Pith. sign in

REVIEW 4 major objections 7 minor 39 references

Episode-specific Fine-tuning for Metric-based Few-shot Learners with Optimization-based Training

T0 review · 4 major / 7 minor · reviewed 2026-08-15 · deepseek-v4-flash

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

Pith's one-line read The paper claims that episode-specific fine-tuning of the support set, trained with Meta-Curvature, consistently improves metric-based few-shot audio classifiers across three datasets.

desk verdict Useful combined fine-tuning + Meta-Curvature recipe for audio few-shot learning, but the paper never isolates fine-tuning from MC training and the tables contain internal inconsistencies that undercut the exact gains. read the letter →

arxiv 2506.17499 v1 pith:IXN5GVFR submitted 2025-06-20 cs.LG cs.MMcs.SD

Xuanyu Zhuang , Geoffroy Peeters , Gaël Richard This is my paper · ORCID
classification cs.LGcs.MMcs.SD
keywords few-shotclassificationmetric-basedmodelsepisode-specificfine-tuningMeta-Curvaturemeta-learningaudioPrototypicalNetworksCrossAttention
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

Metric-based few-shot classifiers use the labeled support set only as comparison points, even though those labels could adapt the model to the current episode. The paper proposes three during-inference fine-tuning procedures—Rotational, Iterative, and Augmented Division Fine-Tuning—that carve the support set into pseudo support/query pairs, giving a supervised loss where none existed because the real query set is unlabeled. Because each episode has very few samples, the paper couples these procedures with Meta-Curvature, an optimization-based meta-learning method whose learnable curvature matrix makes few-step fine-tuning stable. Across ESC-50, Speech Commands V2, and Medley-solos-DB, the combination beats the vanilla baselines for Prototypical, Matching, and Cross Attention Networks, with the largest gains for the attention-based model. If true, this offers a simple recipe for improving existing metric-based few-shot audio classifiers without changing architecture or data.

What carries the argument

The load-bearing object is the pseudo support/query division of the support set, operationalized as Rotational Division Fine-Tuning: for a K-way-N-shot support set, hold out one shot as a pseudo query set and use the remaining K-way-(N-1)-shot set as pseudo support, repeating N times so every sample serves once as pseudo query. This constructed task supplies a supervised loss where none existed, since the true query set is unlabeled. The second mechanism is Meta-Curvature's learnable curvature matrix M, which transforms the gradient in each inner-update step; training the metric model inside this optimization-based loop makes the fine-tuning converge in a few steps without overfitting on the K times N labeled samples.

What would settle it

Run the same fine-tuning procedures on episodes where the real query set is drawn from classes or recording conditions deliberately different from the support set, for example support from one set of speakers or instruments and query from unseen speakers or timbres. If accuracy relative to the vanilla baseline falls to zero or goes negative in those mismatched episodes, the pseudo-query proxy is not faithful and the method's benefit depends on support-query similarity rather than on general adaptation.

Watch

Extended reading notes

Core claim

The central discovery is that the support set can be recycled into pseudo episodes—holding out one shot as a pseudo query set and fine-tuning on the remainder—and that doing so during inference, while training the base model with Meta-Curvature, reshapes the embedding space toward the classes of the current episode. This makes even non-parametric metric models like Prototypical and Matching Networks adaptable at test time. The paper reports consistent accuracy gains over vanilla baselines for all three models on all three audio datasets, with gains up to about eight percentage points. For Cross Attention Networks, fine-tuning also redistributes the attention maps, which initially concentrate on only part of the feature space; this joint adaptation of embedding and attention explains the especially large improvements for attention-based models.

Load-bearing premise

The method assumes that the pseudo query samples carved out of the support set are a faithful stand-in for the real query samples, so fine-tuning on them points the model in the right direction; if real queries are systematically more diverse or come from different conditions, the adaptation could be misdirected.

Editorial extensions

If this is right

  • Vanilla Prototypical, Matching, and Cross Attention Networks all improve when their support set is used for episode-specific fine-tuning under Meta-Curvature training.
  • Larger support sets yield larger gains: the 25-sample episodes on ESC-50 and Speech Commands V2 improve more than the 15-sample episodes on Medley-solos-DB.
  • Attention-based models benefit most, because fine-tuning adapts not only the embedding space but also the attention maps to the current episode's classes.
  • ADFT with a well-chosen audio augmentation is the best overall variant, while replication-only ADFT is safer than risky augmentations for non-attention models.
  • The episodic training principle matters: the variant that violates it (IDFT) consistently underperforms the variant that adheres to it (ADFT).

Reading between the lines

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

  • A testable extension is to apply the same pseudo-division fine-tuning to image or text metric-based few-shot learners; the paper's mechanism is architecture-agnostic, but it only evaluates audio.
  • The attention-map visualizations suggest a concrete diagnostic: track attention entropy before and after fine-tuning; if gains disappear when attention is already uniform, the benefit may be mainly variance reduction rather than class-specific focus.
  • Because ADFT's replication can bias prototypes, augmentation that restores within-class variance may be a general recipe for any metric learner whose prediction depends on a similarity matrix, not only audio models.
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

4 major / 7 minor

Summary. The paper proposes episode-specific fine-tuning methods for metric-based few-shot classifiers. The core idea is to construct pseudo support/query splits from the labeled support set at inference time (Rotational Division Fine-Tuning, RDFT, and variants IDFT and ADFT) and to train the metric-based model within a Meta-Curvature optimization-based meta-learning framework so that the model can adapt to the support set without overfitting. The methods are evaluated on three audio datasets (ESC-50, Speech Commands V2, Medley-solos-DB) with Prototypical Networks, Matching Networks, and Cross Attention Networks. The authors report that the combined MC+fine-tuning systems consistently outperform vanilla metric-based baselines, with especially large gains for attention-based CAN models, and that augmentation choices in ADFT can further improve CAN results.

Significance. If the central claim holds, the proposed framework is a simple and general recipe for adding inference-time adaptation to metric-based few-shot learners, with potential value across audio few-shot classification tasks. The paper is strong in breadth: three model families, three datasets from different audio domains, and a public code repository are provided, which supports reproducibility. The main weakness is that the headline numerical gains conflate two different interventions, optimization-based meta-training and episode-specific fine-tuning, so the specific contribution of the proposed fine-tuning methods is not cleanly established by the reported comparisons. Additional ablations and a correction of the inconsistent tables would make the empirical case substantially stronger.

major comments (4)
  1. [Section IV-C, Tables II-IV] The column labeled 'Gain' is computed as vanilla-baseline accuracy minus the accuracy of the Meta-Curvature model after fine-tuning, so it does not measure the contribution of the proposed fine-tuning methods. On ESC-50, MC-PN-RDFT already reaches 84.12% without inference-time fine-tuning, versus 82.34% for vanilla PN; the reported +3.63% gain therefore includes a +1.78% contribution from MC training alone, and the fine-tuning-only increment is +1.85%. For MC-CAN-RDFT, MC training without fine-tuning is 81.67%, below the vanilla CAN 83.01%, so the reported +5.22% is an interaction effect rather than a property of RDFT. No condition trains a metric-based model with Meta-Curvature but without RDFT/IDFT/ADFT in the inner loop, so Contributions 1-3 are not isolated. Please add this ablation and compute fine-tuning-only gains from w/o-FT to w/FT columns, and rework the text that attributes the gains to the fine-tuning method.
  2. [Tables IV and V] The Medley-solos-DB numbers are internally inconsistent across the two tables. MC-PN-RDFT is reported as 81.95% with fine-tuning in Table IV but 83.93% with fine-tuning in Table V; MC-CAN-RDFT is 73.14% without fine-tuning in Table IV but 75.81% without fine-tuning in Table V. Since the abstract and Section IV-C rely on these exact magnitudes, the authors must reconcile the two tables and explain which numbers correspond to the official evaluation.
  3. [Section IV-B/IV-C, Tables III and V] For Prototypical Networks on Speech Commands V2, the fine-tuning-only effect is numerically negligible: RDFT moves from 86.84% to 86.94% and ADFT from 87.02% to 87.07%, both within the reported 95% confidence intervals of roughly ±0.13%. No significance tests are reported anywhere. The abstract's claim that the approach 'consistently improves performance for all evaluated metric-based models' is therefore not supported for these cells; either add a significance analysis or soften the claim.
  4. [Section IV-F, Tables VI-VII] The best audio augmentation for each dataset is selected from Table VI, which reports accuracy on the same test sets used for the final evaluation in Table VII. Choosing augmentation policies based on test-set performance and then reporting those test numbers as the result makes the gains optimistic. The selection should be made on a validation split (or fixed in advance), or the authors should disclose that Table VII is a test-set selection result.
minor comments (7)
  1. [Algorithm 1] The loop 'for j = 0 to size(S_i)' is ambiguous: a pseudo query set should contain one shot (one sample per class), but the notation {S_ij} suggests a single sample, and S_{i,j-1} is undefined for j=0. Please clarify whether j indexes shots or samples and define the indexing convention.
  2. [Section I, line 4] There is a typo: 'spoken keywors' should be 'spoken keywords'.
  3. [Section II-B, Eq. (6)] The displayed formula for the meta fusion layer has unbalanced parentheses; the closing parenthesis for W_2 is missing.
  4. [References] References [30] and [36] are the same paper (Warden, Speech Commands); please consolidate or cite distinct versions.
  5. [Section III-D] The statement that 'the total weight of each sample remains equal' is only true for the per-iteration prototype construction, not for the final model parameters after sequential gradient updates; the order of updates matters.
  6. [Figure 2] The caption says the x- and y-axes are the two principal components of the features, but the figure itself is not shown in the text; please ensure the figure and caption clearly explain what is plotted and how the attention maps are visualized.
  7. [Section III-B] The paper would be strengthened by an analysis validating that pseudo queries constructed from the support set are a faithful proxy for real queries; currently only the final accuracy numbers support this assumption.

Circularity Check

0 steps flagged · score 1.0 of 10

No significant circularity: the empirical gains are evaluated on held-out classes and are not derived from their own inputs, though the Gain column conflates Meta-Curvature training with fine-tuning.

full rationale

The paper contains no mathematical derivation in which an output is defined as its own input. RDFT, IDFT, and ADFT construct pseudo support and query sets from the support set (Section III-B), and the meta-training loop in Section III-E updates initial parameters and curvature using the true query loss after inner-loop adaptation (Algorithm 1). Reported accuracies are measured on test classes disjoint from the training classes, with vanilla PN, MN, and CAN as independently trained baselines, so the headline gains are not equivalent to the training loss by construction. The prior self-citation [24] proposes RDFT for Prototypical Networks and is used as motivation, but the present paper re-tests RDFT on new architectures and datasets rather than importing the conclusion as evidence. The main caveat is attribution, not circularity: the Gain column in Tables II--IV compares the vanilla baseline to the full MC+RDFT system, so it does not isolate the fine-tuning procedure from Meta-Curvature training; and because Section III-E explicitly optimizes for post-fine-tuning performance, part of the improvement seen after fine-tuning is anticipated by the training objective. That is a research-design confound and a correctness risk, not a reduction by construction. There is also an internal inconsistency between Tables IV and V for Medley-solos-DB (e.g., MC-PN-RDFT w/ FT is reported as 81.95% versus 83.93%, and MC-CAN-RDFT w/o FT as 73.14% versus 75.81%), which further weakens the exact gain magnitudes but does not constitute circularity. Weighing all of this, the central claim retains independent empirical content, so the circularity score is low.

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

No new physical entities are produced. The paper's free parameters are standard meta-learning hyperparameters and augmentation strengths, and the key load-bearing assumptions concern the validity of pseudo tasks and the class-preserving nature of augmentations.

free parameters (4)
  • inner learning rate alpha = 0.2 for ESC-50 and Medley-solos-DB, 0.02 for Speech Commands V2
    Chosen from prior work [2], not fitted to test data; affects fine-tuning convergence.
  • meta learning rate beta = 1e-3
    Standard hyperparameter for Meta-Curvature, inherited from prior work.
  • number of fine-tuning steps n = 8
    Fixed for all experiments; controls how much adaptation occurs during each episode.
  • augmentation scale s = 1 for additive noise, 4 for frequency filter and pitch shifting
    Controls augmentation strength and was chosen by the authors without a principled rule.
assumptions (3)
  • domain assumption Constructed pseudo support-query pairs from the support set are a valid training signal for adapting the embedding space.
    The whole method relies on this; Section III-A/B state it but provide no theoretical or empirical validation of the pseudo-task distribution.
  • domain assumption Applying audio augmentation to a replicated support sample preserves class identity and adds useful variance.
    Stated in Section III-D; the paper acknowledges it can fail for pitch shifting on instruments in Section IV-E.
  • domain assumption Episodic training principle, matching the number of shots between training and evaluation, is beneficial for few-shot performance.
    The paper uses this to interpret why IDFT underperforms and ADFT outperforms; it is inherited from few-shot learning literature.

how reviews work

0 comments
Cite this review

Pith. "Pith review of Episode-specific Fine-tuning for Metric-based Few-shot Learners with Optimization-based Training." pith.science (2026). https://pith.science/paper/IXN5GVFR

@misc{pith2026250617499,
  author       = {Pith},
  title        = {Pith review of: Episode-specific Fine-tuning for Metric-based Few-shot Learners with Optimization-based Training},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/IXN5GVFR}},
  note         = {Machine review of arXiv:2506.17499}
}
read the original abstract

In few-shot classification tasks (so-called episodes), a small set of labeled support samples is provided during inference to aid the classification of unlabeled query samples. Metric-based models typically operate by computing similarities between query and support embeddings within a learned metric space, followed by nearest-neighbor classification. However, these labeled support samples are often underutilized--they are only used for similarity comparison, despite their potential to fine-tune and adapt the metric space itself to the classes in the current episode. To address this, we propose a series of simple yet effective episode-specific, during-inference fine-tuning methods for metric-based models, including Rotational Division Fine-Tuning (RDFT) and its two variants, Iterative Division Fine-Tuning (IDFT) and Augmented Division Fine-Tuning (ADFT). These methods construct pseudo support-query pairs from the given support set to enable fine-tuning even for non-parametric models. Nevertheless, the severely limited amount of data in each task poses a substantial risk of overfitting when applying such fine-tuning strategies. To mitigate this, we further propose to train the metric-based model within an optimization-based meta-learning framework. With the combined efforts of episode-specific fine-tuning and optimization-based meta-training, metric-based models are equipped with the ability to rapidly adapt to the limited support samples during inference while avoiding overfitting. We validate our approach on three audio datasets from diverse domains, namely ESC-50 (environmental sounds), Speech Commands V2 (spoken keywords), and Medley-solos-DB (musical instrument). Experimental results demonstrate that our approach consistently improves performance for all evaluated metric-based models (especially for attention-based models) and generalizes well across different audio domains.

Figures

Figures reproduced from arXiv: 2506.17499 by the authors.

Figure 1
Figure 1. The illustration of the three proposed fine-tuning methods for metric-based few-shot learners with the example of a [PITH_FULL_IMAGE:figures/full_fig_p005_1.png] view at source ↗
Figure 2
Figure 2. The cross attention map of CAN on the extracted [PITH_FULL_IMAGE:figures/full_fig_p008_2.png] view at source ↗
Figure 3
Figure 3. An illustration of how audio augmentation changes the [PITH_FULL_IMAGE:figures/full_fig_p009_3.png] view at source ↗

Review history (2 revisions) →

Discussion (0). Continue with ORCID to comment.

Reference graph

Works this paper leans on

39 extracted references · 20 canonical work pages

  1. [1]

    A closer look at few-shot classification,

    W.-Y . Chen, Y .-C. Liu, Z. Kira, Y .-C. F. Wang, and J.-B. Huang, “A closer look at few-shot classification,”arXiv preprint arXiv:1904.04232, 2019

    arXiv 1904

  2. [2]

    Metaaudio: A few-shot audio classification benchmark,

    C. Heggan, S. Budgett, T. Hospedales, and M. Yaghoobi, “Metaaudio: A few-shot audio classification benchmark,” inInternational Conference on Artificial Neural Networks. Springer, 2022, pp. 219–230

    work page 2022

  3. [3]

    Few- shot continual learning for audio classification,

    Y . Wang, N. J. Bryan, M. Cartwright, J. P. Bello, and J. Salamon, “Few- shot continual learning for audio classification,” inIEEE International Conference on Acoustics, Speech and Signal Processing (ICASSP), 2021, pp. 321–325

    work page 2021

  4. [4]

    Few-shot class-incremental audio classification via discriminative prototype learning,

    W. Xie, Y . Li, Q. He, and W. Cao, “Few-shot class-incremental audio classification via discriminative prototype learning,”Expert Systems with Applications, vol. 225, p. 120044, 2023

    work page 2023

  5. [5]

    Few-shot audio classification with attentional graph neural networks

    S. Zhang, Y . Qin, K. Sun, and Y . Lin, “Few-shot audio classification with attentional graph neural networks.” inInterspeech, 2019, pp. 3649–3653

    work page 2019

  6. [6]

    Learning from taxonomy: Multi- label few-shot classification for everyday sound recognition,

    J. Liang, H. Phan, and E. Benetos, “Learning from taxonomy: Multi- label few-shot classification for everyday sound recognition,” in2024 IEEE International Conference on Acoustics, Speech and Signal Pro- cessing (ICASSP). IEEE, 2024, pp. 771–775

    work page 2024

  7. [7]

    Few-shot sound event detection,

    Y . Wang, J. Salamon, N. J. Bryan, and J. Pablo Bello, “Few-shot sound event detection,” inIEEE International Conference on Acoustics, Speech and Signal Processing (ICASSP), 2020, pp. 81–85

    work page 2020

  8. [9]

    Active few-shot learning for sound event detection

    Y . Wang, M. Cartwright, and J. P. Bello, “Active few-shot learning for sound event detection.” inInterspeech, 2022, pp. 1551–1555

    work page 2022

Show all 39 references
  1. [10]

    Few-shot acoustic event detection via meta learning,

    B. Shi, M. Sun, K. C. Puvvada, C.-C. Kao, S. Matsoukas, and C. Wang, “Few-shot acoustic event detection via meta learning,” in2020 IEEE International Conference on Acoustics, Speech and Signal Processing (ICASSP). IEEE, 2020, pp. 76–80

    2020

  2. [11]

    Few-shot bioacoustic event detection at the dcase 2022 challenge,

    I. Nolasco, S. Singh, E. Vidana-Villa, E. Grout, J. Morford, M. Emmer- son, F. Jensens, H. Whitehead, I. Kiskin, A. Strandburg-Peshkinet al., “Few-shot bioacoustic event detection at the dcase 2022 challenge,” arXiv preprint arXiv:2207.07911, 2022

    2022 arXiv

  3. [12]

    Overview of birdclef 2022: Endangered bird species recognition in soundscape recordings

    S. Kahl, A. Navine, T. Denton, H. Klinck, P. Hart, H. Glotin, H. Go ¨eau, W.-P. Vellinga, R. Planqu ´e, and A. Joly, “Overview of birdclef 2022: Endangered bird species recognition in soundscape recordings.” inCLEF (Working Notes), 2022, pp. 1929–1939

    2022

  4. [13]

    Few-shot musical source separation,

    Y . Wang, D. Stoller, R. M. Bittner, and J. P. Bello, “Few-shot musical source separation,” in2022 IEEE International Conference on Acoustics, Speech and Signal Processing (ICASSP). IEEE, 2022, pp. 121–125

    2022

  5. [14]

    Leveraging hierarchical structures for few-shot musical instrument recognition,

    H. F. Garcia, A. Aguilar, E. Manilow, and B. Pardo, “Leveraging hierarchical structures for few-shot musical instrument recognition,” arXiv preprint arXiv:2107.07029, 2021

    2021 arXiv

  6. [15]

    Lc-protonets: Multi- label few-shot learning for world music audio tagging,

    C. Papaioannou, E. Benetos, and A. Potamianos, “Lc-protonets: Multi- label few-shot learning for world music audio tagging,”IEEE Open Journal of Signal Processing, 2025. 11

    2025

  7. [16]

    Few-shot drum transcription in polyphonic music,

    Y . Wang, J. Salamon, M. Cartwright, N. J. Bryan, and J. P. Bello, “Few-shot drum transcription in polyphonic music,”arXiv preprint arXiv:2008.02791, 2020

    2008 arXiv

  8. [17]

    Prototypical networks for few-shot learning,

    J. Snell, K. Swersky, and R. Zemel, “Prototypical networks for few-shot learning,”Advances in neural information processing systems, vol. 30, 2017

    2017

  9. [18]

    Cross attention network for few-shot classification,

    R. Hou, H. Chang, B. Ma, S. Shan, and X. Chen, “Cross attention network for few-shot classification,” 2019. [Online]. Available: https://arxiv.org/abs/1910.07677

    2019 arXiv

  10. [19]

    Learning to compare: Relation network for few-shot learning,

    F. Sung, Y . Yang, L. Zhang, T. Xiang, P. H. Torr, and T. M. Hospedales, “Learning to compare: Relation network for few-shot learning,” in Proceedings of the IEEE conference on computer vision and pattern recognition, 2018, pp. 1199–1208

    2018

  11. [20]

    Matching net- works for one shot learning,

    O. Vinyals, C. Blundell, T. Lillicrap, D. Wierstraet al., “Matching net- works for one shot learning,”Advances in neural information processing systems, vol. 29, 2016

    2016

  12. [21]

    Model-agnostic meta-learning for fast adaptation of deep networks,

    C. Finn, P. Abbeel, and S. Levine, “Model-agnostic meta-learning for fast adaptation of deep networks,” inInternational conference on machine learning. PMLR, 2017, pp. 1126–1135

    2017

  13. [22]

    Meta-sgd: Learning to learn quickly for few-shot learning,

    Z. Li, F. Zhou, F. Chen, and H. Li, “Meta-sgd: Learning to learn quickly for few-shot learning,”arXiv preprint arXiv:1707.09835, 2017

    2017 arXiv

  14. [23]

    Meta-curvature,

    E. Park and J. B. Oliva, “Meta-curvature,”Advances in neural informa- tion processing systems, vol. 32, 2019

    2019

  15. [24]

    Episodic fine-tuning prototypical networks for optimization-based few-shot learning: Application to audio classification,

    X. Zhuang, G. Peeters, and G. Richard, “Episodic fine-tuning prototypical networks for optimization-based few-shot learning: Application to audio classification,” 2024. [Online]. Available: https://arxiv.org/abs/2410.05302

    2024

  16. [25]

    Adapted deep embeddings: A synthesis of methods for k-shot inductive transfer learning,

    T. Scott, K. Ridgeway, and M. C. Mozer, “Adapted deep embeddings: A synthesis of methods for k-shot inductive transfer learning,”Advances in Neural Information Processing Systems, vol. 31, 2018

    2018

  17. [26]

    Adaptive prototypical networks,

    M. Gogoi, S. Tiwari, and S. Verma, “Adaptive prototypical networks,” arXiv preprint arXiv:2211.12479, 2022

    2022 arXiv

  18. [27]

    Meta- dataset: A dataset of datasets for learning to learn from few examples,

    E. Triantafillou, T. Zhu, V . Dumoulin, P. Lamblin, U. Evci, K. Xu, R. Goroshin, C. Gelada, K. Swersky, P.-A. Manzagolet al., “Meta- dataset: A dataset of datasets for learning to learn from few examples,” arXiv preprint arXiv:1903.03096, 2019

    1903 arXiv

  19. [28]

    A hybrid approach with optimization and metric-based meta-learner for few-shot learning,

    D. Wang, Y . Cheng, M. Yu, X. Guo, and T. Zhang, “A hybrid approach with optimization and metric-based meta-learner for few-shot learning,” 2019

    2019

  20. [29]

    Esc: Dataset for environmental sound classification,

    K. J. Piczak, “Esc: Dataset for environmental sound classification,” in Proceedings of the 23rd ACM international conference on Multimedia, 2015, pp. 1015–1018

    2015

  21. [31]

    Deep convolutional networks on the pitch spiral for musical instrument recognition,

    V . Lostanlen and C.-E. Cella, “Deep convolutional networks on the pitch spiral for musical instrument recognition,”arXiv preprint arXiv:1605.06644, 2016

    2016 arXiv

  22. [32]

    Learning to match transient sound events using attentional similarity for few- shot sound recognition,

    S.-Y . Chou, K.-H. Cheng, J.-S. R. Jang, and Y .-H. Yang, “Learning to match transient sound events using attentional similarity for few- shot sound recognition,” inIEEE International Conference on Acoustics, Speech and Signal Processing (ICASSP), 2019, pp. 26–30

    2019

  23. [33]

    Nearest neighbor pattern classification,

    T. Cover and P. Hart, “Nearest neighbor pattern classification,”IEEE transactions on information theory, vol. 13, no. 1, pp. 21–27, 1967

    1967

  24. [34]

    Rectified linear units improve restricted boltz- mann machines,

    V . Nair and G. E. Hinton, “Rectified linear units improve restricted boltz- mann machines,” inProceedings of the 27th international conference on machine learning (ICML-10), 2010, pp. 807–814

    2010

  25. [35]

    Network in network,

    M. Lin, Q. Chen, and S. Yan, “Network in network,” 2014. [Online]. Available: https://arxiv.org/abs/1312.4400

    2014 arXiv

  26. [36]

    Speech commands: A dataset for limited-vocabulary speech recognition,

    P. Warden, “Speech commands: A dataset for limited-vocabulary speech recognition,” 2018. [Online]. Available: https://arxiv.org/abs/1804.03209

    2018 arXiv

  27. [37]

    Medleydb: A multitrack dataset for annotation-intensive mir research,

    R. Bittner, J. Salamon, M. Tierney, M. Mauch, C. Cannam, and J. Bello, “Medleydb: A multitrack dataset for annotation-intensive mir research,” 10 2014

    2014

  28. [38]

    Musical instrument recognition based on class pairwise feature selection,

    S. Essid, G. Richard, and B. David, “Musical instrument recognition based on class pairwise feature selection,” inInternational conference on music information retrieval (ISMIR), 2004

    2004

  29. [39]

    Pytorch: An imperative style, high-performance deep learning library,

    A. Paszke, S. Gross, F. Massa, A. Lerer, J. Bradbury, G. Chanan, T. Killeen, Z. Lin, N. Gimelshein, L. Antiga, A. Desmaison, A. K ¨opf, E. Yang, Z. DeVito, M. Raison, A. Tejani, S. Chilamkurthy, B. Steiner, L. Fang, J. Bai, and S. Chintala, “Pytorch: An imperative style, high-...

    2019 arXiv

  30. [40]

    learn2learn: A library for meta-learning research,

    S. M. Arnold, P. Mahajan, D. Datta, I. Bunner, and K. S. Zarkias, “learn2learn: A library for meta-learning research,”arXiv preprint arXiv:2008.12284, 2020

    2008 arXiv

  31. [41]

    librosa: Audio and music signal analysis in python

    B. McFee, C. Raffel, D. Liang, D. P. Ellis, M. McVicar, E. Battenberg, and O. Nieto, “librosa: Audio and music signal analysis in python.” SciPy, vol. 2015, pp. 18–24, 2015

    2015

Pith tools

Reviewed August 15, 2026 · model on record in the stance chip above.