Pith. sign in

REVIEW 4 major objections 4 minor 36 references

Open Universal Arabic ASR Leaderboard

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 The paper introduces the Open Universal Arabic ASR Leaderboard, a continuous five-dataset benchmark that ranks open Arabic speech models by average zero-shot WER.

desk verdict Useful, well-scoped Arabic ASR benchmark that deserves referee time, but the unquantified Common Voice leakage means the mid-table rankings shouldn't be treated as final. read the letter →

arxiv 2412.13788 v1 pith:VT2WJFRU submitted 2024-12-18 cs.CL

Yingzhi Wang , Anas Alhmoud , Muhammad Alqurishi This is my paper · ORCID
classification cs.CL
keywords ArabicASRmultidialectspeechrecognitionbenchmarkleaderboardzero-shotevaluationWordErrorRatespeakerrobustnessConformer-CTC
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 establishes a common reference for open-source Arabic speech recognition by building a multi-dialect benchmark leaderboard. It evaluates fourteen models zero-shot on five multi-dialect test sets—SADA, Common Voice 18.0, MASC test-clean, MASC test-noisy, and MGB-2—and ranks them by average Word Error Rate. The top-ranked system is the Conformer-CTC-large-Arabic model combined with a 4-gram language model, at 25.71% average WER, ahead of the Whisper large series and Seamless M4T. The same evaluation also measures robustness to noise and music, dialect-specific behavior, age and gender differences, and inference speed and memory, so the leaderboard doubles as a production-oriented survey.

What carries the argument

The named object is the Open Universal Arabic ASR Leaderboard, a continuously updated ranking built from a multi-dataset zero-shot evaluation protocol. Fourteen open models transcribe five exclusively multi-dialectal Arabic test sets without any adaptation, and a text-normalization pipeline removes punctuation and diacritics, normalizes Hamza and Madda, and unifies numerals before WER and CER are computed. The average WER across the five sets is the single ranking number that carries the paper's argument, while the per-dataset columns expose dialect-specific strengths and weaknesses.

What would settle it

Compute utterance-level text overlap between the Common Voice 18.0 test split and the training data of any evaluated model that was fine-tuned on earlier Common Voice versions; if there is substantial overlap, re-run the ranking without Common Voice and see whether the Conformer-CTC plus language model still holds first place.

Watch

Extended reading notes

Core claim

The paper's central claim is that a single continuous leaderboard can fairly rank general Arabic ASR models, and that when ranked this way the current open-source landscape has a clear order. The Conformer-CTC-large-Arabic model, decoded with its released 4-gram language model, leads with 25.71% average WER and 10.02% average CER across the five zero-shot test sets; greedy decoding places it second, still ahead of Whisper large-v3, Seamless M4T, and the Whisper large-v2/v1 variants. The authors find that performance tracks the amount of labeled Arabic training data more than model scale: the Conformer was trained on about 3,600 hours, the Whisper models on about 739 hours, and the lowest-ranked self-supervised models on under 100 hours. Across dialects, every model does best on Modern Standard Arabic and clearly worse on Egyptian and Khaliji speech, and most models show higher error rates for elderly and female speakers.

Load-bearing premise

The rankings are not materially distorted by training-data overlap with the evaluation sets, particularly Common Voice, whose test splits change between releases.

Editorial extensions

If this is right

  • Arabic ASR practitioners get a single yardstick: new open models can be added to the same five test sets and compared directly against the published ranking.
  • The leaderboard's ranking suggests that investing in labeled Arabic speech data (the Conformer's 3,600 hours) pays off more than relying on large multilingual pretraining (Whisper's 739 Arabic hours).
  • The consistent MSA-over-dialect gap indicates that dialect coverage in training data, not model architecture, is the main obstacle to universal Arabic ASR.
  • The efficiency measurements give a practical trade-off: self-supervised models decode fastest and use least memory, while the top-ranking Conformer-CTC is still small and fast; Whisper large models are several times slower.
  • Because the benchmark is designed as a continuous project, the same evaluation framework can absorb future models and datasets without changing the comparison protocol.

Reading between the lines

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

  • A natural stress test the paper does not run: drop Common Voice 18.0 from the average and recompute the ranking, since the paper itself flags possible test leakage from Common Voice's reshuffled splits; if the top order shifts, the benchmark's headline ranking is partly an artifact of that leakage.
  • The per-dataset columns show that a model like the HuBERT fine-tune is much stronger on MASC clean than on MGB-2, so a practitioner targeting one dialect should not rely on the average rank alone; the leaderboard can support dialect-specific subsets as the community adds them.
  • If the leaderboard is adopted as a standard, it would be cheap to add native dialect labels (already present in SADA) as a first-class column, turning the robustness analysis into a per-dialect benchmark rather than a single average.
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 Open Universal Arabic ASR Leaderboard, a continuous benchmark for open-source Arabic ASR models. It evaluates 14 models in zero-shot mode on five multi-dialect test sets (SADA, Common Voice 18.0, MASC test-clean, MASC test-noisy, and MGB-2), reporting WER/CER, average rankings, robustness across acoustic conditions and dialects, speaker age/gender analysis, and inference efficiency/memory. The authors claim this is the largest Arabic ASR benchmark to date and release evaluation code and a public leaderboard.

Significance. If the reported rankings are trustworthy, this would be a valuable community resource: it covers most currently available open Arabic ASR models, uses multiple multi-dialect datasets, and adds practical efficiency and robustness dimensions that most prior Arabic ASR comparisons lack. The paper ships open-source evaluation code, states a clear text-normalization recipe, and documents model sources. The main risk is training/test overlap with Common Voice, which the paper acknowledges in Section 2.2 and Section 3.5 but does not quantify; the suspiciously low CV-18.0 WER of one model suggests leakage and could change mid-table rankings.

major comments (4)
  1. [Section 2.2, Section 3.5, Table 1] The claim that Common Voice split shuffling 'will not significantly impact the overall benchmark' is not supported by evidence. In Table 1, asafaya/hubert-large-arabic-transcribe achieves 8.01% WER on CV-18.0, versus 32.94% on MASC clean, 50.16% on MASC noisy, 37.51% on MGB-2, and 67.82% on SADA; excluding CV-18.0 raises its average WER from 39.29% to about 47.1%, dropping it below whisper-medium. whitefox123/w2v-bert-2.0-ft also shows a disproportionately low CV-18.0 WER (24.18% vs 35.93% on MASC clean and 78.02% on SADA). Because the paper does not report the training data for these community models, leakage cannot be ruled out. Please quantify the overlap (e.g., deduplicate CV-18.0 test against previous CV releases used in training, or report the training corpus compositions) and show the leaderboard with and without CV-18.0, or flag CV-18.0 results for CV-trained models as potentially biased.
  2. [Section 2.2, Section 3.5, Table 1] MGB-2 is stated to be over 70% MSA, yet the Average WER gives MGB-2 the same weight as the other four test sets. This tilts the 'general' ranking toward MSA-dominated performance and away from a balanced multi-dialect assessment. Please report the dialect distribution of every test set and provide a sensitivity analysis (e.g., per-dialect WERs, dialect-weighted averaging, or one average excluding MGB-2) to justify the unweighted five-way average as a 'general' performance measure.
  3. [Section 2.1, Table 1] For Seamless M4T, the authors run three separate inferences with different Arabic dialect configuration tags and report only the best result. This is a model-specific selection that is not applied to any other model in the benchmark, so Seamless's rank-4 position is not directly comparable with the zero-shot scores of the other models. Please report the WER for all three configurations or justify a fixed, principled configuration for leaderboard reporting.
  4. [Table 1] The leaderboard ranks are point estimates with no confidence intervals or significance tests. Several adjacent ranks differ by less than 1% WER (e.g., ranks 5 and 6: 33.30% vs 34.04%; ranks 8 and 9: 39.29% vs 39.60%), so the stability of the ordering is unclear. Please add bootstrap confidence intervals or pairwise significance tests for at least the top and mid-table comparisons, or explicitly state that adjacent small differences should not be interpreted as reliable ordering.
minor comments (4)
  1. [Table 5] The column header '#Mem model' appears twice; based on the caption, the second column should be labeled '#Mem infer' to distinguish model-loading memory from per-sample inference memory.
  2. [Table 1] Table 1 uses commas as decimal separators (e.g., '25,71%') while the text uses periods (e.g., '39.29%'); please make the decimal separator consistent throughout.
  3. [Section 3.5] The HuggingFace and GitHub URLs for the leaderboard and code should be checked for line-break corruption in the PDF; the GitHub URL in the text appears to be broken across lines.
  4. [General] The phrase 'largest Arabic ASR benchmark to date' in Sections 1 and 4 is used without defining 'largest' (number of models, number of datasets, or total test hours); please state the criterion explicitly.

Circularity Check

0 steps flagged · score 0.0 of 10

No circularity: the paper reports directly measured benchmark scores; the acknowledged Common Voice overlap is a data-contamination risk, not a circular derivation.

full rationale

This is an evaluation and benchmarking paper, not a derivation. The leaderboard scores are obtained by running zero-shot inference with existing open-source ASR models on public multi-dialect test sets and then computing WER/CER with a stated normalization recipe. No parameter is fitted to the benchmark outputs, and no quantity that the paper claims to predict is defined in terms of the benchmark scores themselves. The ranking is a direct aggregation of measured errors, so there is no derivation chain that could reduce to its own inputs. The paper does disclose a potential training/test overlap for Common Voice, noting that 'Common Voice shuffles the dataset splits from one release to another, the results may be biased for those models trained on previous versions due to the test data leakage,' and it asserts without a quantitative check that this 'will not significantly impact the overall benchmark.' That is a legitimate validity concern about possible data contamination, and it could affect the reliability of some reported numbers, but it is not circularity: the benchmark scores are not constructed from the models' training data, and the paper does not use the overlap as evidence for any derived claim. Similarly, reporting only the best Seamless M4T result across three dialect configurations is a clearly stated reporting convention, not a fitted parameter disguised as a prediction. The references are external datasets, toolkits, and prior model papers; there is no load-bearing self-citation, no imported uniqueness theorem, and no renaming of a known result under new coordinates. Accordingly, no circular step is present and the appropriate score is 0.

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

No numerical parameters are fitted in this paper. The benchmark relies on the assumed validity of its evaluation design and dataset choices.

assumptions (3)
  • domain assumption Multi-dialect datasets are sufficient to evaluate generalization across Arabic dialects.
    The paper selects only multi-dialect datasets and argues this is methodologically superior; this is a design assumption, not proven.
  • domain assumption Average WER across the five test sets is a valid primary ranking metric.
    The leaderboard ranks by unweighted average WER, implicitly treating all test sets as equally important and ignoring sample-size and dialect-distribution differences.
  • domain assumption Character normalization following [31] preserves dialect information while enabling fair comparison.
    The normalization choices affect WER/CER values; if too aggressive or inconsistent, rankings could change.

how reviews work

0 comments
Cite this review

Pith. "Pith review of Open Universal Arabic ASR Leaderboard." pith.science (2026). https://pith.science/paper/VT2WJFRU

@misc{pith2026241213788,
  author       = {Pith},
  title        = {Pith review of: Open Universal Arabic ASR Leaderboard},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/VT2WJFRU}},
  note         = {Machine review of arXiv:2412.13788}
}
read the original abstract

In recent years, the enhanced capabilities of ASR models and the emergence of multi-dialect datasets have increasingly pushed Arabic ASR model development toward an all-dialect-in-one direction. This trend highlights the need for benchmarking studies that evaluate model performance on multiple dialects, providing the community with insights into models' generalization capabilities. In this paper, we introduce Open Universal Arabic ASR Leaderboard, a continuous benchmark project for open-source general Arabic ASR models across various multi-dialect datasets. We also provide a comprehensive analysis of the model's robustness, speaker adaptation, inference efficiency, and memory consumption. This work aims to offer the Arabic ASR community a reference for models' general performance and also establish a common evaluation framework for multi-dialectal Arabic ASR models.

Discussion (0). Continue with ORCID to comment.

Reference graph

Works this paper leans on

36 extracted references · 26 canonical work pages

  1. [1]

    Introduction Arabic Automatic Speech Recognition (ASR) research faces unique challenges compared to other languages like English . The complexity of Arabic, with its rich morphology, dialec- tal variations, and lack of diacritics in written text, pose s dif- ficulties for both data collection and model development [1, 2]. Particularly regarding dialects, w...

    work page

  2. [2]

    Experimental Settings We conducted zero-shot inference for a diverse set of popu- lar open-source Arabic ASR models on a collection of multi- dialectal Arabic datasets. 2.1. Models To facilitate the collection of state-of-the-art Arabic AS R mod- els, we divided them into the following four main categories . Whisper: The OpenAI’s Whisper series of models ...

    work page

  3. [3]

    Overall Leaderboard The overall leaderboard is shown in Table 1

    Results and Analysis 3.1. Overall Leaderboard The overall leaderboard is shown in Table 1. For all the col- lected models, we performed zero-shot inference across five dif- ferent multi-dialect test sets. We then averaged the WER and CER over the test sets and ranked the models based on the aver- age WER. As can be seen from the results, Nvidia’s Conformer...

    work page

  4. [4]

    Conclusions In this work, we implement the most comprehensive Arabic ASR benchmark to date, which includes an Open Universal Arabic ASR Leaderboard and a thorough analysis covering ro- bustness, speaker adaptation, efficiency, and resource uti liza- tion. As a continuous effort, our work aims to provide the com - munity with the latest references on model ...

    work page

  5. [5]

    Masc: Massive arabic speech corpus,

    M. Al-Fetyani, M. Al-Barham, G. Abandah, A. Alsharkawi, and M. Dawas, “Masc: Massive arabic speech corpus,” in 2022 IEEE Spoken Language Technology W orkshop (SLT). IEEE, 2023, pp. 1006–1013

    work page 2022

  6. [6]

    These datasets offer rich linguistic re- sources that capture the nuances of regional dialects, layi ng the groundwork for multi-dialectal ASR model training

    includes over 2000 hours of speech samples collected fro m diverse sources, including broadcast media and conversati onal contexts, and provides labeled data for both MSA and promi- nent Arabic dialects. These datasets offer rich linguistic re- sources that capture the nuances of regional dialects, layi ng the groundwork for multi-dialectal ASR model trai...

    work page 2000

  7. [7]

    Arabic automatic speech recognition: a syste m- atic literature review,

    A. Dhouib, A. Othman, O. El Ghoul, M. K. Khribi, and A. Al Sinani, “Arabic automatic speech recognition: a syste m- atic literature review,” Applied Sciences, vol. 12, no. 17, p. 8898, 2022

    work page 2022

  8. [8]

    End-to-end arabic speech recognition: A review,

    A. A. Abdelhamid, H. A. Alsayadi, I. Hegazy, and Z. T. Faye d, “End-to-end arabic speech recognition: A review,” in Proceedings of the 19th Conference of Language Engineering (ESOLEC’19) , Alexandria, Egypt, 2020, pp. 26–30

    work page 2020

Show all 36 references
  1. [9]

    The mgb-2 challenge: Arabic multi-dialect broad cast media recognition,

    A. Ali, P . Bell, J. Glass, Y . Messaoui, H. Mubarak, S. Rena ls, and Y . Zhang, “The mgb-2 challenge: Arabic multi-dialect broad cast media recognition,” in 2016 IEEE Spoken Language Technology W orkshop (SLT). IEEE, 2016, pp. 279–284

    2016

  2. [10]

    Sada: Saudi audio dataset for arabic,

    S. Alharbi, A. Alowisheq, Z. T ¨ uske, K. Darwish, A. Alraj eh, A. Alrowithi, A. B. Tamran, A. Ibrahim, R. Aloraini, R. Alnaj im et al. , “Sada: Saudi audio dataset for arabic,” in ICASSP 2024- 2024 IEEE International Conference on Acoustics, Speech an d Signal Processing (ICA...

    2024

  3. [11]

    Qasr : Qcri aljazeera speech resource–a large scale annotated ara bic speech corpus,

    H. Mubarak, A. Hussein, S. A. Chowdhury, and A. Ali, “Qasr : Qcri aljazeera speech resource–a large scale annotated ara bic speech corpus,” arXiv preprint arXiv:2106.13000, 2021

    2021 arXiv

  4. [12]

    Casablanca: Data and models for multidialec- tal arabic speech recognition,

    B. Talafha, K. Kadaoui, S. M. Magdy, M. Habiboullah, C. M. Chafei, A. O. El-Shangiti, H. Zayed, R. Alhamouri, R. Assi, A. Alraeesi et al., “Casablanca: Data and models for multidialec- tal arabic speech recognition,” arXiv preprint arXiv:2410.04527 , 2024

    arXiv

  5. [13]

    Data2vec: A general framework for self-supervised learni ng in speech, vision and language,

    A. Baevski, W.-N. Hsu, Q. Xu, A. Babu, J. Gu, and M. Auli, “Data2vec: A general framework for self-supervised learni ng in speech, vision and language,” in International Conference on Ma- chine Learning. PMLR, 2022, pp. 1298–1312

    2022

  6. [14]

    wav2vec 2.0: A framework for self-supervised learning of speech rep re- sentations,

    A. Baevski, Y . Zhou, A. Mohamed, and M. Auli, “wav2vec 2.0: A framework for self-supervised learning of speech rep re- sentations,” Advances in neural information processing systems , vol. 33, pp. 12 449–12 460, 2020

    2020

  7. [15]

    Hubert: Self-supervised speech repre sen- tation learning by masked prediction of hidden units,

    W.-N. Hsu, B. Bolte, Y .-H. H. Tsai, K. Lakhotia, R. Salak hutdi- nov, and A. Mohamed, “Hubert: Self-supervised speech repre sen- tation learning by masked prediction of hidden units,” IEEE/ACM transactions on audio, speech, and language processing , vol. 29, pp. 3451–3460, 2021

    2021

  8. [16]

    Conformer: Convolution- augmented transformer for speech recognition,

    A. Gulati, J. Qin, C.-C. Chiu, N. Parmar, Y . Zhang, J. Y u, W. Han, S. Wang, Z. Zhang, Y . Wu et al. , “Conformer: Convolution- augmented transformer for speech recognition,” arXiv preprint arXiv:2005.08100, 2020

    2005 arXiv

  9. [17]

    Robust speech recognition via large-scale w eak supervision,

    A. Radford, J. W. Kim, T. Xu, G. Brockman, C. McLeavey, an d I. Sutskever, “Robust speech recognition via large-scale w eak supervision,” in International conference on machine learning . PMLR, 2023, pp. 28 492–28 518

    2023

  10. [18]

    Huggingface’s transformers: State-of-the- art natural language processing,

    T. Wolf, “Huggingface’s transformers: State-of-the- art natural language processing,” arXiv preprint arXiv:1910.03771, 2019

    1910 arXiv

  11. [19]

    Nemo: a toolkit for building ai applications using neural modules ,

    O. Kuchaiev, J. Li, H. Nguyen, O. Hrinchuk, R. Leary, B. G ins- burg, S. Kriman, S. Beliaev, V . Lavrukhin, J. Cook et al., “Nemo: a toolkit for building ai applications using neural modules ,” arXiv preprint arXiv:1909.09577, 2019

    1909 arXiv

  12. [20]

    Open-source conversational ai with speechbrain 1.0,

    M. Ravanelli, T. Parcollet, A. Moumen, S. de Langen, C. S ubakan, P . Plantinga, Y . Wang, P . Mousavi, L. Della Libera, A. Ploujnikov et al., “Open-source conversational ai with speechbrain 1.0,” Jour- nal of Machine Learning Research , vol. 25, no. 333, pp. 1–11, 2024

    2024

  13. [21]

    N-shot ben ch- marking of whisper on diverse arabic speech recognition,

    B. Talafha, A. Waheed, and M. Abdul-Mageed, “N-shot ben ch- marking of whisper on diverse arabic speech recognition,” arXiv preprint arXiv:2306.02902, 2023

    2023 arXiv

  14. [22]

    Unsupervised cross-lingual representation learning for speech recognition,

    A. Conneau, A. Baevski, R. Collobert, A. Mohamed, and M. Auli, “Unsupervised cross-lingual representation learning for speech recognition,” arXiv preprint arXiv:2006.13979, 2020

    2006 arXiv

  15. [23]

    Common voice: A massively-multilingual speech corpus,

    R. Ardila, M. Branson, K. Davis, M. Henretty, M. Kohler, J. Meyer, R. Morais, L. Saunders, F. M. Tyers, and G. Weber, “Common voice: A massively-multilingual speech corpus,” arXiv preprint arXiv:1912.06670, 2019

    1912 arXiv

  16. [24]

    Speech recognition cha llenge in the wild: Arabic mgb-3,

    A. Ali, S. V ogel, and S. Renals, “Speech recognition cha llenge in the wild: Arabic mgb-3,” in 2017 IEEE Automatic Speech Recog- nition and Understanding W orkshop (ASRU) . IEEE, 2017, pp. 316–322

    2017

  17. [25]

    The mgb-5 challenge: Recogni- tion and dialect identification of dialectal arabic speech,

    A. Ali, S. Shon, Y . Samih, H. Mubarak, A. Abdelali, J. Gla ss, S. Renals, and K. Choukri, “The mgb-5 challenge: Recogni- tion and dialect identification of dialectal arabic speech, ” in 2019 IEEE Automatic Speech Recognition and Understanding W ork- shop (ASRU). IEEE, 2019, pp...

    2019

  18. [26]

    Fleurs: Few-shot learnin g evaluation of universal representations of speech,

    A. Conneau, M. Ma, S. Khanuja, Y . Zhang, V . Axelrod, S. Dalmia, J. Riesa, C. Rivera, and A. Bapna, “Fleurs: Few-shot learnin g evaluation of universal representations of speech,” in 2022 IEEE Spoken Language Technology W orkshop (SLT). IEEE, 2023, pp. 798–805

    2022

  19. [27]

    Larabench: Benchmarking arabic ai with large language models,

    A. Abdelali, H. Mubarak, S. Chowdhury, M. Hasanain, B. M ousi, S. Boughorbel, S. Abdaljalil, Y . El Kheir, D. Izham, F. Dalvi et al. , “Larabench: Benchmarking arabic ai with large language models,” in Proceedings of the 18th Conference of the European Chapter of the Associati...

    2024

  20. [28]

    Open automatic speech recognition leaderboard,

    V . Srivastav, S. Majumdar, N. Koluguri, A. Moumen, S. Ga ndhi et al. , “Open automatic speech recognition leaderboard,” https: //huggingface.co/spaces/hf-audio/open asr leaderboard, 2023

    2023

  21. [29]

    Con- nectionist temporal classification: labelling unsegmente d se- quence data with recurrent neural networks,

    A. Graves, S. Fern´ andez, F. Gomez, and J. Schmidhuber, “Con- nectionist temporal classification: labelling unsegmente d se- quence data with recurrent neural networks,” in Proceedings of the 23rd international conference on Machine learning, 2006, pp. 369–376

    2006

  22. [30]

    Sequence transduction with recurrent neur al net- works,

    A. Graves, “Sequence transduction with recurrent neur al net- works,” arXiv preprint arXiv:1211.3711, 2012

    2012 arXiv

  23. [31]

    Stateful conformer with cache-based inference for stream ing au- tomatic speech recognition,

    V . Noroozi, S. Majumdar, A. Kumar, J. Balam, and B. Ginsb urg, “Stateful conformer with cache-based inference for stream ing au- tomatic speech recognition,” in ICASSP 2024-2024 IEEE Inter- national Conference on Acoustics, Speech and Signal Proces sing (ICASSP). IEEE, 2024, ...

    2024

  24. [32]

    Arabic speech corpus,

    N. Halabi et al., “Arabic speech corpus,” Oxford Text Archive Core Collection, 2016

    2016

  25. [33]

    Scaling speech technology to 1,000+ languages,

    V . Pratap, A. Tjandra, B. Shi, P . Tomasello, A. Babu, S. K undu, A. Elkahky, Z. Ni, A. Vyas, M. Fazel-Zarandi et al. , “Scaling speech technology to 1,000+ languages,” Journal of Machine Learning Research, vol. 25, no. 97, pp. 1–52, 2024

    2024

  26. [34]

    W2v-bert: Combining contrastive learning and maske d language modeling for self-supervised speech pre-trainin g,

    Y .-A. Chung, Y . Zhang, W. Han, C.-C. Chiu, J. Qin, R. Pang , and Y . Wu, “W2v-bert: Combining contrastive learning and maske d language modeling for self-supervised speech pre-trainin g,” in 2021 IEEE Automatic Speech Recognition and Understanding W orkshop (ASRU). IEEE, 202...

    2021

  27. [35]

    Seamlessm4t-massively multilingual & multimodal ma- chine translation,

    L. Barrault, Y .-A. Chung, M. C. Meglioli, D. Dale, N. Don g, P .- A. Duquenne, H. Elsahar, H. Gong, K. Heffernan, J. Hoffman et al. , “Seamlessm4t-massively multilingual & multimodal ma- chine translation,” arXiv preprint arXiv:2308.11596, 2023

    2023 arXiv

  28. [36]

    To wards one model to rule all: Multilingual strategy for dialectal c ode- switching arabic asr,

    S. A. Chowdhury, A. Hussein, A. Abdelali, and A. Ali, “To wards one model to rule all: Multilingual strategy for dialectal c ode- switching arabic asr,” arXiv preprint arXiv:2105.14779, 2021

    2021 arXiv

Pith tools

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