to-end sign language translation . In Findings of the Association for Computational Linguistics: EMNLP 2023, page...

    2023 doi

  32. [40]

    Kayo Yin, Amit Moryossef, Julie Hochgesang, Yoav Goldberg, and Malihe Alikhani. 2021. https://doi.org/10.18653/v1/2021.acl-long.570 Including signed languages in natural language processing . In Proceedings of the 59th Annual Meeting of the Association for Computational Lingui...

    2021 doi

  33. [41]

    Kayo Yin and Jesse Read. 2020. https://doi.org/10.18653/v1/2020.coling-main.525 Better sign language translation with STMC -transformer . In Proceedings of the 28th International Conference on Computational Linguistics, pages 5975--5989, Barcelona, Spain (Online). Internationa...

    2020 doi

  34. [42]

    Biao Zhang, Barry Haddow, and Rico Sennrich. 2022. Revisiting end-to-end speech-to-text translation from scratch. In International Conference on Machine Learning, pages 26193--26205. PMLR

    2022

  35. [43]

    Biao Zhang, Mathias M \"u ller, and Rico Sennrich. 2023. https://openreview.net/forum?id=EBS4C77p_5S SLTUNET : A simple unified model for sign language translation . In The Eleventh International Conference on Learning Representations

    2023

  36. [44]

    Hao Zhou, Wengang Zhou, Weizhen Qi, Junfu Pu, and Houqiang Li. 2021. Improving sign language translation with monolingual data by sign back-translation. In Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR), pages 1316--1325

    2021

  37. [45]

    Hao Zhou et al. 2022. https://doi.org/10.1109/TMM.2021.3059098 Spatial-temporal multi-cue network for sign language recognition and translation . IEEE Transactions on Multimedia, 24:768--779

    2022

  38. [46]

    Dele Zhu, Vera Czehmann, and Eleftherios Avramidis. 2023. https://doi.org/10.18653/v1/2023.acl-long.700 Neural machine translation methods for translating text to sign language glosses . In Proceedings of the 61st Annual Meeting of the Association for Computational Linguistics...

    2023 doi

  39. [47]

    Ronglai Zuo, Fangyun Wei, and Brian Mak. 2023. Natural language-assisted sign language recognition. In Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR), pages 14890--14900

    2023

  40. [48]

    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...

  41. [49]

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