Pith. sign in

REVIEW 3 major objections 6 minor 33 references

Towards Robust Overlapping Speech Detection: A Speaker-Aware Progressive Approach Using WavLM

T0 review · 3 major / 6 minor · reviewed 2026-08-07 · deepseek-v4-flash

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

Pith's one-line read This paper reports that a speaker-aware progressive OSD model, built on WavLM-Large with Campplus speaker attention and VAD-gated temporal masking, achieves 82.76% F1 on the AMI test set, the best published result on that benchmark.

desk verdict Solid OSD paper with a plausible SOTA claim, but the missing AMI split specification makes the central number unverifiable until the authors clarify training/test separation. read the letter →

arxiv 2505.23207 v1 pith:Q7FP3C5P submitted 2025-05-29 cs.SD cs.LGeess.AS

Zhaokai Sun , Li Zhang , Qing Wang , Pan Zhou , Lei Xie This is my paper · ORCID
classification cs.SDcs.LGeess.AS
keywords overlappingspeechdetectionspeakerattentionprogressivelearningWavLMself-supervisedvoiceactivityAMIcorpusembeddings
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 claims that overlapping speech detection improves when a model treats voice activity detection as a gate for overlap detection instead of a parallel classification target. Its proposed system runs WavLM-Large over the audio, fuses frame-level Campplus speaker embeddings into those features by cross-attention, masks the fused features with VAD logits, and only then feeds them to an OSD decoder. On the AMI meeting test set this reaches an F1 of 82.76%, a relative improvement of about 4.4% over the previous published best, with ablations attributing the gain to both the speaker attention module and the progressive masking strategy. If the result is what it appears to be, explicit modeling of the VAD-to-OSD hierarchy and of speaker identity are effective design choices for multi-party overlap detection.

What carries the argument

The load-bearing mechanism is the progressive temporal mask: the OSD decoder's input is $\mathbf{R}_{\text{mask}} = F_{\text{mask}}(S_{\text{vad}}, \mathbf{R}_{\text{att}})$, where $S_{\text{vad}}$ are the VAD decoder's logits and $\mathbf{R}_{\text{att}}$ is the speaker-enhanced representation. This gate forces overlap detection to operate only on frames the VAD believes contain speech, making the VAD-to-OSD hierarchy explicit. The second mechanism is the frame-level speaker attention module: a cross-attention layer using $\mathbf{R}_{\text{raw}}$ (WavLM hidden states) as query and $\mathbf{R}_{\text{spk}}$ (Campplus frame-level speaker embeddings) as key and value, with a residual addition, which enriches the acoustic features with speaker identity before both decoders. Both decoders are stacks of Conformer blocks, and the fuzzy-label MSE loss converts prediction into regression.

What would settle it

Reproduce the same recipe with a strict meeting-disjoint AMI split: hold out entire meetings for testing, check that no test meeting ID or segment appears in the fine-tuning data, and report F1 on that held-out set. If 82.76% cannot be reproduced under that separation, or if the gain over the XLSR-Conformer baseline disappears, the central claim is an artifact of data leakage rather than the proposed architecture.

Watch

Extended reading notes

Core claim

On the paper's own terms, the central discovery is a new state of the art for overlapping speech detection on AMI: an F1 of 82.76% with recall 81.48% and precision 84.08%. The system combines WavLM-Large as a frozen pretrained SSL encoder, a Campplus speaker-verification frontend whose frame-level embeddings are injected into the WavLM hidden states through a cross-attention module with a residual connection, and a progressive training design in which VAD logits temporally mask the encoder output before the OSD decoder sees it. Training first pretrains on LibriHeavyMix, then fine-tunes on realistic meeting data with balanced 1:1:1 sampling of silence, single-speaker, and overlap frames and fuzzy labels whose boundaries decay linearly over ten frames. Ablation experiments show that removing the speaker attention drops F1 from 82.76% to 79.47%, replacing attention with MSE alignment drops it to 81.62%, and replacing the progressive masking with a unified multi-task output drops it to 82.20%, while WavLM outperforms XLSR as the frontend.

Load-bearing premise

The entire empirical claim rests on the AMI test set being genuinely unseen during fine-tuning, and the paper never specifies how AMI was split, so leakage from the same meetings would invalidate the reported 82.76% F1.

Editorial extensions

If this is right

  • On AMI, the proposed system beats every previously reported method in Table 2, with F1 82.76% versus 79.21% for XLSR-Conformer, a relative gain of roughly 4.4%.
  • The progressive masking strategy improves over the unified multi-task variant under identical components (82.76% versus 82.20%), so the paper's claimed gain is not solely from the stronger frontend.
  • The speaker attention module is the largest single contributor: F1 falls from 82.76% to 79.47% when it is removed, and attention beats MSE-based speaker alignment (82.76% versus 81.62%).
  • Choosing WavLM-Large over XLSR as the SSL frontend adds about 2.7 F1 points when speaker attention is present (82.76% versus 80.07%).

Reading between the lines

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

  • Editorial inference: the same VAD-gating design could transfer to other hierarchical speech tasks, such as speaker-change detection or laughter detection, where a coarse speech/non-speech decision can safely restrict where the finer detector looks.
  • Editorial inference: because Campplus embeddings are speaker-discriminative and the attention module is the main contributor, the method may generalize to unseen speakers better than speaker-agnostic OSD; a cross-corpus experiment with held-out speakers would test this.
  • Editorial inference: the paper does not isolate the contribution of the balanced 1:1:1 sampling or the fuzzy-label decay, so part of the 82.76% could come from the data-curation recipe rather than the architecture; ablating those factors separately would clarify the mechanism.
  • Editorial inference: the reported system uses WavLM-Large plus Campplus plus two Conformer decoders, so the practical cost is nontrivial; reporting inference speed or parameter count would determine whether the F1 gain survives deployment constraints.
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

3 major / 6 minor

Summary. The paper proposes a speaker-aware progressive OSD model that combines WavLM-Large self-supervised features with a Campplus-based frame-level speaker attention module and a VAD-gated temporal masking module. The model is first pretrained on LibriHeavyMix and then fine-tuned on realistic data including AliMeeting and AMI, and evaluated on the AMI test set. The central claim is an F1 score of 82.76%, presented as state of the art and supported by ablations on the speaker attention module, the progressive versus unified decoding strategy, and the SSL frontend.

Significance. If the AMI evaluation protocol is clean, the reported result is a meaningful advance: a relative F1 improvement of about 4.4% over the listed XLSR-Conformer baseline, with controlled internal ablations showing consistent gains from the speaker attention module and the progressive masking strategy. The use of pretrained WavLM and Campplus is appropriate, and the component ablations are well designed. However, the empirical claim currently rests on a single F1 number with no variance or significance testing, and the comparison to prior work is made against published numbers rather than re-run baselines. Moreover, the paper does not clearly separate the AMI training and test partitions, which is essential to rule out leakage.

major comments (3)
  1. [§3.1.1, Table 1] The training protocol is not sufficiently specified. The paper states that the model is 'first pretrained for five epochs on LibriHeavyMix before being fine-tuned using realistic data' and lists AMI (75 hours) as a training corpus, but it never states which AMI partition is used for fine-tuning versus which partition is used as the 'AMI test set.' If any portion of the AMI test set, or overlapping segments from the same meetings, was included in the fine-tuning data, the reported 82.76 F1 in Table 2 would be an artifact of leakage. Please specify the exact train/dev/test partition (e.g., the standard AMI meeting split) and confirm that no test material or same-meeting segments were used during training.
  2. [§3.2, Table 2] All results are reported as single F1, recall, and precision values with no variance across runs, no significance tests, and no confidence intervals. Since the claimed SOTA rests on a 3.55-point F1 gap over the XLSR-Conformer baseline, it is important to know whether this gap is stable across random seeds or within run-to-run noise. Please report mean ± std over at least three training runs for the main result and the key ablations, or otherwise provide a significance analysis.
  3. [§3.2, Table 2] The comparison with prior work is made against previously published numbers, with no indication that the same evaluation protocol, label definition (including fuzzy labeling), and scoring script were used. In particular, the citation labels in Table 2 are inconsistent with the reference list (e.g., 'x-vectors [15]' points to the AMI corpus paper, and 'pyannote [16]' to a multi-task learning survey), which raises a question about whether the baseline numbers are correctly attributed. Please clarify how each baseline was obtained and ensure the evaluation conditions match.
minor comments (6)
  1. [§2.5] Equation (8) defines R_mask = F_mask(S_vad, R_att) but does not describe the actual masking operation; please specify how the VAD logits are converted into a mask (e.g., threshold, sigmoid, or hard masking).
  2. [§3.1.1] There is a typo in the Table 1 caption: 'conversastion' should be 'conversation'.
  3. [§2.4] The module name is written inconsistently as 'Campplus' in the text and 'CampPlus' in Section 2.4; please use a single consistent name.
  4. [§3.2] In the discussion of Table 4, the paper states that the progressive model yields 'superior recall and false alarm rates,' but the table reports precision, not a false alarm rate; please either report the false alarm rate or revise the wording.
  5. [§4] The line 'The Interspeech 2025 organisers' immediately after the conclusion paragraph appears to be a leftover template string and should be removed.
  6. [§3.1.2] The paper says all setups maintain 'consistent batch sizes and optimizer configurations' but does not report the batch size; please add it for reproducibility.

Circularity Check

0 steps flagged · score 0.0 of 10

No significant circularity: the SOTA claim is evaluated against an external benchmark and controlled ablations, with no load-bearing self-citation or fitted-input prediction.

full rationale

No circular reasoning is present in the derivation chain. The system first extracts frozen WavLM-Large representations and Campplus frame-level speaker embeddings, then fuses them via cross-attention (Eq. 7). A VAD decoder produces logits from that fused representation (Eq. 9), and the temporal masking module uses those predicted VAD logits to gate the representation before the OSD decoder produces overlap logits (Eqs. 8 and 10). The VAD logits used for masking are model outputs, not ground-truth labels, so the OSD prediction is not defined in terms of its own target. The central empirical claim, the 82.76% F1 on the AMI test set, is compared against external previously reported methods (Table 2) and is further supported by internal ablations (Tables 3–5). The use of pretrained WavLM and Campplus is not circular because their pretraining did not use AMI OSD labels. The only substantive concern, the undisclosed AMI train/test split, is a data-leakage and reproducibility risk rather than a circularity, so it does not affect the circularity score.

Assumptions & free parameters 4 free parameters · 3 assumptions · 0 invented entities

Since the paper makes an empirical benchmark claim, the ledger records the hand-chosen training and labeling choices plus background assumptions that the final F1 depends on. No new physical or theoretical entities are introduced.

free parameters (4)
  • Fuzzy label decay span = 10 frames
    Labels decay linearly from 1 to 0 over 10 frames around VAD/OSD boundaries (Section 3.1.1); no sensitivity ablation is given.
  • Speech class balance ratio = 1:1:1 (silence:single-speaker:overlap)
    Training segments are curated per session to balance classes (Section 3.1.1); the ratio is chosen by hand and no variation is tested.
  • Fine-tuning duration and schedule = not reported
    Pretraining on LibriHeavyMix is 5 epochs, but the fine-tuning epoch count and learning rate schedule on realistic data are unspecified (Section 3.1.2), and these choices affect the final F1.
  • Temporal mask operating parameters = undefined
    Eq. 8 only defines R_mask = F_mask(S_vad, R_att); whether the mask is a threshold, sigmoid, or learned gate, and any associated values, are not given. The OSD decoder input depends directly on this operation.
assumptions (3)
  • domain assumption WavLM-Large hidden states are a suitable frame-level feature source for OSD after Conformer decoding.
    The system relies on the pretrained SSL encoder without OSD-specific pretraining or feature analysis (Section 2.3).
  • domain assumption Campplus frame-level speaker embeddings contain information useful for distinguishing single-speaker from overlapping frames.
    The speaker attention module is built on this; Table 3 gives empirical support on AMI only.
  • domain assumption Manual annotations and the fuzzy-label smoothing define a valid regression target.
    No annotation reliability analysis is provided; the fuzzy labeling is adopted from the literature (Section 2.6).

how reviews work

0 comments
Cite this review

Pith. "Pith review of Towards Robust Overlapping Speech Detection: A Speaker-Aware Progressive Approach Using WavLM." pith.science (2026). https://pith.science/paper/Q7FP3C5P

@misc{pith2026250523207,
  author       = {Pith},
  title        = {Pith review of: Towards Robust Overlapping Speech Detection: A Speaker-Aware Progressive Approach Using WavLM},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/Q7FP3C5P}},
  note         = {Machine review of arXiv:2505.23207}
}
read the original abstract

Overlapping Speech Detection (OSD) aims to identify regions where multiple speakers overlap in a conversation, a critical challenge in multi-party speech processing. This work proposes a speaker-aware progressive OSD model that leverages a progressive training strategy to enhance the correlation between subtasks such as voice activity detection (VAD) and overlap detection. To improve acoustic representation, we explore the effectiveness of state-of-the-art self-supervised learning (SSL) models, including WavLM and wav2vec 2.0, while incorporating a speaker attention module to enrich features with frame-level speaker information. Experimental results show that the proposed method achieves state-of-the-art performance, with an F1 score of 82.76\% on the AMI test set, demonstrating its robustness and effectiveness in OSD.

Figures

Figures reproduced from arXiv: 2505.23207 by the authors.

Figure 2
Figure 2. Frame-level speaker attention module. 2.2. Progressive OSD modeling The progressive modeling approach we propose is depicted in [PITH_FULL_IMAGE:figures/full_fig_p002_2.png] view at source ↗

Discussion (0). Continue with ORCID to comment.

Reference graph

Works this paper leans on

33 extracted references · 25 canonical work pages

  1. [15]

    End-to-end speaker segmentation for overlap-aware resegmentation,

    H. Bredin and A. Laurent, “End-to-end speaker segmentation for overlap-aware resegmentation,” 2021. [Online]. Available: https://arxiv.org/abs/2104.04045

    arXiv 2021

  2. [16]

    Large-Scale Learning on Overlapped Speech Detection: New Benchmark and New General System

    Z. Yin, J. Tian, X. Hu, X. Xu, and Y . Xiang, “Large-scale learning on overlapped speech detection: New benchmark and new general system,” 2023. [Online]. Available: https: //arxiv.org/abs/2308.05987

    work page Pith review arXiv 2023

  3. [1]

    Overlapping regions present unique challenges, as they involve both overlap of speech segments and speaker identities

    Introduction Multi-party conversation recognition is one of the most chal- lenging problems in speech processing [1], particularly due to difficulties in accurately identifying overlapping speech re- gions. Overlapping regions present unique challenges, as they involve both overlap of speech segments and speaker identities. Effectively addressing these ch...

    work page

  4. [2]

    We be- gin with an overview of the progressive OSD architecture, fol- lowed by a detailed explanation of the progressive OSD training strategy

    Methods In this section, we provide a comprehensive description of the proposed speaker-aware progressive OSD framework. We be- gin with an overview of the progressive OSD architecture, fol- lowed by a detailed explanation of the progressive OSD training strategy. Finally, we present a thorough discussion of the indi- vidual modules within the system. 2.1...

    work page

  5. [3]

    Experiments setup In this section, we describe the experimental setup used for our paper, including the datasets and training configurations

    Experiments 3.1. Experiments setup In this section, we describe the experimental setup used for our paper, including the datasets and training configurations. 3.1.1. Data Corpus An ideal OSD system should be robust to variations in language, speaking style, and speaker-to-microphone distance. To en- hance generalization, we compile a diverse multi-speaker...

    work page

  6. [4]

    Ex- perimental results consistently demonstrate the advantages of this approach across different configurations

    Conclusion In this paper, we propose a speaker-aware progressive OSD model that leverages a progressive learning strategy to effec- tively exploit the correlations between subtasks, thereby en- hancing the robustness and performance of OSD systems. Ex- perimental results consistently demonstrate the advantages of this approach across different configurati...

    work page 2025

  7. [5]

    Detection of overlapping speech for the purposes of speaker diarization,

    M. Kune ˇsov´a, M. Hr ´uz, Z. Zaj ´ıc, and V . Radov´a, “Detection of overlapping speech for the purposes of speaker diarization,” in Speech and Computer, A. A. Salah, A. Karpov, and R. Potapova, Eds. Cham: Springer International Publishing, 2019, pp. 247– 257

    work page 2019

  8. [6]

    Overlap-aware diarization: resegmentation using neural end-to-end overlapped speech detection

    L. Bullock, H. Bredin, and L. P. Garcia-Perera, “Overlap- aware diarization: resegmentation using neural end-to-end overlapped speech detection,” 2019. [Online]. Available: https: //arxiv.org/abs/1910.11646

    work page Pith review arXiv 2019

Show all 33 references
  1. [7]

    Diarization is hard: Some experiences and lessons learned for the jhu team in the inaugural dihard challenge,

    G. Sell, D. Snyder, A. McCree, D. Garcia-Romero, J. Villalba, M. Maciejewski, V . Manohar, N. Dehak, D. Povey, S. Watanabe, and S. Khudanpur, “Diarization is hard: Some experiences and lessons learned for the jhu team in the inaugural dihard challenge,” inInterspeech 2018, 201...

    2018

  2. [8]

    Overlapped speech detection for improved speaker diarization in multiparty meetings,

    K. Boakye, B. Trueba-Hornero, O. Vinyals, and G. Friedland, “Overlapped speech detection for improved speaker diarization in multiparty meetings,” in2008 IEEE International Conference on Acoustics, Speech and Signal Processing, 2008, pp. 4353–4356

    2008

  3. [9]

    Impact of overlapping speech detection on speaker diarization for broadcast news and debates,

    D. Charlet, C. Barras, and J.-S. Li ´enard, “Impact of overlapping speech detection on speaker diarization for broadcast news and debates,” in2013 IEEE International Conference on Acoustics, Speech and Signal Processing, 2013, pp. 7707–7711

    2013

  4. [10]

    But system for dihard speech diariza- tion challenge 2018,

    M. Diez, F. Landini, L. Burget, J. Rohdin, A. Silnova, K. ˇZmol´ıkov´a, O. Novotn ´y, K. Vesel ´y, O. Glembek, O. Plchot, L. Moˇsner, and P. Matˇejka, “But system for dihard speech diariza- tion challenge 2018,” inInterspeech 2018, 2018, pp. 2798–2802

    2018

  5. [11]

    Over- lapped speech detection in broadcast streams using x-vectors

    L. Mateju, F. Kynych, P. Cerva, J. Malek, and J. Zd ´ansk`y, “Over- lapped speech detection in broadcast streams using x-vectors.” in INTERSPEECH, 2022, pp. 4606–4610

    2022

  6. [12]

    Libriheavymix: a 20,000-hour dataset for single-channel reverberant multi-talker speech separation, asr and speaker diarization,

    Z. Jin, Y . Yang, M. Shi, W. Kang, X. Yang, Z. Yao, F. Kuang, L. Guo, L. Meng, L. Linet al., “Libriheavymix: a 20,000-hour dataset for single-channel reverberant multi-talker speech separation, asr and speaker diarization,”arXiv preprint arXiv:2409.00819, 2024

    2024 arXiv

  7. [13]

    Spatial-aware speaker diarization for multi- channel multi-party meeting,

    J. Wang, Y . Liu, B. Wang, Y . Zhi, S. Li, S. Xia, J. Zhang, F. Tong, L. Li, and Q. Hong, “Spatial-aware speaker diarization for multi- channel multi-party meeting,”arXiv preprint arXiv:2209.12002, 2022

    2022 arXiv

  8. [14]

    In [9], Yin et al

    and AMI [15], also serve as benchmarks. In [9], Yin et al. propose a large-scale benchmark, with filtered open-source datasets. Multi-task learning (MTL) [16] has shown promise in im- proving OSD performance by integrating related tasks, such as voice activity detection (V AD)...

    2025 arXiv

  9. [17]

    Automatic detection of multi- speaker fragments with high time resolution,

    E. Kazimirova and A. Belyaev, “Automatic detection of multi- speaker fragments with high time resolution,” inProceedings of the Annual Conference of the International Speech Communica- tion Association, INTERSPEECH, 2018, pp. 1388–1392

    2018

  10. [18]

    Beamtransformer: Microphone array-based overlap- ping speech detection,

    S. Zheng, S. Zhang, W. Huang, Q. Chen, H. Suo, M. Lei, J. Feng, and Z. Yan, “Beamtransformer: Microphone array-based overlap- ping speech detection,”arXiv preprint arXiv:2109.04049, 2021

    2021 arXiv

  11. [19]

    Multitask detection of speaker changes, overlapping speech and voice activity using wav2vec 2.0,

    M. Kune ˇsov´a and Z. Zaj ´ıc, “Multitask detection of speaker changes, overlapping speech and voice activity using wav2vec 2.0,” inICASSP 2023-2023 IEEE International Conference on Acoustics, Speech and Signal Processing (ICASSP). IEEE, 2023, pp. 1–5

    2023

  12. [20]

    M2met: The icassp 2022 multi- channel multi-party meeting transcription challenge,

    F. Yu, S. Zhang, Y . Fu, L. Xie, S. Zheng, Z. Du, W. Huang, P. Guo, Z. Yan, B. Maet al., “M2met: The icassp 2022 multi- channel multi-party meeting transcription challenge,” inICASSP 2022-2022 IEEE International Conference on Acoustics, Speech and Signal Processing (ICASSP). I...

    2022

  13. [21]

    Recognition and under- standing of meetings the ami and amida projects,

    S. Renals, T. Hain, and H. Bourlard, “Recognition and under- standing of meetings the ami and amida projects,” in2007 IEEE Workshop on Automatic Speech Recognition & Understanding (ASRU), 2007, pp. 238–247

    2007

  14. [22]

    A survey on multi-task learning,

    Y . Zhang and Q. Yang, “A survey on multi-task learning,”IEEE transactions on knowledge and data engineering, vol. 34, no. 12, pp. 5586–5609, 2021

    2021

  15. [23]

    Three-class overlapped speech detection using a convolutional recurrent neural network,

    J.-w. Jung, H.-S. Heo, Y . Kwon, J. S. Chung, and B.-J. Lee, “Three-class overlapped speech detection using a convolutional recurrent neural network,”arXiv preprint arXiv:2104.02878, 2021

    2021 arXiv

  16. [24]

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

    I. Medennikov, M. Korenevsky, T. Prisyach, Y . Khokhlov, M. Ko- renevskaya, I. Sorokin, T. Timofeeva, A. Mitrofanov, A. An- drusenko, I. Podluzhny, A. Laptev, and A. Romanenko, “Target- speaker voice activity detection: A novel approach for multi- speaker diarization in a dinn...

  17. [25]

    A survey on text-dependent and text-independent speaker verification,

    Y . Tu, W. Lin, and M.-W. Mak, “A survey on text-dependent and text-independent speaker verification,”IEEE Access, vol. 10, pp. 99 038–99 049, 2022

    2022

  18. [26]

    In this paper, we adapt WavLM for OSD tasks

    and spoofed audio detection [27]. In this paper, we adapt WavLM for OSD tasks. As illustrated in Fig. 1, the WavLM en- coder converts input audio signals into a sequence of high-level abstract representations, serving as the foundation for subse- quent processing. The input au...

  19. [27]

    Ecapa-tdnn embeddings for speaker di- arization,

    N. Dawalatabad, M. Ravanelli, F. Grondin, J. Thienpondt, B. De- splanques, and H. Na, “Ecapa-tdnn embeddings for speaker di- arization,”arXiv preprint arXiv:2104.01466, 2021

    2021 arXiv

  20. [28]

    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,”arXiv preprint arXiv:2303.00332, 2023

    2023 arXiv

  21. [29]

    Wavlm: Large-scale self- supervised pre-training for full stack speech processing,

    S. Chen, C. Wang, Z. Chen, Y . Wu, S. Liu, Z. Chen, J. Li, N. Kanda, T. Yoshioka, X. Xiaoet al., “Wavlm: Large-scale self- supervised pre-training for full stack speech processing,”IEEE Journal of Selected Topics in Signal Processing, vol. 16, no. 6, pp. 1505–1518, 2022

    2022

  22. [30]

    Frame-wise and overlap-robust speaker em- beddings for meeting diarization,

    T. Cord-Landwehr, C. Boeddeker, C. Zoril ˘a, R. Doddipatla, and R. Haeb-Umbach, “Frame-wise and overlap-robust speaker em- beddings for meeting diarization,” inICASSP 2023-2023 IEEE International Conference on Acoustics, Speech and Signal Pro- cessing (ICASSP). IEEE, 2023, pp. 1–5

    2023

  23. [31]

    Joint speech and overlap detection: a benchmark over multiple audio setup and speech domains,

    M. Lebourdais, T. Mariotte, M. Tahon, A. Larcher, A. Laurent, S. Montresor, S. Meignier, and J.-H. Thomas, “Joint speech and overlap detection: a benchmark over multiple audio setup and speech domains,”arXiv preprint arXiv:2307.13012, 2023

    2023 arXiv

  24. [32]

    Unified audio event detection,

    Y . Jiang, R. Tao, W. Huang, Q. Chen, and W. Wang, “Unified audio event detection,” 2024. [Online]. Available: https://arxiv.org/abs/2409.08552

    2024 arXiv

  25. [33]

    A robust audio deepfake detection system via multi-view feature,

    Y . Yang, H. Qin, H. Zhou, C. Wang, T. Guo, K. Han, and Y . Wang, “A robust audio deepfake detection system via multi-view feature,” 2024. [Online]. Available: https: //arxiv.org/abs/2403.01960

    2024 arXiv

Pith tools

Reviewed August 7, 2026 · model on record in the stance chip above.