Pith. sign in

REVIEW 4 major objections 5 minor 47 references

Streaming Speaker Change Detection and Gender Classification for Transducer-Based Multi-Talker Speech Translation

T0 review · 4 major / 5 minor · reviewed 2026-08-09 · deepseek-v4-flash

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

Pith's one-line read A streaming translation model can detect speaker changes and gender from token-level speaker embeddings, with no retraining of the translation model.

desk verdict A plausible, modest extension of the authors' own t-vector work, but the evaluation dodges the real streaming multi-talker scenario and the abstract overclaims 'high accuracy.' read the letter →

arxiv 2502.02683 v1 pith:BSQA6OLV submitted 2025-02-04 cs.SD cs.AIcs.CLeess.AS

Peidong Wang , Naoyuki Kanda , Jian Xue , Jinyu Li , Xiaofei Wang , Aswin Shanmugam Subramanian , Junkun Chen , Sunit Sivasankaran
show 2 more authors
Xiong Xiao Yong Zhao
This is my paper · ORCID
classification cs.SDcs.AIcs.CLeess.AS
keywords streamingspeechtranslationspeakerchangedetectiongenderclassificationt-vectortransducermulti-talkertext-to-speechtoken-levelembedding
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 is trying to establish that the same streaming, transducer-based speech translation model that turns English audio into translated text can also tell when the speaker changes and what gender the current speaker is, without retraining the translation model. It does this by attaching a t-vector module that emits a speaker embedding for every non-blank output token, then treating both tasks as cosine-similarity comparisons in that embedding space. On real recorded conversations the method reaches speaker change detection F1 scores between 0.61 and 0.68 across five target languages, with a tunable threshold trading recall against precision, and token-level gender classification accuracy of 0.989. The payoff is a single streaming system whose translated text, speaker change marks, and gender labels are naturally synchronized, ready to drive a downstream text-to-speech module.

What carries the argument

The central object is the t-vector, a token-level speaker embedding emitted for every non-blank output token while the model streams. It is produced by a speaker encoder that attends to the frozen translation encoder's hidden states and a two-layer recurrent decoder that also reads the current output-token embedding. It carries the argument because both downstream tasks reduce to cosine comparisons in one 128-dimensional embedding space: adjacent-token cosine similarity marks a speaker change when it drops below a threshold, and profile cosine similarity assigns a gender to each token.

What would settle it

Run the method on continuous multi-talker recordings that contain consecutive speaker changes within a few seconds of one another, and compare predicted change times against manual labels; if F1 on adjacent turns drops well below the reported 0.61-0.68 range, the single-change test clips were masking a real failure mode. A second check is to add noise at increasing levels to the same clips and observe whether precision collapses as the paper's own observation that most errors land on noise and silence would predict.

Watch

Extended reading notes

Core claim

A well-trained transducer-based streaming multilingual speech translation model can be extended with a t-vector speaker encoder and a two-layer recurrent speaker decoder, trained on VoxCeleb alignments while all translation parameters stay frozen, so that every output token carries a 128-dimensional speaker embedding. Speaker changes are then detected by thresholding the cosine similarity between adjacent t-vectors, and gender is classified by comparing each t-vector with the average speaker vectors of male and female profiles. The reported experiments show speaker change detection F1 of 0.68, 0.66, and 0.61 at thresholds 0.99, 0.94, and 0.89, and gender accuracy of 0.989 that is nearly identical across all five language pairs. Because the embeddings live in token space, the speaker change marks and timestamps are synchronized with the translated text without any separate diarization or alignment step.

Load-bearing premise

The load-bearing premise is that a speaker-embedding model trained on clean single-speaker VoxCeleb audio transfers to noisy, conversational multi-talker audio, so that cosine similarity between adjacent token-level speaker vectors reliably marks real speaker turns even at noisy and silent points.

Editorial extensions

If this is right

  • A streaming speech-to-speech pipeline can receive translation text, speaker change markers, and gender labels from one model, with latency governed by the 1-second chunked encoder rather than by an offline diarization pass.
  • Because the ST model is frozen during t-vector training, translation quality is preserved while the auxiliary speaker information is added.
  • The single cosine threshold gives system builders a direct control on the operating point, favoring recall at high thresholds or precision at low thresholds.
  • The same token-space machinery should transfer from English-to-many to many-to-English translation, since the speaker embeddings do not depend on the target language.
  • The offline baselines compared in the paper, Whisper and EEND, can reach similar F1 on the same clips but require an extra alignment step to connect detected speaker changes to the translated text, which this method avoids.

Reading between the lines

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

  • The test set was constructed from real recordings segmented so each clip contains exactly one speaker change, so behavior on consecutive speaker turns inside a continuous stream is unmeasured; a natural extension is to score the method on unsegmented multi-talker audio with several turns per clip.
  • The frozen speaker encoder is trained on clean single-speaker VoxCeleb audio, and the paper reports that most wrong detections occur at noisy or silent points; adding noise-robust training or threshold smoothing is a direct, testable improvement path.
  • Gender classification here is binary and anchored to two profile centroids; the same embedding comparisons could be repurposed to select among many TTS voice profiles by nearest-neighbor distance rather than by binary gender.
  • The cosine-gap detector assumes each token belongs to one speaker; overlapping speech or very short interjections would likely blur adjacent t-vectors, so the method will probably need a voice-activity or overlap gate before it can be deployed in fully conversational settings.
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 / 5 minor

Summary. The paper addresses streaming speaker change detection and gender classification for transducer-based end-to-end speech translation. The authors append a t-vector module to a fixed streaming multilingual Transformer transducer; at inference, cosine similarity between adjacent t-vectors is thresholded to insert speaker-change tokens, and gender is assigned by comparing token t-vectors with male and female profile vectors. On five English-to-X language pairs, the method reports speaker-change detection F1 scores of 0.61-0.68 depending on the threshold (Table 1) and token-level gender classification accuracy of 0.989 on the VoxCeleb-1 test set (Table 2), while claiming that the fixed ST model is not degraded.

Significance. The idea of producing speaker-change and gender metadata in token space from a streaming transducer is relevant and the paper's core mechanism is plausible. The strengths include keeping the ST model fixed, producing text-synchronized labels that are directly usable by downstream TTS, and providing a comparison with offline Whisper and EEND baselines. If the evaluation covered the actual streaming multi-talker condition, the contribution would be solid. However, the headline numbers are currently obtained under simplified conditions: each speaker-change clip contains exactly one change, gender is tested only on single-speaker VoxCeleb audio, and the detection threshold is swept rather than fixed by a validation rule. These limitations prevent me from treating the results as established evidence for the stated 'high accuracy' in streaming multi-talker speech translation.

major comments (4)
  1. [Section 3.1.2 and Section 4.1, Table 1] The speaker-change test set is constructed by concatenating segments into 688 audio samples, each containing exactly one speaker change. This does not exercise the streaming multi-talker scenario that is central to the paper: the detector is never required to track consecutive speaker turns, so missed changes in a multi-turn stream and spurious insertions within a turn are not measured in the way they would occur in continuous use. I request an evaluation on continuous recordings with multiple speaker changes per clip, with metrics computed over the full stream and, ideally, a per-turn analysis.
  2. [Table 1, Section 4.1] Results are reported for three cosine-similarity thresholds (0.99, 0.94, 0.89), with no validation-based procedure for selecting the threshold and no confidence intervals. Because the threshold is swept and the operating point is chosen ex post, the reported F1 range may overstate expected performance on new data. Please fix the threshold using a separate validation set, or report a calibration curve with a pre-specified selection rule, and provide error bars or significance tests over the 688 clips. The choice of the +/-2s tolerance window should also be justified or varied.
  3. [Section 4.2, Table 2] Gender classification is evaluated on the VoxCeleb-1 test set, which contains single-speaker utterances and does not include the conversational, noisy, multi-talker audio of the target application. The 0.989 accuracy therefore does not establish gender classification quality for the speaker-change segments in the collected conversational test data. Please report gender accuracy on the 688 conversational clips, or on another multi-talker test set, at least for utterances around speaker changes.
  4. [Section 4.1, error analysis] The paper notes that most incorrect speaker-change detections occur at noisy or silent points, which are exactly the conditions where t-vectors trained on VoxCeleb are least reliable. This is a stated limitation, but no corresponding analysis or mitigation is provided. I would like an error analysis conditioned on noise, silence, and overlapping speech, or at least a discussion of how the method is expected to behave on the overlapping-speech portions of real conversations, since that is where multi-talker ST is most challenging.
minor comments (5)
  1. [Section 3.1.1] The language label 'Germany (DE)' should be 'German (DE)'.
  2. [Throughout] There are repeated typographical issues such as 'V oxceleb' instead of 'VoxCeleb'; please correct these and check the reference list for similar formatting errors.
  3. [Section 4.1, Table 1] The comparison with Whisper and EEND is under-specified: please state explicitly whether the same 688 clips and the same reference tolerance were used, and how the offline systems' speaker-change outputs were aligned to the reference timestamps.
  4. [Section 4.2, Table 2] Please define the denominator for 'token-level' gender accuracy, especially how punctuation tokens are treated, and clarify whether decisions are made per token independently or per utterance with a segment-level aggregation.
  5. [Abstract and Section 5] The claim of 'high accuracy' is stronger than what Table 1 supports at the lower threshold (F1 = 0.61); consider qualifying the claim with the operating point or reporting a single threshold-selection policy.

Circularity Check

0 steps flagged · score 0.0 of 10

No significant circularity: the speaker-change and gender results are measured on held-out data, and the threshold sweep is an operating-point trade-off rather than a fitted prediction.

full rationale

The paper's derivation chain is self-contained and non-circular. Speaker change detection is implemented as cosine similarity between adjacent t-vectors compared with a threshold, and the reported F1 scores are measured on 688 held-out conversational clips with human-annotated speaker-change timestamps. The three thresholds (0.99, 0.94, 0.89) are presented as a precision-recall operating-point sweep, not as parameters fitted to the test labels, so the result is not forced by construction. Gender classification is evaluated on the Voxceleb-1 test set using gender profiles derived from Voxceleb training data; the reported 0.989 accuracy is an external held-out measurement and is not entailed by the cross-entropy speaker-ID training objective. The t-vector component is cited from prior work (Kanda et al., 2022b) and used as an architectural module, but it is re-initialized and evaluated here; no uniqueness theorem or load-bearing self-citation is invoked to forbid alternatives. The evaluation limitations noted in the paper (single speaker change per test clip, errors at noisy or silent points, gender tested only on single-speaker Voxceleb audio) concern external validity and evaluation scope, not circularity. Therefore no circular step can be identified by the paper's own equations or citations.

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

The central experiment relies on a hand-chosen detection threshold and an evaluation tolerance window, plus a set of transfer assumptions about the t-vector model generalizing from clean single-speaker data to noisy conversations. No new physical or formal entities are introduced.

free parameters (2)
  • cosine similarity threshold for speaker change = 0.99, 0.94, 0.89
    A speaker change is declared when the cosine similarity between adjacent t-vectors falls below this threshold. The paper reports results for three values with no validation-based selection, so the operating point is chosen by hand.
  • evaluation tolerance window = +/-2 seconds
    A detected speaker change is counted correct if it falls within 2 seconds of the reference timestamp; this choice directly affects the reported recall, precision, and F1.
assumptions (4)
  • domain assumption The ST model's token sequence and Viterbi alignment correctly associate each t-vector with its corresponding speech frames
    Section 3.1.1 states the t-vector training stage uses Viterbi alignment to associate t-vectors with non-blank tokens; if alignment is wrong, the t-vector is not tied to the actual speaker of that audio region.
  • domain assumption A t-vector model trained on single-speaker Voxceleb audio generalizes to noisy multi-talker conversational audio
    Section 3.1.1 trains the t-vector model on Voxceleb; Section 4.1 reports that most speaker change errors occur at noisy or silent points, showing the generalization is imperfect.
  • domain assumption A fixed cosine threshold is valid across output languages and acoustic conditions
    Table 1 applies the same thresholds to all five language pairs; the paper provides no per-language threshold adaptation or analysis of acoustic variation.
  • domain assumption The pre-trained Res2Net SID module remains speaker-discriminative when the ST encoder is used as external attention value
    Section 3.2 says the SID module is initialized with a pre-trained model and remains fixed; if the mismatch between ST encoder and SID features is large, t-vectors would not carry speaker identity.

how reviews work

0 comments
Cite this review

Pith. "Pith review of Streaming Speaker Change Detection and Gender Classification for Transducer-Based Multi-Talker Speech Translation." pith.science (2026). https://pith.science/paper/BSQA6OLV

@misc{pith2026250202683,
  author       = {Pith},
  title        = {Pith review of: Streaming Speaker Change Detection and Gender Classification for Transducer-Based Multi-Talker Speech Translation},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/BSQA6OLV}},
  note         = {Machine review of arXiv:2502.02683}
}
read the original abstract

Streaming multi-talker speech translation is a task that involves not only generating accurate and fluent translations with low latency but also recognizing when a speaker change occurs and what the speaker's gender is. Speaker change information can be used to create audio prompts for a zero-shot text-to-speech system, and gender can help to select speaker profiles in a conventional text-to-speech model. We propose to tackle streaming speaker change detection and gender classification by incorporating speaker embeddings into a transducer-based streaming end-to-end speech translation model. Our experiments demonstrate that the proposed methods can achieve high accuracy for both speaker change detection and gender classification.

Figures

Figures reproduced from arXiv: 2502.02683 by the authors.

Figure 1
Figure 1. Illustration of Transformer transducer for ST. [PITH_FULL_IMAGE:figures/full_fig_p002_1.png] view at source ↗
Figure 4
Figure 4. Illustration of t-vector model for ST. d-vector, which is obtained from audio segments. A t-vector module is typically added to a transducer model that has been well-trained. It was first devel￾oped for multi-talker ASR and has been recently used for ST tasks (Yang et al., 2023). The t-vector model, which is built on top of an ST model, is illustrated in [PITH_FULL_IMAGE:figures/full_fig_p003_4.png] view at source ↗
Figure 2
Figure 2. Illustration of the reception field of a stream [PITH_FULL_IMAGE:figures/full_fig_p003_2.png] view at source ↗
Figures from the paper (3 more)
Figure 3
Figure 3. Figure 3: Illustration of LAMASSU-UNI. 2.2 t-vector for ST A t-vector is a type of speaker embedding vector that captures the speaker characteristics at the to￾ken level (Kanda et al., 2022b). It is based on the [PITH_FULL_IMAGE:figures/full_fig_p003_3.png]
Figure 6
Figure 6. Figure 6: Illustration of speaker change detection using t-vectors. [PITH_FULL_IMAGE:figures/full_fig_p005_6.png]
Figure 7
Figure 7. Figure 7: Illustration of gender classification using t-vectors. See the caption of Figure [PITH_FULL_IMAGE:figures/full_fig_p006_7.png]

Discussion (0). Continue with ORCID to comment.

Reference graph

Works this paper leans on

47 extracted references · 34 canonical work pages

  1. [1]

    Naveen Arivazhagan, Colin Cherry, Wolfgang Macherey, Chung-Cheng Chiu, Semih Yavuz, Ruoming Pang, Wei Li, and Colin Raffel. 2019. Monotonic infinite lookback attention for simultaneous machine translation. In Proceedings of the Annual Meeting of the Association for Computational Linguistics, pages 1313--1323

    work page 2019

  2. [2]

    Berard, O

    A. Berard, O. Pietquin, C. Servan, and L. Besacier. 2016. Listen and translate: A proof of concept for end-to-end speech-to-text translation. In NIPS Workshop on End-to-end Learning for Speech and Audio Processing

    work page 2016

  3. [3]

    Alexandre B \'e rard, Laurent Besacier, Ali Can Kocabiyikoglu, and Olivier Pietquin. 2018. End-to-end automatic speech translation of audiobooks. In IEEE International Conference on Acoustics, Speech and Signal Processing, pages 6224--6228. IEEE

    work page 2018

  4. [4]

    William Chan, Navdeep Jaitly, Quoc V Le, and Oriol Vinyals. 2015. Listen, attend and spell. arXiv preprint arXiv:1508.01211

    arXiv 2015

  5. [5]

    X. Chen, Y. Wu, Z. Wang, S. Liu, and J. Li. 2021. Developing real-time streaming transformer transducer for speech recognition on large-scale dataset. In Proc. of ICASSP, pages 5904--5908. IEEE

    work page 2021

  6. [6]

    C. C. Chiu and C. Raffel. 2018. Monotonic chunkwise attention. In ICLR

    work page 2018

  7. [7]

    Yusuke Fujita, Naoyuki Kanda, Shota Horiguchi, Kenji Nagamatsu, and Shinji Watanabe. 2019 a . End-to-end neural speaker diarization with permutation-free objectives. arXiv preprint arXiv:1909.05952

    work page Pith review arXiv 2019

  8. [8]

    Yusuke Fujita, Naoyuki Kanda, Shota Horiguchi, Yawen Xue, Kenji Nagamatsu, and Shinji Watanabe. 2019 b . End-to-end neural speaker diarization with self-attention. In 2019 IEEE Automatic Speech Recognition and Understanding Workshop (ASRU), pages 296--303. IEEE

    work page 2019

Show all 47 references
  1. [9]

    A. Graves. 2012. Sequence transduction with recurrent neural networks. arXiv preprint arXiv:1211.3711

    2012 arXiv

  2. [10]

    Hirofumi Inaguma, Yashesh Gaur, Liang Lu, Jinyu Li, and Yifan Gong. 2020. Minimum latency training strategies for streaming sequence-to-sequence asr. In IEEE International Conference on Acoustics, Speech and Signal Processing, pages 6064--6068. IEEE

    2020

  3. [11]

    Naoyuki Kanda, Jian Wu, Yu Wu, Xiong Xiao, Zhong Meng, Xiaofei Wang, Yashesh Gaur, Zhuo Chen, Jinyu Li, and Takuya Yoshioka. 2022 a . Streaming multi-talker ASR with token-level serialized output training. In Proc. Interspeech, pages 3774--3778

    2022

  4. [12]

    Naoyuki Kanda, Jian Wu, Yu Wu, Xiong Xiao, Zhong Meng, Xiaofei Wang, Yashesh Gaur, Zhuo Chen, Jinyu Li, and Takuya Yoshioka. 2022 b . Streaming speaker-attributed asr with token-level speaker embeddings. arXiv preprint arXiv:2203.16685

    2022 arXiv

  5. [13]

    J. Li, R. Zhao, Z. Meng, Y. Liu, W. Wei, S. Parthasarathy, V. Mazalov, Z. Wang, L. He, S. Zhao, and et al. 2020. Developing rnnt models surpassing high-performance hybrid models with customization capability. In Proceedings of Interspeech, pages 3590--3594

    2020

  6. [14]

    Jinyu Li. 2022. Recent advances in end-to-end automatic speech recognition. APSIPA Transactions on Signal and Information Processing, 11(1)

    2022

  7. [15]

    D. Liu, M. Du, X. Li, Y. Li, and E. Chen. 2021. Cross attention augmented transducer networks for simultaneous translation. In Proceedings of EMNLP, pages 39--55

    2021

  8. [16]

    Xutai Ma, Juan Miguel Pino, James Cross, Liezl Puzon, and Jiatao Gu. 2019. Monotonic multihead attention. In Proceedings of International Conference on Learning Representations

    2019

  9. [17]

    Xutai Ma, Yongqiang Wang, Mohammad Javad Dousti, Philipp Koehn, and Juan Pino. 2021. Streaming simultaneous speech translation with augmented memory transformer. In IEEE International Conference on Acoustics, Speech and Signal Processing, pages 7523--7527. IEEE

    2021

  10. [18]

    Matusov, S

    E. Matusov, S. Kanthak, and H. Ney. 2005. On the integration of speech recognition and statistical machine translation. In European Conference on Speech Communicaton and Technology

    2005

  11. [19]

    Haoran Miao, Gaofeng Cheng, Pengyuan Zhang, Ta Li, and Yonghong Yan. 2019. Online hybrid ctc/attention architecture for end-to-end speech recognition. Proceedings of Interspeech, pages 2623--2627

    2019

  12. [20]

    Arsha Nagrani, Joon Son Chung, Weidi Xie, and Andrew Zisserman. 2020. Voxceleb: Large-scale speaker verification in the wild. Computer Speech & Language, 60:101027

    2020

  13. [21]

    Arsha Nagrani, Joon Son Chung, and Andrew Zisserman. 2017. Voxceleb: a large-scale speaker identification dataset. Telephony, 3:33--039

    2017

  14. [22]

    Hermann Ney. 1999. Speech translation: Coupling of recognition and translation. In Proceedings of ICASSP, pages 517--520

    1999

  15. [23]

    Sara Papi, Peter Polak, Ond r ej Bojar, and Dominik Mach \'a c ek. 2024. How" real" is your real-time simultaneous speech-to-text translation system? arXiv preprint arXiv:2412.18495

    2024 arXiv

  16. [24]

    Tae Jin Park, Naoyuki Kanda, Dimitrios Dimitriadis, Kyu J Han, Shinji Watanabe, and Shrikanth Narayanan. 2021. A review of speaker diarization: Recent advances with deep learning. arXiv:2101.09624

    2021 arXiv

  17. [25]

    Alexis Plaquet and Herv\' e Bredin. 2023. Powerset multi-class cross entropy loss for neural speaker diarization. In Proc. Interspeech 2023

    2023

  18. [26]

    M. Post, G. Kumar, A. Lopez, D. Karakos, C. Callison-Burch, and S. Khudanpur. 2013. Improved speech-to-text translation with the fisher and callhome spanish-english speech translation corpus. In Proceedings of IWSLT

    2013

  19. [27]

    Prabhavalkar, K

    R. Prabhavalkar, K. Rao, T. N. Sainath, B. Li, L. Johnson, and N. Jaitly. 2017. A comparison of sequence-to-sequence models for speech recognition. In Proceedings of Interspeech, pages 939--943

    2017

  20. [28]

    Alec Radford, Jong Wook Kim, Tao Xu, Greg Brockman, Christine McLeavey, and Ilya Sutskever. 2023. Robust speech recognition via large-scale weak supervision. In International Conference on Machine Learning, pages 28492--28518. PMLR

    2023

  21. [29]

    T. N. Sainath, Y. He, B. Li, A. Narayanan, R. Pang, A. Bruguier, S.-Y. Chang, W. Li, R. Alvarez, Z. Chen, and et al. 2020. A streaming on-device end-to-end model surpassing server-side conventional model quality and latency. In Proceedings of ICASSP, pages 6059--6003

    2020

  22. [30]

    G. Saon, Z. Tüske, D. Bolanos, and B. Kingsbury. 2021. Advancing rnn transducer technology for speech recognition. In Proceedings of ICASSP, pages 5654--5658

    2021

  23. [31]

    Matthias Sperber and Matthias Paulik. 2020. Speech translation and the end-to-end promise: Taking stock of where we are. In Proceedings of the 58th Annual Meeting of the Association for Computational Linguistics, pages 7409--7421

    2020

  24. [32]

    Yun Tang, Anna Y Sun, Hirofumi Inaguma, Xinyue Chen, Ning Dong, Xutai Ma, Paden D Tomasello, and Juan Pino. 2023. Hybrid transducer and attention based encoder-decoder modeling for speech-to-text tasks. arXiv preprint arXiv:2305.03101

    2023 arXiv

  25. [33]

    Laura Cross Vila, Carlos Escolano, Jos \'e AR Fonollosa, and Marta R Costa-Jussa. 2018. End-to-end speech translation with the transformer. In Proceedings of Interspeech, pages 60--63

    2018

  26. [34]

    P. Wang, Z. Chen, D. L. Wang, J. Li, and Y. Gong. 2020. Speaker separation using speaker inventories and estimated speech. IEEE/ACM TASLP, 29:537--546

    2020

  27. [35]

    P. Wang, Z. Chen, X. Xiao, Z. Meng, T. Yoshioka, T. Zhou, L. Lu, and J. Li. 2019 a . Speech separation using speaker inventory. In Proc. of ASRU, pages 230--236. IEEE

    2019

  28. [36]

    P. Wang, J. Cui, C. Weng, and D. Yu. 2019 b . Large margin training for attention based end-to-end speech recognition. In Proc. of INTERSPEECH, pages 246--250

    2019

  29. [37]

    P. Wang, J. Cui, C. Weng, and D. Yu. 2019 c . Token-wise training for attention based end-to-end speech recognition. In Proc. of ICASSP, pages 6276--6280

    2019

  30. [38]

    Wang and D

    P. Wang and D. L. Wang. 2018. Utterance-wise recurrent dropout and iterative speaker adaptation for robust monaural speech recognition. In Proc. of ICASSP, pages 4814--4818. IEEE

    2018

  31. [39]

    Sainath, and Ron J

    Peidong Wang, Tara N. Sainath, and Ron J. Weiss. 2021. https://doi.org/10.21437/Interspeech.2021-683 Multitask training with text data for end-to-end speech recognition . In Proc. of Interspeech, pages 2566--2570

    2021 doi

  32. [40]

    Peidong Wang, Eric Sun, Jian Xue, Yu Wu, Long Zhou, Yashesh Gaur, Shujie Liu, and Jinyu Li. 2022. Lamassu: Streaming language-agnostic multilingual speech recognition and translation using neural transducers. arXiv preprint arXiv:2211.02809

    2022 arXiv

  33. [41]

    R. J. Weiss, J. Chorowski, N. Jaitly, Y. Wu, and Z. Chen. 2017. Sequence-to-sequence models can directly translate foreign speech. In Proc. of INTERSPEECH, pages 2625--2629

    2017

  34. [42]

    Jian Xue, Peidong Wang, Jinyu Li, Matt Post, and Yashesh Gaur. 2022. Large-scale streaming end-to-end speech translation with neural transducers. arXiv preprint arXiv:2204.05352

    2022 arXiv

  35. [43]

    Jian Xue, Peidong Wang, Jinyu Li, and Eric Sun. 2023. A weakly-supervised streaming multilingual speech model with truly zero-shot capability. In 2023 IEEE Automatic Speech Recognition and Understanding Workshop (ASRU), pages 1--7. IEEE

    2023

  36. [44]

    Mu Yang, Naoyuki Kanda, Xiaofei Wang, Junkun Chen, Peidong Wang, Jian Xue, Jinyu Li, and Takuya Yoshioka. 2023. Diarist: Streaming speech translation with speaker diarization. arXiv preprint arXiv:2309.08007

    2023 arXiv

  37. [45]

    Juan Zuluaga-Gomez, Zhaocheng Huang, Xing Niu, Rohit Paturi, Sundararajan Srinivasan, Prashant Mathur, Brian Thompson, and Marcello Federico. 2023. End-to-end single-channel speaker-turn aware conversational speech translation. arXiv preprint arXiv:2311.00697

    2023 arXiv

  38. [46]

    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...

  39. [47]

    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 9, 2026 · model on record in the stance chip above.