D3RM: A Discrete Denoising Diffusion Refinement Model for Piano Transcription

Hounsu Kim; Juhan Nam; Taegyun Kwon

REVIEW 3 major objections 6 minor 1 cited by

D3RM: A Discrete Denoising Diffusion Refinement Model for Piano Transcription

T0 review · 3 major / 6 minor · reviewed 2026-08-10 · deepseek-v4-flash

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

Pith's one-line read A discrete diffusion refinement decoder raises piano transcription F1 to 97.57, beating feed-forward baselines and prior diffusion models.

desk verdict First discrete diffusion for piano transcription with a solid baseline gain, but the AS-sampling trick's manifold equivalence is unproven and likely wrong; the main result survives anyway. read the letter →

arxiv 2501.05068 v2 pith:XN6DOBWU submitted 2025-01-09 cs.SD cs.AIcs.LGeess.AS

Hounsu Kim , Taegyun Kwon , Juhan Nam This is my paper · ORCID

classification cs.SDcs.AIcs.LGeess.AS

keywords automaticmusictranscriptionpianodiscretediffusionmodeldenoisingrefinementNeighborhoodAttentionabsorbingstatesamplingMAESTROdatasetinformationretrieval

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 tries to show that discrete denoising diffusion can act as a refinement module for piano transcription, not just a generator. The authors attach a lightweight Neighborhood Attention denoising decoder to a pretrained acoustic model, and let it iteratively correct a multi-state piano roll. On MAESTRO, the resulting D3RM model reaches 97.57 note F1 (90.44 with offsets), beating its feed-forward baseline (96.45/87.60) and the prior continuous diffusion model DiffRoll (78.1). The paper also introduces 'absorbing state sampling,' an inference-only change to the diffusion posterior that improves refinement further (97.21 to 97.55). If correct, the result shows that diffusion-based iterative refinement is a practical way to make note predictions depend on each other.

What carries the argument

The load-bearing mechanism is the discrete diffusion transition matrix and the asymmetric posterior used at sampling. The forward transition $Q_\tau$ mixes a uniform replacement probability $\beta_\tau$ with a masking probability $\gamma_\tau$ and an absorbing [MASK] state; the training loss is the variational lower bound with the auxiliary denoising objective from VQ-diffusion. During refinement, the model first predicts the clean sample $\hat{y}_0$, then applies the posterior transition. For 'absorbing state sampling,' the posterior $q(y_{\tau-1}|y_\tau,\hat{y}_0)$ is replaced by a masking-only transition $\tilde{q}$ with total corrupted-token probability $\gamma_\tau + K\beta_\tau$, which the authors argue places the sample on the same diffusion manifold as training while avoiding token-replacement disruption. Neighborhood Attention layers with window size 3 and dilation pattern [1,2,4,8,1,2,4,8] provide the locality bias that lets each piano-roll token attend to neighboring tokens, enabling iterative refinement.

What would settle it

Run D3RM on a second piano dataset (e.g., MAPS) and compare inference with AS sampling against ordinary posterior sampling; if ordinary sampling matches or beats AS sampling, or if the empirical fraction of masked tokens at each timestep under AS sampling deviates from the training-time distribution, then the manifold-equivalence assumption is not holding.

Watch

Extended reading notes

Core claim

The central claim is that a discrete diffusion model, trained to denoise a multi-label piano roll with a transition matrix that both replaces and masks tokens, can serve as an effective refinement decoder for automatic piano transcription. The target is a $T \times 88 \times 6$ tensor with five note states plus a mask, and the decoder uses a pitchwise bidirectional LSTM plus stacked Neighborhood Attention (NA) layers with AdaLN conditioning, cross-attending to penultimate features of the pretrained HPPNet-style acoustic encoder. At inference, the authors replace the posterior $q(y_{\tau-1}|y_\tau,y_0)$ with a mask-only posterior $\tilde{q}$ in which the total corruption probability is $\gamma_\tau + K\beta_\tau$, based on the assumption that diffusion manifolds are identified by the total fraction of corrupted tokens. With this asymmetric train/inference schedule, D3RM obtains 97.57/90.44 note F1 (with/without offsets) on MAESTRO, versus 96.45/87.60 for the baseline and 78.1 for DiffRoll.

Load-bearing premise

The paper's improvement hinges on the assumption that diffusion manifolds are exactly the sets of samples with the same total number of corrupted tokens, which lets the authors swap in a mask-only posterior during inference; if wrong, the sampling is heuristic and the F1 gains may not generalize.

Editorial extensions

If this is right

Piano transcription can be improved by a lightweight diffusion refinement decoder rather than a larger feed-forward model: D3RM uses 2.0M parameters and beats models with up to 86.8M parameters (DiffRoll) and 19.7M (PAR).
Learning state-to-state transitions during training is what gives refinement power: setting $\gamma_T=1.0$ (absorbing-only training) drops F1 from 97.55 to 96.82.
Applying the mask-only 'absorbing state sampling' at inference consistently helps (F1 97.21 → 97.55; with offsets 89.10 → 90.28), so the refinement benefit is not just from the architecture.
Computing the acoustic encoder features once and reusing them across diffusion timesteps makes the refinement loop fast, needing only one forward pass of the encoder.

Reading between the lines

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

The 'manifold equals total corrupted-token fraction' assumption, if it holds, suggests a general recipe: any discrete diffusion discriminative decoder can be trained with replace+mask and sampled with mask-only transitions; the trick may transfer to other dense prediction tasks such as sound event detection or instrument separation.
A principled derivation of the optimal inference-time transition, rather than the simple mask-only posterior, could yield further gains; one testable variant is to choose the posterior that minimizes the KL to the training-time marginal at each timestep.
Because $\gamma_T$ values of 0.4–0.9 all performed within 0.07 F1, the refinement benefit appears robust to the exact corruption schedule; a follow-up could test whether the same holds on datasets with more polyphony or pedal complexity (e.g., MAPS).
The authors' own future direction of note-level token representations might let the same refinement framework handle note onsets/offsets as structured events rather than frame-wise states, potentially improving offset F1 more than the 2.84-point current gain.

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.

Desk Editor's Note

read the letter

Colleague,

Quick take: this is the first paper to apply discrete multinomial diffusion to piano transcription, and the empirical core holds up: D3RM beats its own feed-forward baseline clearly (97.57 vs 96.45 note F1) and crushes DiffRoll (78.1). The model is small (2M params), the code is public, and the ablations are honest. That is real, if incremental, progress for the transcription subfield.

The genuinely new piece is the 'absorbing state sampling' (AS) trick: train with a mixture of substitution and masking transitions, then at inference replace the posterior with a masking-only transition using total corruption gamma_T + K beta_T. The paper justifies this with an explicit 'simply assume' that the model sees two corruption processes as the same diffusion manifold. That is the soft spot, and the stress-test is right that it is not just unproven but likely false under a straightforward reading. In Eq. (1), the probability a token actually changes state is gamma + (K-1)beta, not gamma + K beta; the latter includes the beta self-loop mass. So the two processes do not even match on the expected number of corrupted tokens. Without an invariance argument for p_theta, AS sampling is a heuristic that happens to give +0.34 F1 on one dataset. The paper says 'simply assume' rather than 'we prove', so it is honest, but the claim that it 'places the sample on the manifold' is unsupported and probably wrong.

That said, the main result is not load-bearing on AS. Even without it, D3RM gets 97.21, still above the baseline and above DiffRoll. So the paper's central claim survives. The other weaknesses are minor: no error bars, and the gamma_T sweep (0.4-1.0) is reported on the test set, so there is a mild selection-on-test issue. The ablation differences among 0.4/0.6/0.9 are within noise, so claiming any one is optimal is overreach.

This is a solid engineering paper with a weak theoretical justification for one trick. It deserves a serious referee, and a good referee will ask the authors to either prove the manifold equivalence under a defined metric (e.g., expected fraction of changed tokens) or reframe AS as a heuristic with a matching experimental defense (e.g., multiple seeds, error bars). I would send it out, not desk-reject.

Referee Report

3 major / 6 minor

Summary. The paper proposes D3RM, a piano transcription system that couples a pretrained HPPNet-style acoustic encoder with a discrete denoising diffusion decoder built from Neighborhood Attention layers. The target label is a multi-state (5 note states plus mask) piano-roll representation, and the model is trained with a standard discrete diffusion loss plus an auxiliary denoising objective. The main methodological novelty is 'absorbing state sampling' (AS sampling): during inference, the transition posterior is replaced by a masking-only posterior with total corruption probability γτ + Kβτ, based on an explicit assumption in Section III-D that the model identifies diffusion manifolds as samples with the same total portion of corrupted tokens. Experiments on MAESTRO V3 report note F1 of 97.57 (with offsets 90.44), outperforming the baseline HPPNet-based model (96.45) and DiffRoll (78.1). Ablations show AS sampling improves F1 from 97.21 to 97.55, and that training with partial substitution (γT = 0.9) outperforms training with a pure absorbing state (γT = 1.0).

Significance. If the results are reliable, the paper makes a useful empirical contribution: it shows that a lightweight discrete diffusion refinement decoder can improve over a strong feedforward acoustic model for piano transcription, and it provides a concrete architecture (NA-based denoising, cross-attention conditioning, multi-state labels) with a modest parameter count. The code is released, which supports reproducibility. However, the significance is tempered by two factors. First, the headline gain of the proposed AS-sampling mechanism rests on an unproven manifold-invariance assumption, so the paper's central methodological novelty is not yet validated as a principled inference procedure. Second, all reported F1 differences are small and come from single runs, with no uncertainty estimates; the AS-sampling gain of +0.34 F1 and the γT selection (0.9 vs 0.4 or 0.6) are within a range that could plausibly be noise. The baseline comparison is internally consistent and the empirical framework is standard, but the lack of statistical grounding weakens the confidence one can place in the specific claims about the proposed components.

major comments (3)

[III-D] The manifold equivalence assumption is load-bearing for the AS-sampling contribution and is neither derived nor tested. The paper states, 'we first simply assume that the model identifies diffusion manifolds as samples with the same total portion of corrupted tokens,' and then replaces q(y_{τ−1}|yτ,y0) with a masking-only posterior using probability γτ + Kβτ. As the skeptical note observes, under Eq. (1) the probability that a non-mask token actually changes state is γ + (K−1)β (the diagonal is α+β), not γ + Kβ; the two processes agree only on the number of tokens that undergo a transition operation, not on the expected number of changed tokens. No invariance of pθ(y0|yτ,x) to substitution-versus-mask corruption is shown. This matters because the claimed +0.34 F1 improvement (97.21 to 97.55) is attributed to this substitution. I request either a formal argument or a concrete empirical test: for example, feed the same partially corrupted sample through the model with the two corruption schemes at matched corruption levels and compare the predicted y0 distributions, or validate AS sampling against an alternative inference that uses the true learned posterior. Without this, AS sampling is an unexplained heuristic that happens to help on one benchmark.
[V-A / Table I] All main results are reported from what appear to be single training runs, with no variance, confidence intervals, or statistical significance tests. The differences among the ablation conditions are small: AS sampling changes F1 by 0.34 (97.21 vs 97.55), encoder initialization by 0.09 (97.55 vs 97.46), and γT variations across 0.4/0.6/0.8/0.9 span only 0.07 (97.51–97.58). Without repeated runs or a significance test, it is impossible to know whether the AS-sampling improvement or the γT ranking is a reliable effect or run-to-run noise. The paper should provide mean±std over at least three seeds for the main configuration and for the key ablations, or otherwise justify that the differences exceed the noise floor.
[IV-B and V-B / Table II] The choice of γT = 0.9 as the main configuration is not justified by any stated validation procedure. The ablation table shows γT = 0.4 gives note F1 97.58 and F1-with-offsets 90.37, while γT = 0.9 gives 97.55 and 90.28; the differences are tiny and the best numbers actually favor γT = 0.4. Since the paper linearly increases γτ to a selected γT but does not describe how the grid {0.4,0.6,0.8,0.9,1.0} was chosen or post-selected, the reader cannot tell whether the headline configuration was chosen based on held-out data or on the test set. Please specify the validation protocol used to select γT and report test-set metrics for the selected configuration.

minor comments (6)

[I] The phrase 'Building on these foundings' should be 'Building on these findings'.
[II-D] 'hierachically' is a typo for 'hierarchically'.
[Table II] The column header 'END. COND.' should be 'ENC. COND.' for consistency with the caption's 'encoder conditioning method'.
[Fig. 3 caption] The caption begins with stray text, '4MIDI-Unprocessed 12 R1 2006 01-08 ORIG MID–', which appears to be a leftover from a dataset path; it should be removed.
[III-C] The notation is confusing because the single-step transition matrix is written as Qτ and the cumulative product is also written as Qτ (for example, 'The cumulative products Qτ = Qτ . . . Q1'). The two should be distinguished, e.g., using Q‾τ or Q(τ) for the cumulative product.
[III-C, Eq. (4)] The first term DKL[q(yT|y0)||p(yT)] is written with a subscript 'T' but the equation is then summed over τ = 1 to T; please check that the prior term is intended to be the Kullback-Leibler divergence for the final timestep and clarify the indexing.

Circularity Check

0 steps flagged · score 0.0 of 10

No significant circularity: D3RM's F1 gains are measured against external and baseline systems, and the AS-sampling assumption, while unproven, is a heuristic rather than an input-output equivalence.

full rationale

The paper's central claim is an empirical comparison (Section V-A, Table I): D3RM note F1 97.57/90.44 versus HPPNet-sp 97.18/83.80, HFT-transformer 97.44/90.53, Yan et al. 98.32/93.48, PAR 97.01/87.88, DiffRoll 78.1, and its own baseline 96.45/87.60. These numbers come from mir_eval on MAESTRO test data, not from fitting the target into the model. The AS-sampling proposal in Section III-D is introduced with an explicit assumption ('we first simply assume that the model identifies diffusion manifolds as samples with the same total portion of corrupted tokens'), and the claimed benefit is a measured ablation (Table II: 97.21 -> 97.55). That assumption may be a correctness risk—under Eq. (1), the probability that a token's state actually changes is gamma+(K-1)beta, whereas the masking posterior uses gamma+Kbeta, so the two corruption budgets do not exactly coincide—but an unproven heuristic is not circularity because the reported improvement is not defined to be true by construction. Self-citations [11] and [14] supply the multi-state label convention and a pitchwise-LSTM design, and [11] also appears as a benchmark row in Table I; these are supporting design choices and comparison points, not premises that force the D3RM F1 score. There is no self-citation uniqueness theorem, no fitted parameter renamed as a prediction, and no known result relabeled as new. The result is therefore self-contained against external benchmarks.

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

The central result rests on standard discrete diffusion math (borrowed from [19,20]), a domain representation assumption, an assumption about pretrained encoder features, and the AS sampling manifold assumption introduced by this paper. No new physical or conceptual entities are introduced. The headline hyperparameter gamma_T is tuned empirically, with lambda and the diffusion schedule carried from prior work.

free parameters (4)

gamma_T = 0.9 for the reported main model; grid-searched over {0.4, 0.6, 0.8, 0.9, 1.0}
Endpoint of the linear masking schedule. Selected empirically in Section IV-B; the gap between gamma_T=0.4 and 0.9 is only 0.01 F1 in Table II.
lambda = 0.0005
Weight for the auxiliary denoising objective, taken from VQ-Diffusion [20] rather than derived or tuned in this paper.
diffusion timesteps T = 100
Number of diffusion steps chosen in Section IV-B; affects training and sampling cost but is not derived from first principles.
NA decoder hyperparameters = 8 layers, window size 3, dilation pattern [1,2,4,8,1,2,4,8], hidden size 48
Architecture hyperparameters for the denoising decoder, set in Section IV-B without sensitivity analysis.

assumptions (4)

standard math Tractability and closed form of multinomial diffusion posterior (Eqs. 2-5)
The KL training objective relies on the posterior formulas from D3PM [19] and VQ-Diffusion [20]; the paper does not re-derive them.
ad hoc to paper The model identifies diffusion manifolds as samples with the same total portion of corrupted tokens
Section III-D; this is the core justification for AS sampling and is asserted, not proven.
domain assumption Multi-state note labels (onset, sustain, re-onset, offset, off) plus mask capture transcription targets adequately
Section II-A and III-C; representation adopted from [14] and central to the discrete diffusion formulation.
domain assumption Penultimate HPPNet features have local coherence with the piano roll and serve as good cross-attention conditioning
Section III-B; the design choice that the encoder output spatially aligns with the target roll.

how reviews work

0 comments

Cite this review

Pith. "Pith review of D3RM: A Discrete Denoising Diffusion Refinement Model for Piano Transcription." pith.science (2026). https://pith.science/paper/XN6DOBWU

@misc{pith2026250105068,
  author       = {Pith},
  title        = {Pith review of: D3RM: A Discrete Denoising Diffusion Refinement Model for Piano Transcription},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/XN6DOBWU}},
  note         = {Machine review of arXiv:2501.05068}
}

read the original abstract

Diffusion models have been widely used in the generative domain due to their convincing performance in modeling complex data distributions. Moreover, they have shown competitive results on discriminative tasks, such as image segmentation. While diffusion models have also been explored for automatic music transcription, their performance has yet to reach a competitive level. In this paper, we focus on discrete diffusion model's refinement capabilities and present a novel architecture for piano transcription. Our model utilizes Neighborhood Attention layers as the denoising module, gradually predicting the target high-resolution piano roll, conditioned on the finetuned features of a pretrained acoustic model. To further enhance refinement, we devise a novel strategy which applies distinct transition states during training and inference stage of discrete diffusion models. Experiments on the MAESTRO dataset show that our approach outperforms previous diffusion-based piano transcription models and the baseline model in terms of F1 score. Our code is available in https://github.com/hanshounsu/d3rm.

Figures

Figures reproduced from arXiv: 2501.05068 by the authors.

**Figure 2.** Illustration of the overall discrete diffusion process including the [PITH_FULL_IMAGE:figures/full_fig_p003_2.png] view at source ↗

**Figure 3.** Transcription example from a single 10-second sample [PITH_FULL_IMAGE:figures/full_fig_p004_3.png] view at source ↗

Discussion (0). Continue with ORCID to comment.

Forward citations

Cited by 1 Pith paper

Reviewed papers in the Pith corpus that reference this work. Sorted by Pith novelty score. Full citation record

Quantum Algorithm Software for Condensed Matter Physics
cond-mat.str-el 2025-06 reject novelty 2.0 of 10

A review of quantum algorithm software that advertises a benchmark suite, yet the body contains no benchmarks, data, or code.

Reference graph

Works this paper leans on

23 extracted references · 13 canonical work pages · cited by 1 Pith paper

[1]

Automatic piano transcription with hierarchical frequency-time transformer,

K. Toyama, T. Akama, Y . Ikemiya, Y . Takida, W.-H. Liao, and Y . Mit- sufuji, “Automatic piano transcription with hierarchical frequency-time transformer,” in Proc. of the 24th International Society for Music Information Retrieval Conference (ISMIR) , 2023

work page 2023
[2]

Scoring intervals using non-hierarchical transformer for automatic piano transcription,

Y . Yan and Z. Duan, “Scoring intervals using non-hierarchical transformer for automatic piano transcription,” arXiv preprint arXiv:2404.09466, 2024

arXiv 2024
[3]

Naturalspeech 3: Zero-shot speech syn- thesis with factorized codec and diffusion models,

Z. Ju, Y . Wang, K. Shen, X. Tan, D. Xin, D. Yang, Y . Liu, Y . Leng, K. Song, S. Tang et al. , “Naturalspeech 3: Zero-shot speech syn- thesis with factorized codec and diffusion models,” arXiv preprint arXiv:2403.03100, 2024

arXiv 2024
[4]

Analog bits: Generating discrete data using diffusion models with self-conditioning,

T. Chen, R. Zhang, and G. Hinton, “Analog bits: Generating discrete data using diffusion models with self-conditioning,” in The Eleventh International Conference on Learning Representations , 2023

work page 2023
[5]

Dformer: Diffusion-guided transformer for universal image segmen- tation,

H. Wang, J. Cao, R. M. Anwer, J. Xie, F. S. Khan, and Y . Pang, “Dformer: Diffusion-guided transformer for universal image segmen- tation,” arXiv preprint arXiv:2306.03437 , 2023

arXiv 2023
[6]

Diffroll: Diffusion-based gen- erative music transcription with unsupervised pretraining capability,

K. W. Cheuk, R. Sawata, T. Uesaka, N. Murata, N. Takahashi, S. Taka- hashi, D. Herremans, and Y . Mitsufuji, “Diffroll: Diffusion-based gen- erative music transcription with unsupervised pretraining capability,” in ICASSP 2023-2023 IEEE International Conference on Acoustics, Speech and Signal Processing (ICASSP) . IEEE, 2023, pp. 1–5

work page 2023
[7]

Neighborhood attention transformer,

A. Hassani, S. Walton, J. Li, S. Li, and H. Shi, “Neighborhood attention transformer,” in IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR), 2023

work page 2023
[8]

Onsets and frames: Dual-objective piano transcription,

C. Hawthorne, E. Elsen, J. Song, A. Roberts, I. Simon, C. Raffel, J. Engel, S. Oore, and D. Eck, “Onsets and frames: Dual-objective piano transcription,” in Proc. of the 19th International Society for Music Information Retrieval Conference (ISMIR) , October 2018, pp. 50–57

work page 2018

Show all 23 references

[9]

High-resolution piano transcription with pedals by regressing onset and offset times,

Q. Kong, B. Li, X. Song, Y . Wan, and Y . Wang, “High-resolution piano transcription with pedals by regressing onset and offset times,” IEEE/ACM Transactions on Audio, Speech, and Language Processing , vol. 29, pp. 3707–3717, 2021

2021
[10]

Hppnet: Modeling the harmonic struc- ture and pitch invariance in piano transcription,

W. Wei, P. Li, Y . Yu, and W. Li, “Hppnet: Modeling the harmonic struc- ture and pitch invariance in piano transcription,” in Proceedings of the 23rd International Society for Music Information Retrieval Conference, ISMIR 2022, Bengaluru, India, December 4-8, 2022, 2022, pp. 709–716

2022
[11]

Towards efficient and real-time piano transcription using neural autoregressive models,

T. Kwon, D. Jeong, and J. Nam, “Towards efficient and real-time piano transcription using neural autoregressive models,” arXiv preprint arXiv:2404.06818, 2024

2024 arXiv
[12]

Sequence-to-sequence piano transcription with transformers,

C. Hawthorne, I. Simon, R. Swavely, E. Manilow, and J. Engel, “Sequence-to-sequence piano transcription with transformers,” in Proc. of the 22th International Society for Music Information Retrieval Con- ference (ISMIR), 2021

2021
[13]

Piano transcription with harmonic attention,

R. Wu, X. Wang, Y . Li, W. Xu, and W. Cheng, “Piano transcription with harmonic attention,” in ICASSP 2024-2024 IEEE International Conference on Acoustics, Speech and Signal Processing (ICASSP) . IEEE, 2024, pp. 1256–1260

2024
[14]

Polyphonic piano transcription using autoregressive multi-state note model,

T. Kwon, D. Jeong, and J. Nam, “Polyphonic piano transcription using autoregressive multi-state note model,” inInternational Society for Music Information Retrieval Conference , 2020

2020
[15]

Dif- fwave: A versatile diffusion model for audio synthesis,

Z. Kong, W. Ping, J. Huang, K. Zhao, and B. Catanzaro, “Dif- fwave: A versatile diffusion model for audio synthesis,” arXiv preprint arXiv:2009.09761, 2020

2009 arXiv
[16]

Wavenet: A genera- tive model for raw audio,

A. v. d. Oord, S. Dieleman, H. Zen, K. Simonyan, O. Vinyals, A. Graves, N. Kalchbrenner, A. Senior, and K. Kavukcuoglu, “Wavenet: A genera- tive model for raw audio,” arXiv preprint arXiv:1609.03499 , 2016

2016 arXiv
[17]

Deep unsupervised learning using nonequilibrium thermodynamics,

J. Sohl-Dickstein, E. Weiss, N. Maheswaranathan, and S. Ganguli, “Deep unsupervised learning using nonequilibrium thermodynamics,” in International conference on machine learning . PMLR, 2015, pp. 2256–2265

2015
[18]

Argmax flows and multinomial diffusion: Learning categorical distributions,

E. Hoogeboom, D. Nielsen, P. Jaini, P. Forr ´e, and M. Welling, “Argmax flows and multinomial diffusion: Learning categorical distributions,” Advances in Neural Information Processing Systems, vol. 34, pp. 12 454– 12 465, 2021

2021
[19]

Structured denoising diffusion models in discrete state-spaces,

J. Austin, D. D. Johnson, J. Ho, D. Tarlow, and R. Van Den Berg, “Structured denoising diffusion models in discrete state-spaces,” Ad- vances in Neural Information Processing Systems , vol. 34, pp. 17 981– 17 993, 2021

2021
[20]

Vector quantized diffusion model for text-to-image synthesis,

S. Gu, D. Chen, J. Bao, F. Wen, B. Zhang, D. Chen, L. Yuan, and B. Guo, “Vector quantized diffusion model for text-to-image synthesis,” in Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition, 2022, pp. 10 696–10 706

2022
[21]

Diffsound: Discrete diffusion model for text-to-sound generation,

D. Yang, J. Yu, H. Wang, W. Wang, C. Weng, Y . Zou, and D. Yu, “Diffsound: Discrete diffusion model for text-to-sound generation,” IEEE/ACM Transactions on Audio, Speech, and Language Processing , 2023

2023
[22]

Enabling factorized piano music modeling and generation with the MAESTRO dataset,

C. Hawthorne, A. Stasyuk, A. Roberts, I. Simon, C.-Z. A. Huang, S. Dieleman, E. Elsen, J. Engel, and D. Eck, “Enabling factorized piano music modeling and generation with the MAESTRO dataset,” in International Conference on Learning Representations , 2019

2019
[23]

mir eval: A transparent implementation of common MIR metrics,

C. Raffel, B. McFee, E. J. Humphrey, J. Salamon, O. Nieto, D. Liang, and D. P. Ellis, “mir eval: A transparent implementation of common MIR metrics,” in Proc. of the 15th International Society for Music Information Retrieval Conference (ISMIR) , 2014, pp. 367–372

2014

Pith tools

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

[1] [1]

Automatic piano transcription with hierarchical frequency-time transformer,

K. Toyama, T. Akama, Y . Ikemiya, Y . Takida, W.-H. Liao, and Y . Mit- sufuji, “Automatic piano transcription with hierarchical frequency-time transformer,” in Proc. of the 24th International Society for Music Information Retrieval Conference (ISMIR) , 2023

work page 2023

[2] [2]

Scoring intervals using non-hierarchical transformer for automatic piano transcription,

Y . Yan and Z. Duan, “Scoring intervals using non-hierarchical transformer for automatic piano transcription,” arXiv preprint arXiv:2404.09466, 2024

arXiv 2024

[3] [3]

Naturalspeech 3: Zero-shot speech syn- thesis with factorized codec and diffusion models,

Z. Ju, Y . Wang, K. Shen, X. Tan, D. Xin, D. Yang, Y . Liu, Y . Leng, K. Song, S. Tang et al. , “Naturalspeech 3: Zero-shot speech syn- thesis with factorized codec and diffusion models,” arXiv preprint arXiv:2403.03100, 2024

arXiv 2024

[4] [4]

Analog bits: Generating discrete data using diffusion models with self-conditioning,

T. Chen, R. Zhang, and G. Hinton, “Analog bits: Generating discrete data using diffusion models with self-conditioning,” in The Eleventh International Conference on Learning Representations , 2023

work page 2023

[5] [5]

Dformer: Diffusion-guided transformer for universal image segmen- tation,

H. Wang, J. Cao, R. M. Anwer, J. Xie, F. S. Khan, and Y . Pang, “Dformer: Diffusion-guided transformer for universal image segmen- tation,” arXiv preprint arXiv:2306.03437 , 2023

arXiv 2023

[6] [6]

Diffroll: Diffusion-based gen- erative music transcription with unsupervised pretraining capability,

K. W. Cheuk, R. Sawata, T. Uesaka, N. Murata, N. Takahashi, S. Taka- hashi, D. Herremans, and Y . Mitsufuji, “Diffroll: Diffusion-based gen- erative music transcription with unsupervised pretraining capability,” in ICASSP 2023-2023 IEEE International Conference on Acoustics, Speech and Signal Processing (ICASSP) . IEEE, 2023, pp. 1–5

work page 2023

[7] [7]

Neighborhood attention transformer,

A. Hassani, S. Walton, J. Li, S. Li, and H. Shi, “Neighborhood attention transformer,” in IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR), 2023

work page 2023

[8] [8]

Onsets and frames: Dual-objective piano transcription,

C. Hawthorne, E. Elsen, J. Song, A. Roberts, I. Simon, C. Raffel, J. Engel, S. Oore, and D. Eck, “Onsets and frames: Dual-objective piano transcription,” in Proc. of the 19th International Society for Music Information Retrieval Conference (ISMIR) , October 2018, pp. 50–57

work page 2018