Pith. sign in

REVIEW 3 major objections 4 minor 29 references

SCDiar: a streaming diarization system based on speaker change detection and speech recognition

T0 review · 3 major / 4 minor · reviewed 2026-08-10 · deepseek-v4-flash

Pith's one-line read SCDiar, a streaming diarization system, uses token-level speaker change detection and a representative-segment selection optimization to make online speaker-attributed ASR nearly as accurate as offline methods on multi-hour meetings.

desk verdict SCDiar has a real new idea in segment selection, but the optimization as written doesn't enforce one-to-one speaker mapping, and the evaluation is too thin to take the numbers at face value. read the letter →

arxiv 2501.16641 v1 pith:LF7FTCSL submitted 2025-01-28 eess.AS cs.HCcs.SD

classification eess.AScs.HCcs.SD
keywords speakerdiarizationstreamingchangedetectionspeaker-attributedASRrepresentativesegmentselectionCIF-ASRsegment-tokensimilaritybounded-variableleastsquares
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

SCDiar is a streaming system for turning a live meeting audio stream into 'who said what' text. It splits the stream at token-level speaker-change boundaries found jointly by an ASR aligner and a speaker embedding extractor, then—instead of clustering every short segment—solves a small optimization to pick one reliable 'representative segment' per speaker and uses those anchors to map tokens to speakers and update a running speaker cache. The paper's central empirical claim is that this design makes online diarization almost as accurate as offline methods: on AISHELL-4 the ΔcpWER gap to the offline spectral-clustering baseline is only 1.3 percentage points, and on an in-house 7-hour corpus with 10+ speakers per meeting it is 0.7 percentage points, while clearly beating prior online systems. The authors also show via ablation that representative-segment selection is the main source of the gain, and that removing the split-strategy augmentation hurts on the harder corpus.

What carries the argument

The central mechanism is a rectangular segment-token similarity matrix A of shape tokens × segments, produced by a learned aligner that refines cosine similarities using token-level speaker embeddings, global segment embeddings, and a cross-attention decoder. On this matrix the system solves min ∥Ax − 1∥² with box constraints (bounded-variable least squares) to find representative segments; the binary vector x indicates which segments best cover all tokens exactly once. A token-count-weighted running mean updates cached speaker centers, and a threshold θcache decides whether a new segment becomes a new speaker.

What would settle it

Take a controlled meeting recording in which one participant deliberately moves to a different seat or changes emotional tone for a few minutes. If SCDiar assigns that participant two or more distinct speaker IDs that persist beyond the change, the cache-stability assumption fails and the reported gains would not hold under normal meeting dynamics.

Watch

Extended reading notes

Core claim

The paper claims that a streaming speaker-attributed ASR system can operate on speech segments produced by a token-level SCD module and still achieve near-offline diarization accuracy, provided the clustering step selects a single representative segment per speaker instead of using all segments. The authors argue that this one-to-one mapping, obtained by relaxing an integer optimization into bounded-variable least squares on a token-by-segment similarity matrix, solves the problem of short, information-poor utterances that plague streaming clustering. With this mechanism, SCDiar reports ΔcpWER of 3.42 and WDER of 3.56 on AISHELL-4, and ΔcpWER of 10.66 and WDER of 15.36 on the in-house 10+ speaker corpus, against 2.13/2.69 and 9.95/13.51 for the best offline baseline. The paper concludes that the method significantly narrows the gap between online and offline diarization.

Load-bearing premise

The load-bearing assumption is that a single cached speaker embedding centroid stays a valid representation of a participant across an hours-long meeting, despite changes in emotion, spatial position, and head movement; the paper itself acknowledges that such variations can result in the creation of erroneous virtual speakers.

Editorial extensions

If this is right

  • Streaming speaker-attributed ASR for multi-hour meetings can operate with a latency bounded by the VAD chunk length and still keep diarization error within about a percentage point of offline systems.
  • Representative segment selection is the key ingredient: removing it raises ΔcpWER from 3.42 to 11.96 on AISHELL-4, showing that picking one reliable segment per speaker matters more than using all segments.
  • The system runs in real time on a CPU+GPU setup, with RTF values of 0.072 for ASR, 0.004 for SCD, and 0.009 for SD, so the accuracy gains do not come at the cost of streaming feasibility.
  • Short VAD chunks below 3 seconds degrade performance sharply, which means the method is suited to meeting-style speech where turns can be batched, not to extremely low-latency interactive dialogue.

Reading between the lines

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

  • Because the representative-segment selection is formulated as a generic least-squares cover problem, the same idea could be reused for any streaming clustering task where some incoming data points are too short to be reliable—for example, online speaker counting in smart speakers—though the paper does not test this.
  • The paper evaluates only Mandarin meetings; since the SCD and SD modules are trained on acoustic features plus ASR alignments rather than on language-specific text, the same architecture should transfer to other languages once a CIF-style ASR backbone with token timestamps exists, but that transfer is untested.
  • The fixed thresholds θcache=0.55 and θsd=0.3 are set experimentally; a confidence-adaptive update rule for the speaker cache might reduce the virtual-speaker failures the authors flag, and could be validated by measuring the number of spurious speakers on meetings with strong vocal variation.
  • The reported 1.3% and 0.7% ΔcpWER gaps are relative to a single offline baseline (sliding-window spectral clustering); the approach may appear better or worse against other offline methods such as VBx with oracle segmentation, so the gap should be read as indicative rather than absolute.
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

3 major / 4 minor

Summary. SCDiar is a streaming speaker-attributed ASR/diarization system. The audio stream is processed by a CIF-based ASR; a token-level speaker change detection module splits the transcript into segments; a speaker-diarization network produces a token-segment similarity matrix; and a bounded-variable least-squares problem selects 'representative segments' that are then mapped to a streaming cache of speaker centroids. Results are reported on the AISHELL-4 test set and on a five-meeting in-house corpus, with ablations for the split augmentation and for the representative-selection stage.

Significance. The representative-segment selection idea is timely and potentially useful for streaming diarization, where short segments carry little speaker information. The paper's strengths are its integrated system design, the public AISHELL-4 benchmark, an RTF measurement, and ablations isolating the split strategy and representative selection. If the mechanism were implemented as described, the reported margins over online baselines would be important. However, the current evidence is not yet sufficient to establish the central claim: the optimization as written does not enforce one-to-one selection, and the evaluation rests on a very small in-house corpus without variance or significance information.

major comments (3)
  1. [Section II-D, Eqs. (14)-(15)] The objective as written does not enforce the claimed one-to-one mapping. The sentence following Eq. (14) states that 'the number of ones in x equals to the predicted number of speakers,' but no cardinality constraint appears in Eq. (14) or in the relaxed problem in Eq. (15); the only subsequent selection mechanism is the fixed threshold θsd. Because A contains sigmoid similarities, for a speaker whose acoustic states are split across two segments the row sum A[i,:]x can be closer to 1 when both segments are selected than when only one is selected, so the optimizer is rewarded for creating duplicate representatives. This is exactly the 'erroneous virtual speakers' failure mode admitted in the Conclusion, and it means the large gain of '-w.o. rep. select' in Table I cannot be attributed to the stated optimization without additional evidence. The paper should either add an explicit constraint (e.g., sum(x)=K with the predicted speaker count) or provide an analysis of the selected x vectors showing that thresholding does not admit multiple representatives of the same speaker in the evaluated meetings.
  2. [Section III-C, Table I] The evaluation on the in-house corpus uses only five meetings, and no per-meeting numbers, confidence intervals, or significance tests are provided. Since the paper's headline claim ('outperforms previous systems by up to 53.6%') and the conclusion's 'marginal 1.3% and 0.7% increase' refer to this small corpus, the reported margins may not be stable. The baseline systems 3 and 4 are also far weaker than the proposed method (e.g., in-house ΔcpWER 75.98 and 64.29 vs. 10.66), which raises the question of whether the comparison is fair; please report the baseline hyperparameter tuning procedure and, ideally, results on a larger public multi-speaker meeting corpus or a leave-one-meeting-out analysis.
  3. [Section II-E and Conclusion] The streaming cache keeps a single token-count-weighted centroid per speaker and has no merging mechanism. The Conclusion concedes that emotion, spatial position, and head movement 'can result in the creation of erroneous virtual speakers.' This is a load-bearing limitation for the central claim because the target scenario is hours-long meetings with over 10 speakers, where such variations are likely. The paper offers no analysis of how often this failure occurs (e.g., growth of the cache relative to the true number of speakers, or number of merge events). Without that information, the reported WDER/cpWER numbers cannot be distinguished from a system that is accidentally tolerant of virtual speakers via the metric's permutation alignment.
minor comments (4)
  1. [Eq. (14)] The notation x ∈ Z^M should be x ∈ {0,1}^M, and the all-one vector 1 should be explicitly dimensioned as R^N.
  2. [Section II-B] The acronym for the transfer algorithm is written 'TPSP' in the text but 'TPST' in the Figure 2 caption; please make it consistent.
  3. [Section III-C] The abstract's 'up to 53.6% in accuracy' is not directly derivable from any single row of Table I as printed; please state which metric and baseline this improvement refers to.
  4. [Section III-A] Please clarify whether the '5 internal real meetings' are all in Mandarin and whether the reported ASR WER of 17.40% is computed on the same in-house corpus; the current text is ambiguous.

Circularity Check

0 steps flagged · score 0.0 of 10

No circularity: SCDiar's gains are empirical, measured on external corpora, and not derived by construction from its inputs.

full rationale

The paper's central claims are empirical results on AISHELL-4 and an in-house meeting corpus. The proposed system combines pretrained ASR and speaker-embedding extractors with trained SCD and SD modules; no parameter is fitted to the evaluation labels. The optimization in Eqs. (14)-(17) is a heuristic for selecting representative segments, and the thresholds θscd, θsd, and θcache are fixed constants. The reported cpWER and WDER numbers are measured outcomes, not predicted quantities derived from the system's own inputs. All cited building blocks (Paraformer, CAM++, VBx, AISHELL-4, VoxCeleb-2) are external resources or prior work by other research groups, and no load-bearing self-citation or uniqueness theorem is invoked. The reviewer-identified concern about Eq. (14) lacking a cardinality constraint is an optimization/implementation issue that may cause virtual speakers, but it is not a circularity: it does not reduce a claimed result to an input by construction. Therefore no circular steps are present.

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

SCDiar is an empirical engineering system rather than a derivation. The load-bearing choices are the optimization relaxation in Eq. (15), the stability of cached speaker centroids, and the experimentally set thresholds. The paper does not introduce new physical entities.

free parameters (5)
  • θscd = 0.25
    Peak detection threshold for token-level speaker change detection, set experimentally in Section III-B.
  • θsd = 0.3
    Threshold applied to the relaxed representative segment solution x in Eq. (15), set experimentally.
  • θcache = 0.55
    Cosine similarity threshold used to discard new segments that are highly similar to cached speakers, Section II-E.
  • minimum token count = 10
    Segments with fewer than 10 tokens are excluded from representative segment selection and cache update, Section III-B.
  • λ = 10
    Scaling factor for the SCD focal loss in the multi-target training loss of Eq. (18).
assumptions (5)
  • domain assumption The relaxed non-negative least squares solution of Eq. (15), thresholded by θsd, approximates the binary representative-segment selection of Eq. (14) well enough for good clustering.
    The paper relaxes the integer constraint and never bounds the approximation gap; Section II-D.
  • domain assumption Speaker embeddings are stable enough that a token-count-weighted running mean per speaker, maintained in the cache, tracks the speaker over hours.
    Section II-E relies on this for label mapping and cache updates, yet the conclusion lists emotion, spatial position, and head movement as sources of erroneous virtual speakers.
  • ad hoc to paper Minimizing ||Ax - 1||^2 selects representative segments that are stable, close to same-speaker segments, and distinct from other speakers.
    This objective is a design choice in Section II-D with no external validation that the optimized x corresponds to human-annotated representative segments.
  • domain assumption Transcript-preserving speaker transfer (TPSP) from reference [18] maps reference speaker labels to hypothesis transcripts accurately enough to train the SCD module.
    Section II-B relies on TPSP for SCD training labels; errors in this mapping would add noise to the focal loss targets.
  • ad hoc to paper Random split augmentation during training simulates the distribution of SCD segmentations seen at inference.
    Section II-F introduces this augmentation but provides no analysis showing that the simulated splits match real SCD boundary variability.

how reviews work

0 comments
Cite this review

Pith. "Pith review of SCDiar: a streaming diarization system based on speaker change detection and speech recognition." pith.science (2026). https://pith.science/paper/LF7FTCSL

@misc{pith2026250116641,
  author       = {Pith},
  title        = {Pith review of: SCDiar: a streaming diarization system based on speaker change detection and speech recognition},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/LF7FTCSL}},
  note         = {Machine review of arXiv:2501.16641}
}
read the original abstract

In hours-long meeting scenarios, real-time speech stream often struggles with achieving accurate speaker diarization, commonly leading to speaker identification and speaker count errors. To address this challenge, we propose SCDiar, a system that operates on speech segments, split at the token level by a speaker change detection (SCD) module. Building on these segments, we introduce several enhancements to efficiently select the best available segment for each speaker. These improvements lead to significant gains across various benchmarks. Notably, on real-world meeting data involving more than ten participants, SCDiar outperforms previous systems by up to 53.6\% in accuracy, substantially narrowing the performance gap between online and offline systems.

Figures

Figures reproduced from arXiv: 2501.16641 by the authors.

Figure 1
Figure 1. (a) The overview of inference process in the SCDiar system. (b) The structure of the SCD module. (c) The structure of the SD module. [PITH_FULL_IMAGE:figures/full_fig_p002_1.png] view at source ↗
Figure 2
Figure 2. An example for TPST [18]. of individual tokens within each segment. We first employ a speaker embedding extractor to obtain segment-level embeddings E spk seg, i.e., E spk seg,1:M = EmbedExtractor(X1:L, Tseg 1:M), (8) Then, we develop a network to refine the speaker similarity, as illustrated in [PITH_FULL_IMAGE:figures/full_fig_p002_2.png] view at source ↗
Figure 3
Figure 3. Segment-token similarity matrix from (a) cosine distance, (b) [PITH_FULL_IMAGE:figures/full_fig_p003_3.png] view at source ↗
Figures from the paper (1 more)
Figure 4
Figure 4. Figure 4: ASR and diarization results on AISHELL-4 with different maximum [PITH_FULL_IMAGE:figures/full_fig_p004_4.png]

Discussion (0). Continue with ORCID to comment.

Reference graph

Works this paper leans on

29 extracted references · 26 canonical work pages

  1. [1]

    A spectral clustering approach to speaker diarization,

    H. Ning, M. Liu, H. Tang, and T. S. Huang, “A spectral clustering approach to speaker diarization,” in Interspeech, 2006. [Online]. Available: https://api.semanticscholar.org/CorpusID:4547281

  2. [2]

    Bayesian hmm clustering of x-vector sequences (vbx) in speaker diarization: theory, implementa- tion and analysis on standard tasks,

    F. Landini, J. Profant, M. Diez, and L. Burget, “Bayesian hmm clustering of x-vector sequences (vbx) in speaker diarization: theory, implementa- tion and analysis on standard tasks,” Computer Speech & Language , vol. 71, p. 101254, 2022

  3. [3]

    End-to-end neural speaker diarization with permutation- free objectives,

    Y . Fujita, N. Kanda, S. Horiguchi, K. Nagamatsu, and S. Watanabe, “End-to-end neural speaker diarization with permutation- free objectives,” in Interspeech, 2019. [Online]. Available: https://api.semanticscholar.org/CorpusID:202572807

  4. [4]

    End-to-end speaker diarization for an unknown number of speakers with encoder-decoder based attractors,

    S. Horiguchi, Y . Fujita, S. Watanabe, Y . Xue, and K. Nagamatsu, “End-to-end speaker diarization for an unknown number of speakers with encoder-decoder based attractors,” in Interspeech, 2020. [Online]. Available: https://api.semanticscholar.org/CorpusID:218719457

  5. [5]

    Target- speaker voice activity detection: a novel approach for multi-speaker diarization in a dinner party scenario,

    I. Medennikov, M. Korenevsky, T. Prisyach, Y . Y . Khokhlov, M. Korenevskaya, I. Sorokin, T. Timofeeva, A. Mitrofanov, A. Andrusenko, I. Podluzhny, A. Laptev, and A. Romanenko, “Target- speaker voice activity detection: a novel approach for multi-speaker diarization in a dinner party scenario,” in Interspeech, 2020. [Online]. Available: https://api.semant...

  6. [6]

    Online speaker diarization with core samples selection,

    Y . Yue, J. Du, M. He, Y . T. Yeung, and R. Wang, “Online speaker diarization with core samples selection,” in Interspeech, 2022. [Online]. Available: https://api.semanticscholar.org/CorpusID:252336337

  7. [7]

    Frame-wise streaming end-to- end speaker diarization with non-autoregressive self-attention-based attractors,

    D. Liang, N. Shao, and X. Li, “Frame-wise streaming end-to- end speaker diarization with non-autoregressive self-attention-based attractors,” ICASSP 2024 - 2024 IEEE International Conference on Acoustics, Speech and Signal Processing (ICASSP) , pp. 10 521–10 525,

  8. [8]

    Online neural diarization of unlimited numbers of speakers using global and local attractors,

    S. Horiguchi, S. Watanabe, L. P. Garc ´ıa-Perera, Y . Takashima, and Y . Kawaguchi, “Online neural diarization of unlimited numbers of speakers using global and local attractors,” IEEE/ACM Transactions on Audio, Speech, and Language Processing , vol. 31, pp. 706–720,

Show all 29 references
  1. [9]

    Investigation of end-to-end speaker-attributed asr for continuous multi-talker recordings,

    N. Kanda, X. Chang, Y . Gaur, X. Wang, Z. Meng, Z. Chen, and T. Yoshioka, “Investigation of end-to-end speaker-attributed asr for continuous multi-talker recordings,” 2021 IEEE Spoken Language Technology Workshop (SLT) , pp. 809–816, 2020. [Online]. Available: https://api.sema...

  2. [10]

    End-to-End Speaker-Attributed ASR with Transformer,

    N. Kanda, G. Ye, Y . Gaur, X. Wang, Z. Meng, Z. Chen, and T. Yosh- ioka, “End-to-End Speaker-Attributed ASR with Transformer,” in Proc. Interspeech 2021, 2021, pp. 4413–4417

  3. [11]

    Towards word-level end-to-end neural speaker diarization with auxiliary network,

    Y . Huang, W. Wang, G. Zhao, H. Liao, W. Xia, and Q. Wang, “Towards word-level end-to-end neural speaker diarization with auxiliary network,” ArXiv, vol. abs/2309.08489, 2023. [Online]. Available: https://api.semanticscholar.org/CorpusID:262012700

  4. [12]

    Sa-paraformer: Non-autoregressive end-to-end speaker- attributed asr,

    Y . Li, F. Yu, Y . Liang, P. Guo, M. Shi, Z. Du, S. Zhang, and L. Xie, “Sa-paraformer: Non-autoregressive end-to-end speaker- attributed asr,” 2023 IEEE Automatic Speech Recognition and Understanding Workshop (ASRU) , pp. 1–7, 2023. [Online]. Available: https://api.semanticsch...

  5. [13]

    Transcribe-to-diarize: Neural speaker diarization for unlimited number of speakers using end-to-end speaker-attributed asr,

    N. Kanda, X. Xiao, Y . Gaur, X. Wang, Z. Meng, Z. Chen, and T. Yoshioka, “Transcribe-to-diarize: Neural speaker diarization for unlimited number of speakers using end-to-end speaker-attributed asr,” ICASSP 2022 - 2022 IEEE International Conference on Acoustics, Speech and Sign...

  6. [14]

    Highly efficient real-time streaming and fully on-device speaker diarization with multi-stage clustering,

    Q. Wang, Y . Huang, H. Lu, G. Zhao, and I. Moreno, “Highly efficient real-time streaming and fully on-device speaker diarization with multi-stage clustering,” ArXiv, vol. abs/2210.13690, 2022. [Online]. Available: https://api.semanticscholar.org/CorpusID:253107785

  7. [15]

    Streaming Multi-Talker ASR with Token-Level Serialized Output Training,

    N. Kanda, J. Wu, Y . Wu, X. Xiao, Z. Meng, X. Wang, Y . Gaur, Z. Chen, J. Li, and T. Yoshioka, “Streaming Multi-Talker ASR with Token-Level Serialized Output Training,” in Proc. Interspeech 2022, 2022, pp. 3774– 3778

  8. [16]

    Streaming Speaker-Attributed ASR with Token-Level Speaker Embeddings,

    ——, “Streaming Speaker-Attributed ASR with Token-Level Speaker Embeddings,” in Proc. Interspeech 2022 , 2022, pp. 521–525

  9. [17]

    Cif: Continuous integrate-and-fire for end-to- end speech recognition,

    L. Dong and B. Xu, “Cif: Continuous integrate-and-fire for end-to- end speech recognition,” in ICASSP 2020 - 2020 IEEE International Conference on Acoustics, Speech and Signal Processing (ICASSP), 2020, pp. 6079–6083

  10. [18]

    Diarizationlm: Speaker diarization post-processing with large language models,

    Q. Wang, Y . Huang, G. Zhao, E. Clark, W. Xia, and H. Liao, “Diarizationlm: Speaker diarization post-processing with large language models,” ArXiv, vol. abs/2401.03506, 2024. [Online]. Available: https://api.semanticscholar.org/CorpusID:266844688

  11. [19]

    Focal loss for dense object detection,

    T.-Y . Lin, P. Goyal, R. Girshick, K. He, and P. Dollar, “Focal loss for dense object detection,” in Proceedings of the IEEE International Conference on Computer Vision (ICCV) , Oct 2017

  12. [20]

    Bounded-variable least-squares: an algorithm and applica- tions,

    P. B. Stark, “Bounded-variable least-squares: an algorithm and applica- tions,” in Computational Statistics, 1995

  13. [21]

    AISHELL-4: An Open Source Dataset for Speech Enhancement, Separation, Recognition and Speaker Diarization in Conference Scenario,

    Y . Fu, L. Cheng, S. Lv, Y . Jv, Y . Kong, Z. Chen, Y . Hu, L. Xie, J. Wu, H. Bu, X. Xu, J. Du, and J. Chen, “AISHELL-4: An Open Source Dataset for Speech Enhancement, Separation, Recognition and Speaker Diarization in Conference Scenario,” in Proc. Interspeech 2021 , 2021, pp...

  14. [22]

    Acoustic beamforming for speaker diarization of meetings,

    X. Anguera, C. Wooters, and J. Hernando, “Acoustic beamforming for speaker diarization of meetings,” IEEE Transactions on Audio, Speech, and Language Processing , vol. 15, no. 7, pp. 2011–2021, September 2007

  15. [23]

    Paraformer: Fast and Accurate Parallel Transformer for Non-autoregressive End-to-End Speech Recognition,

    Z. Gao, S. Zhang, I. McLoughlin, and Z. Yan, “Paraformer: Fast and Accurate Parallel Transformer for Non-autoregressive End-to-End Speech Recognition,” in Proc. Interspeech 2022 , 2022, pp. 2063–2067

  16. [24]

    CAM++: A Fast and Efficient Network for Speaker Verification Using Context-Aware Masking,

    H. Wang, S. Zheng, Y . Chen, L. Cheng, and Q. Chen, “CAM++: A Fast and Efficient Network for Speaker Verification Using Context-Aware Masking,” in Proc. INTERSPEECH 2023 , 2023, pp. 5301–5305

  17. [25]

    Joint Speech Recognition and Speaker Diarization via Sequence Transduction,

    L. E. Shafey, H. Soltau, and I. Shafran, “Joint Speech Recognition and Speaker Diarization via Sequence Transduction,” in Proc. Interspeech 2019, 2019, pp. 396–400

  18. [26]

    Chime-6 challenge: Tackling multispeaker speech recognition for unsegmented recordings,

    S. Watanabe, M. Mandel, J. Barker, and E. Vincent, “Chime-6 challenge: Tackling multispeaker speech recognition for unsegmented recordings,” ArXiv, vol. abs/2004.09249, 2020. [Online]. Available: https://api.semanticscholar.org/CorpusID:215828091

  19. [27]

    V oxceleb2: Deep speaker recognition,

    J. S. Chung, A. Nagrani, and A. Zisserman, “V oxceleb2: Deep speaker recognition,” Proc. Interspeech 2018 , pp. 1086–1090, 2018

  20. [2022]

    Available: https://api.semanticscholar.org/CorpusID: 249394670

    [Online]. Available: https://api.semanticscholar.org/CorpusID: 249394670

  21. [2023]

    Available: https://api.semanticscholar.org/CorpusID: 262465343

    [Online]. Available: https://api.semanticscholar.org/CorpusID: 262465343

Pith tools

Reviewed August 10, 2026 · model on record in the stance chip above.