Pith. sign in

REVIEW 3 major objections 3 minor 36 references

BeatFM: Improving Beat Tracking with Pre-trained Music Foundation Model

T0 review · 3 major / 3 minor · reviewed 2026-08-05 · deepseek-v4-flash

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

Pith's one-line read The paper claims a frozen music foundation model plus a lightweight side network reaches state-of-the-art beat and downbeat tracking, but the abstract and body describe different architectures.

desk verdict Abstract and full text are two different papers—BeatFM is claimed, HingeNet is written; not reviewable as is. read the letter →

arxiv 2508.09790 v2 pith:HDVOHQ32 submitted 2025-08-13 cs.SD

Ganghui Ru , Jieying Wang , Jiahao Zhao , Yulun Wu , Yi Yu , Nannan Jiang , Wei Wang , Wei Li This is my paper
classification cs.SD
keywords beattrackingdownbeatmusicfoundationmodelparameter-efficientfine-tuningharmonic-awaremodulelearnablegatingfrozenencoderinformationretrieval
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 sets out to show that beat tracking can escape its label-scarcity bottleneck without inserting extra modules into a pre-trained model. Its proposal is to keep a music foundation model completely frozen and to attach a small, separately trainable network that reads the model's intermediate representations, converting them into beat and downbeat predictions; the abstract calls this BeatFM, with a plug-and-play module aggregating semantic features in the temporal, frequency, and channel domains, while the full text describes HingeNet, whose core layers use dimensionality-reducing projections, learnable gating, and harmonic-aware dilated convolutions. The paper claims this recipe reaches state-of-the-art beat and downbeat accuracy on GTZAN, Ballroom, Hainsworth, and SMC, and that it beats both adapter- and LoRA-style fine-tuning, which the authors report overfit on scarce beat labels. The experiments and ablations in the body belong to the HingeNet description; the abstract's architecture is not the one tested, so a careful reader must treat the abstract's specific claims as unverified by the supplied evidence.

What carries the argument

The load-bearing object is the HingeNet core layer, a lightweight side network named for its hinge-like silhouette, attached to every encoder of a frozen music foundation model (MERT or MusicFM). Each core layer combines a projection layer that shrinks the encoder output by a factor $r$ to cap the trainable parameter count; a learnable gate $\mu_i=\mathrm{sigmoid}(\alpha_i)$ that fuses the previous core layer's output with the newly projected features; and a harmonic-aware module (HAM) of $M$ parallel 1D convolutional layers with differing dilation rates, concatenated and mapped by an MLP back to the original dimensionality. The design's bet is that the frozen model's intermediate features c

What would settle it

Open the body's model description and the released code: the body implements projection layers, a learnable gate, and harmonic-aware dilated convolutions, and every reported number comes from that design; the abstract promises parallel temporal, frequency, and channel aggregation sub-modules. If the code and checkpoints match the body rather than the abstract, the abstract's architecture claim has no experimental support. Separately, re-running the GTZAN fine-tuning ablation on identical MERT features would verify whether the roughly eight-point beat-F1 gap over LoRA is real.

Watch

Extended reading notes

Core claim

On the paper's own terms, the discovery is that a frozen pre-trained music foundation model can be adapted to beat and downbeat tracking by a lightweight, separable network that consumes each encoder layer's intermediate features. Each core layer projects those features down by a factor of $r$, blends them with the previous layer's output through a learnable gate, and runs them through a harmonic-aware module of parallel 1D dilated convolutions whose concatenated outputs return through an MLP; a linear layer plus dynamic Bayesian network post-processing yields the beats. The paper reports that this recipe, with MERT or MusicFM as the frozen encoder, beats Adapter and LoRA fine-tuning and rea

Load-bearing premise

The whole argument rests on the submission's abstract and body describing the same method: the abstract promises a multi-domain aggregation module, while the body implements a harmonic-aware projection-and-gating side network, and only the latter is exercised by the reported experiments.

Editorial extensions

If this is right

  • The frozen-encoder recipe transfers across backbones: pairing HingeNet with either MERT or MusicFM improves beat and downbeat accuracy over each backbone alone and over the compared state-of-the-art models on GTZAN, Ballroom, Hainsworth, and SMC.
  • Insertion-style fine-tuning is the loser on scarce data: on GTZAN, MERT+HingeNet reaches 89.7 beat F1 against 73.4 for Adapter and 81.2 for LoRA, which the paper attributes to overfitting.
  • Harmonic structure is doing real work: adding the harmonic-aware module improves beat F1 at every tested projection factor, with the best trade-off at $r = 6$.
  • Beat tracking need not modify the foundation model's weights or architecture, keeping the pre-trained representation space intact while adding only the side network's parameters.

Reading between the lines

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

  • A natural next experiment the paper leaves implicit: unfreeze part of the encoder and compare — if performance drops, the frozen-representation design, not extra capacity, carries the gains.
  • The harmonic-shift assumption is independently testable: check whether harmonic-change onsets in the audio align with predicted beats better than generic energy onsets do, rather than judging only end-to-end F1.
  • Because the abstract's multi-domain aggregation module is never implemented in the body's experiments, the benchmark numbers should be attributed to the projection-gating-HAM design only; a reader cannot infer how the temporal-frequency-channel module would score from this submission.
  • If the recipe generalizes, the same side-network pattern should transfer to other label-scarce frame-level MIR tasks, such as tempo estimation or structural segmentation, which share the mismatch between abundant pre-training data and scarce annotation.
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 / 3 minor

Summary. The manuscript as submitted consists of an abstract that announces a novel beat-tracking paradigm called BeatFM, which introduces a pre-trained music foundation model together with a plug-and-play multi-dimensional semantic aggregation module (temporal, frequency, and channel sub-modules), and claims state-of-the-art results on multiple benchmark datasets. The body text, however, is a different paper titled 'HingeNet: A Harmonic-Aware Fine-Tuning Approach for Beat Tracking' and carries arXiv number 2508.09788 rather than 2508.09790. The body defines HingeNet with projection layers, a learnable gating mechanism, and harmonic-aware modules in Eqs. (1)–(4), and reports experiments only for MERT+HingeNet and MusicFM+HingeNet in Tables I–III. The terms 'BeatFM' and 'multi-dimensional semantic aggregation' do not appear anywhere in the body. Consequently, the method promised in the abstract has no architectural specification, no training details, and no experimental evaluation in this document.

Significance. The claimed contribution—BeatFM, combining a frozen music foundation model with a multi-domain semantic aggregation module—would be a relevant and potentially valuable contribution to beat and downbeat tracking if the architecture and experiments were actually presented. The full-text HingeNet experiments (Tables I–III) may indicate a promising fine-tuning approach in their own right, but they do not support the BeatFM paradigm described in the abstract. The manuscript contains no machine-checked proofs, released code, or parameter-free derivations that could offset the absence of the claimed method. As submitted, the central claims of the abstract are not assessable because the body describes a different method and reports results for a different model. The significance of the submission is therefore not supported by its content.

major comments (3)
  1. [Abstract; Full Text Title] The abstract proposes 'BeatFM' with a 'plug-and-play multi-dimensional semantic aggregation module' composed of temporal, frequency, and channel sub-modules. The body is titled 'HingeNet: A Harmonic-Aware Fine-Tuning Approach for Beat Tracking' and carries arXiv number 2508.09788. The method section (Sec. III) defines HingeNet via projection layers, a learnable gating mechanism, and harmonic-aware modules in Eqs. (1)–(4). The terms 'BeatFM', 'multi-dimensional semantic aggregation', and the three parallel sub-modules do not appear in the body. Thus the claimed method has no architectural definition, no training details, and no analysis in this manuscript.
  2. [Tables I–III; Section IV] The experimental section reports results for 'MERT+HingeNet' and 'MusicFM+HingeNet' on GTZAN, Ballroom, Hainsworth, and SMC. These tables cannot substantiate the abstract's state-of-the-art claim for BeatFM, because the evaluated models are HingeNet variants. No table or figure reports results for a 'BeatFM' model, and there is no ablation of the multi-dimensional semantic aggregation module. The central empirical claim of the abstract is therefore unsupported by the evidence contained in the document.
  3. [Abstract vs. Full Text (global)] The Introduction, Method, Experiments, and Conclusion consistently discuss only HingeNet and its harmonic-aware mechanism. There is no passage connecting the abstract's BeatFM paradigm to the body. The discrepancy is not a terminological inconsistency: the two descriptions differ in architectural components, evaluation objects, and paper title. Even if HingeNet is a valid contribution, this submission as written does not contain the paper promised by its abstract.
minor comments (3)
  1. [Section III.A, Eq. (4)] Equation (4) has an extra closing parenthesis: '... H^M_i(\tilde{h}_i)])))' should be '... H^M_i(\tilde{h}_i)])' or similar. The notation for the dimension of h_i also relies on the implicit compatibility of h_{i-1} and h^P_i; making this explicit would improve readability.
  2. [Tables I–II] Table I uses 'Downbeat Accuracy' with F-Measure, CMLt, and AMLt subcolumns, while Table II uses 'Downbeat F1'. For consistency, the metric naming should be unified across tables.
  3. [Related Work, Section II.B] The text refers to 'Music2Vce [25]' but the cited reference [25] is titled 'Map-Music2Vec'. Please correct the name to match the reference.

Circularity Check

0 steps flagged · score 0.0 of 10

No circular derivation in the body (HingeNet); the BeatFM abstract, however, is not supported by the full text, a manuscript-integrity issue rather than circularity.

full rationale

The derivation chain in the submitted full text is HingeNet's: frozen MERT/MusicFM encoders provide intermediate features, which are projected, gated, and processed by harmonic-aware dilated-conv modules, then evaluated against Adapter/LoRA and published beat trackers on external benchmarks (Tables I-III). No load-bearing step reduces by construction: the harmonic-aware design is an explicit music-theoretic assumption, the projection factor r=6 is an ablation-selected hyperparameter, and the SOTA numbers are empirical comparisons, not quantities fitted from the evaluation labels. The only self-citations ([5], [6]) are background examples and are not load-bearing. The serious problem is that the abstract describes a different paper: it defines 'BeatFM' with a 'plug-and-play multi-dimensional semantic aggregation module... temporal, frequency, and channel domains' and claims SOTA, while the full text is titled 'HingeNet: A Harmonic-Aware Fine-Tuning Approach for Beat Tracking' (arXiv:2508.09788v2) and never defines or evaluates BeatFM. This means the abstract's central claim has no supporting derivation or experiments in the body. That is missing support/manuscript integrity, not circularity—no equation or fitted parameter in BeatFM is shown to be equivalent to its own inputs—so the circularity score is 0.

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

No free parameters appear in the abstract. The axioms list the transferability and module-effectiveness assumptions, plus the critical identity assumption that the body text is about BeatFM, which is contradicted by the different title and arXiv number.

assumptions (3)
  • domain assumption Pre-trained music foundation models contain beat-relevant semantic knowledge that can be adapted to beat tracking.
    The abstract's central claim relies on transferability of foundation model representations to beat tracking; this is asserted, not derived.
  • domain assumption The multi-dimensional semantic aggregation module (temporal, frequency, channel) preserves and combines the information needed for beat and downbeat detection.
    This is the core design premise of BeatFM; the abstract gives no mathematical or empirical justification.
  • ad hoc to paper The full text and abstract describe the same method.
    The submission is only coherent if the body supports the abstract; in fact the body is a different paper, so this axiom is violated.

how reviews work

0 comments
Cite this review

Pith. "Pith review of BeatFM: Improving Beat Tracking with Pre-trained Music Foundation Model." pith.science (2026). https://pith.science/paper/HDVOHQ32

@misc{pith2026250809790,
  author       = {Pith},
  title        = {Pith review of: BeatFM: Improving Beat Tracking with Pre-trained Music Foundation Model},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/HDVOHQ32}},
  note         = {Machine review of arXiv:2508.09790}
}
read the original abstract

Beat tracking is a widely researched topic in music information retrieval. However, current beat tracking methods face challenges due to the scarcity of labeled data, which limits their ability to generalize across diverse musical styles and accurately capture complex rhythmic structures. To overcome these challenges, we propose a novel beat tracking paradigm BeatFM, which introduces a pre-trained music foundation model and leverages its rich semantic knowledge to improve beat tracking performance. Pre-training on diverse music datasets endows music foundation models with a robust understanding of music, thereby effectively addressing these challenges. To further adapt it for beat tracking, we design a plug-and-play multi-dimensional semantic aggregation module, which is composed of three parallel sub-modules, each focusing on semantic aggregation in the temporal, frequency, and channel domains, respectively. Extensive experiments demonstrate that our method achieves state-of-the-art performance in beat and downbeat tracking across multiple benchmark datasets.

Discussion (0). Sign in to comment.

Reference graph

Works this paper leans on

36 extracted references · 30 canonical work pages

  1. [1]

    Explicit beat structure modeling for non-negative matrix factorization-based multipitch analysis,

    Kazuki Ochiai, Hirokazu Kameoka, and Shigeki Sagayama, “Explicit beat structure modeling for non-negative matrix factorization-based multipitch analysis,” in ICASSP. IEEE, 2012, pp. 133–136

    work page 2012

  2. [2]

    Bayesian singing transcription based on a hierarchical generative model of keys, musical notes, and f0 trajectories,

    Ryo Nishikimi, Eita Nakamura, Masataka Goto, Katsutoshi Itoyama, and Kazuyoshi Yoshii, “Bayesian singing transcription based on a hierarchical generative model of keys, musical notes, and f0 trajectories,” IEEE/ACM TASLP, vol. 28, pp. 1678–1691, 2020

    work page 2020

  3. [3]

    Music structure analysis based on an lstm-hsmm hybrid model.,

    Go Shibata, Ryo Nishikimi, and Kazuyoshi Yoshii, “Music structure analysis based on an lstm-hsmm hybrid model.,” in ISMIR, 2020, pp. 23–29

    work page 2020

  4. [4]

    Audio-based music structure analysis: Current trends, open challenges, and applications,

    Oriol Nieto, Gautham J. Mysore, Cheng-i Wang, Jordan B. L. Smith, Jan Schl¨uter, Thomas Grill, and Brian McFee, “Audio-based music structure analysis: Current trends, open challenges, and applications,” Transac- tions of the International Society for Music Information Retrieval , p. 246–263, Dec 2020

    work page 2020

  5. [5]

    Improving music genre classification from multi-modal prop- erties of music and genre correlations perspective,

    Ganghui Ru, Xulong Zhang, Jianzong Wang, Ning Cheng, and Jing Xiao, “Improving music genre classification from multi-modal prop- erties of music and genre correlations perspective,” in ICASSP 2023- 2023 IEEE International Conference on Acoustics, Speech and Signal Processing (ICASSP). IEEE, 2023, pp. 1–5

    work page 2023

  6. [6]

    Melody generation from lyrics with local interpretability,

    Wei Duan, Yi Yu, Xulong Zhang, Suhua Tang, Wei Li, and Keizo Oyama, “Melody generation from lyrics with local interpretability,” ACM Transactions on Multimedia Computing, Communications and Applications, vol. 19, no. 3, pp. 1–21, 2023

    work page 2023

  7. [7]

    Joint beat and downbeat tracking with recurrent neural networks.,

    Sebastian B ¨ock, Florian Krebs, and Gerhard Widmer, “Joint beat and downbeat tracking with recurrent neural networks.,” in ISMIR. New York City, 2016, pp. 255–261

    work page 2016

  8. [8]

    Beat and downbeat tracking of symbolic music data using deep recurrent neural networks,

    Yi-Chin Chuang and Li Su, “Beat and downbeat tracking of symbolic music data using deep recurrent neural networks,” in APSIPA ASC . IEEE, 2020, pp. 346–352

    work page 2020

Show all 36 references
  1. [9]

    Robust downbeat tracking using an ensemble of convolutional net- works,

    Simon Durand, Juan Pablo Bello, Bertrand David, and Gael Richard, “Robust downbeat tracking using an ensemble of convolutional net- works,” IEEE/ACM TASLP, p. 76–89, Jan 2017

    2017

  2. [10]

    Joint beat and downbeat tracking based on crnn models and a comparison of using different context ranges in convolutional layers,

    Tian Cheng, Satoru Fukayama, and Masataka Goto, “Joint beat and downbeat tracking based on crnn models and a comparison of using different context ranges in convolutional layers,” in Proc. ICMC, 2020

    2020

  3. [11]

    Multi-task learning of tempo and beat: Learning one to improve the other.,

    Sebastian B ¨ock, Matthew EP Davies, and Peter Knees, “Multi-task learning of tempo and beat: Learning one to improve the other.,” in ISMIR, 2019, pp. 486–493

    2019

  4. [12]

    Deconstruct, analyse, re- construct: How to improve tempo, beat, and downbeat estimation.,

    Sebastian B ¨ock and Matthew EP Davies, “Deconstruct, analyse, re- construct: How to improve tempo, beat, and downbeat estimation.,” in ISMIR, 2020, pp. 574–582

    2020

  5. [13]

    Modeling beats and downbeats with a time-frequency transformer,

    Yun-Ning Hung, Ju-Chiang Wang, Xuchen Song, Wei-Tsung Lu, and Minz Won, “Modeling beats and downbeats with a time-frequency transformer,” in ICASSP. IEEE, 2022, pp. 401–405

    2022

  6. [14]

    Beat transformer: Demixed beat and downbeat tracking with dilated self-attention,

    Jingwei Zhao, Gus Xia, and Ye Wang, “Beat transformer: Demixed beat and downbeat tracking with dilated self-attention,” arXiv preprint arXiv:2209.07140, 2022

    2022 arXiv

  7. [15]

    Transformer-based beat tracking with low-resolution encoder and high-resolution decoder.,

    Tian Cheng and Masataka Goto, “Transformer-based beat tracking with low-resolution encoder and high-resolution decoder.,” in ISMIR, 2023, pp. 466–473

    2023

  8. [16]

    Side-tuning: a baseline for network adaptation via additive side networks,

    Jeffrey O Zhang, Alexander Sax, Amir Zamir, Leonidas Guibas, and Jitendra Malik, “Side-tuning: a baseline for network adaptation via additive side networks,” in ECCV. Springer, 2020, pp. 698–714

    2020

  9. [17]

    Compacter: Efficient low-rank hypercomplex adapter layers,

    Rabeeh Karimi Mahabadi, James Henderson, and Sebastian Ruder, “Compacter: Efficient low-rank hypercomplex adapter layers,” Advances in Neural Information Processing Systems , vol. 34, pp. 1022–1035, 2021

    2021

  10. [18]

    Lora: Low-rank adaptation of large language models,

    Edward J Hu, Yelong Shen, Phillip Wallis, Zeyuan Allen-Zhu, Yuanzhi Li, Shean Wang, Lu Wang, and Weizhu Chen, “Lora: Low-rank adaptation of large language models,” arXiv preprint arXiv:2106.09685 , 2021

    2021 arXiv

  11. [19]

    Lora dropout as a sparsity regularizer for overfitting control,

    Yang Lin, Xinyu Ma, Xu Chu, Yujie Jin, Zhibang Yang, Yasha Wang, and Hong Mei, “Lora dropout as a sparsity regularizer for overfitting control,” arXiv preprint arXiv:2404.09610 , 2024

    2024 arXiv

  12. [20]

    U-beat: A multi-scale beat tracking model based on wave-u-net,

    Tian Cheng and Masataka Goto, “U-beat: A multi-scale beat tracking model based on wave-u-net,” in ICASSP. IEEE, 2023, pp. 1–5

    2023

  13. [21]

    Singing beat tracking with self-supervised front-end and linear transformers,

    Mojtaba Heydari and Zhiyao Duan, “Singing beat tracking with self-supervised front-end and linear transformers,” arXiv preprint arXiv:2208.14578, 2022

    2022 arXiv

  14. [22]

    Zero- note samba: Self-supervised beat tracking,

    Dorian Desblancs, Vincent Lostanlen, and Romain Hennequin, “Zero- note samba: Self-supervised beat tracking,” IEEE/ACM Transactions on Audio, Speech, and Language Processing , 2023

    2023

  15. [23]

    Local periodicity-based beat tracking for expressive classical piano music,

    Ching-Yu Chiu, Meinard M ¨uller, Matthew EP Davies, Alvin Wen-Yu Su, and Yi-Hsuan Yang, “Local periodicity-based beat tracking for expressive classical piano music,” IEEE/ACM Transactions on Audio, Speech, and Language Processing , 2023

    2023

  16. [24]

    Beat this! accurate beat tracking without dbn postprocessing,

    Francesco Foscarin, Jan Schl ¨uter, and Gerhard Widmer, “Beat this! accurate beat tracking without dbn postprocessing,” arXiv preprint arXiv:2407.21658, 2024

    2024 arXiv

  17. [25]

    Map- music2vec: A simple and effective baseline for self-supervised music audio representation learning,

    Yizhi Li, Ruibin Yuan, Ge Zhang, Yinghao Ma, Chenghua Lin, Xingran Chen, Anton Ragni, Hanzhi Yin, Zhijie Hu, Haoyu He, et al., “Map- music2vec: A simple and effective baseline for self-supervised music audio representation learning,” arXiv preprint arXiv:2212.02508 , 2022

    2022 arXiv

  18. [26]

    Mert: Acoustic music understanding model with large-scale self- supervised training,

    LI Yizhi, Ruibin Yuan, Ge Zhang, Yinghao Ma, Xingran Chen, Hanzhi Yin, Chenghao Xiao, Chenghua Lin, Anton Ragni, Emmanouil Benetos, et al., “Mert: Acoustic music understanding model with large-scale self- supervised training,” in ICLR, 2023

    2023

  19. [27]

    A foundation model for music informatics,

    Minz Won, Yun-Ning Hung, and Duc Le, “A foundation model for music informatics,” in ICASSP 2024-2024 IEEE International Conference on Acoustics, Speech and Signal Processing (ICASSP) . IEEE, 2024, pp. 1226–1230

    2024

  20. [28]

    Self-supervised learning with random-projection quantizer for speech recognition,

    Chung-Cheng Chiu, James Qin, Yu Zhang, Jiahui Yu, and Yonghui Wu, “Self-supervised learning with random-projection quantizer for speech recognition,” in International Conference on Machine Learning . PMLR, 2022, pp. 3915–3924

    2022

  21. [29]

    An efficient state- space model for joint tempo and meter tracking.,

    Florian Krebs, Sebastian B ¨ock, and Gerhard Widmer, “An efficient state- space model for joint tempo and meter tracking.,” in ISMIR, 2015, pp. 72–78

    2015

  22. [30]

    Evalua- tion methods for musical audio beat tracking algorithms,

    Matthew EP Davies, Norberto Degara, and Mark D Plumbley, “Evalua- tion methods for musical audio beat tracking algorithms,” Queen Mary University of London, Centre for Digital Music, Tech. Rep. C4DM-TR- 09-06, 2009

    2009

  23. [31]

    Rwc music database: Popular, classical and jazz music databases.,

    Masataka Goto, Hiroki Hashiguchi, Takuichi Nishimura, and Ryuichi Oka, “Rwc music database: Popular, classical and jazz music databases.,” in ISMIR, 2002, vol. 2, pp. 287–288

    2002

  24. [32]

    Particle filtering applied to musical tempo tracking,

    Stephen W. Hainsworth and Malcolm D. Macleod, “Particle filtering applied to musical tempo tracking,” EURASIP Journal on Advances in Signal Processing, vol. 2004, no. 15, Nov 2004

    2004

  25. [33]

    Rhythmic pattern modeling for beat and downbeat tracking in musical audio.,

    Florian Krebs, Sebastian B ¨ock, and Gerhard Widmer, “Rhythmic pattern modeling for beat and downbeat tracking in musical audio.,” in ISMIR. Citeseer, 2013, pp. 227–232

    2013

  26. [34]

    The harmonix set: Beats, downbeats, and functional segment annotations of western popular music.,

    Oriol Nieto, Matthew C McCallum, Matthew EP Davies, Andrew Robertson, Adam M Stark, and Eran Egozy, “The harmonix set: Beats, downbeats, and functional segment annotations of western popular music.,” in ISMIR, 2019, pp. 565–572

    2019

  27. [35]

    Selective sampling for beat tracking evaluation,

    Andr ´e Holzapfel, Matthew E. P. Davies, Jos ´e R. Zapata, Jo ˜ao Lobato Oliveira, and Fabien Gouyon, “Selective sampling for beat tracking evaluation,” IEEE TASLP, vol. 20, no. 9, pp. 2539–2548, Nov 2012

    2012

  28. [36]

    Swing ratio estimation,

    Ugo Marchand and Geoffroy Peeters, “Swing ratio estimation,” Nov 2015

    2015

Pith tools

Reviewed August 5, 2026 · model on record in the stance chip above.