Pith. sign in

REVIEW 4 major objections 5 minor 41 references

Mel-Spectrogram Inversion via Alternating Direction Method of Multipliers

T0 review · 4 major / 5 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 The paper proposes an ADMM-based algorithm for mel-spectrogram inversion that jointly estimates full-band STFT magnitude and phase, achieving better reconstruction quality and roughly five times fewer iterations than the previous…

desk verdict Solid ADMM-based mel-spectrogram inversion with correct closed-form updates and real gains over iPALM; missing convergence analysis and residual reporting, but worth reviewing. read the letter →

arxiv 2501.05557 v2 pith:FBEXURN5 submitted 2025-01-09 eess.AS cs.SD

Yoshiki Masuyama , Natsuki Ueno , Nobutaka Ono This is my paper · ORCID
classification eess.AScs.SD
keywords mel-spectrograminversionphasereconstructionalternatingdirectionmethodofmultipliersnonconvexoptimizationproximalsplittingspeechsynthesisfoleysoundfull-bandmagnitudeestimation
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

Mel-spectrogram inversion commonly works in two stages: first predict a full-band STFT magnitude from the mel-spectrogram, then run a phase-reconstruction algorithm such as Griffin–Lim. This paper's central claim is that those two steps should be optimized jointly, and that an alternating direction method of multipliers (ADMM) formulation does the joint estimation better and with far fewer iterations than the existing iPALM-based joint method. The authors report that their ADMM-based algorithm, iterated only 100 times, matches the quality the previous joint method reaches after 500 iterations, and that it improves spectral convergence, PESQ, and ESTOI on speech and full-band spectral convergence on foley sounds. The payoff is a training-free inversion method that is applicable to any signal yet competitive with classical signal-processing approaches.

What carries the argument

The central object is the ADMM reformulation of problem (7): the full-band magnitude is split into two variables, one tied to the mel-spectrogram fidelity and one tied to the nonnegative magnitude constraint, and the STFT coefficient is split from a consistency variable constrained to the image of STFT. The workhorse is conditional independence among the blocks, which reduces the joint updates to four separate proximity-operator evaluations. The X-update uses the closed-form prox for magnitude least squares, $\operatorname{prox}_{J(\cdot,Y)/\rho}(\Psi) = (Y + \rho|\Psi|)/(1+\rho)\odot \Psi/|\Psi|$; the W-update solves $(\lambda E^T E + \rho I)^{-1}(\lambda E^T M + \rho \Phi)$, with the inverse precomputable; the Z-update is STFT–iSTFT projection; and the Y-update is a soft-thresholded weighted combination $(|X| + \rho \Upsilon)_+/(1+\rho)$.

What would settle it

Run Algorithm 2 with the paper's recommended $(\lambda, \rho) = (5000, 0.1)$ on a fixed TIMIT utterance and monitor the primal residuals $\|Z - X\|$ and $\|Y - W\|$ across iterations; if these residuals do not decrease while SCM saturates, the assumed ADMM convergence is violated. A broader test: on a held-out corpus with random initializations, if ADMM-Joint's median SCM after 500 iterations is not better than iPALM-Joint's, or if its 100-iteration SCM does not match iPALM-Joint's 500-iteration SCM, the central speed and quality claims fail.

Watch

Extended reading notes

Core claim

The paper builds a mel-spectrogram inversion algorithm, ADMM-Joint, by rewriting the joint full-band magnitude and phase estimation problem as a constrained optimization with auxiliary variables and then applying ADMM. Each of the four block updates becomes a proximal operator calculation: the STFT coefficients are updated by a magnitude-constrained least-squares prox, the auxiliary full-band magnitude by a closed-form ridge-like solve whose matrix inverse can be precomputed, the consistent-STFT variable by projection onto the STFT image, and the magnitude variable by a projection of a weighted combination of magnitude and dual variables. On TIMIT speech and DCASE2023 foley sounds, ADMM-Joint outperforms the cascaded PG-GLA and ADMM-GLA baselines and the iPALM-based joint method [27], with the 100-iteration ADMM run comparable to the 500-iteration iPALM run.

Load-bearing premise

The method's advantage assumes that multi-block ADMM converges to a good stationary point of this nonconvex problem within a few hundred iterations for arbitrary signals and without special initialization; the paper does not provide a convergence proof.

Editorial extensions

If this is right

  • If the central claim holds, training-free mel-spectrogram inversion can be made roughly five times faster at equal quality, which matters for real-time and embedded speech synthesis pipelines.
  • Because the method requires no training data, it can be applied immediately to any signal class, including foley and environmental sounds where neural vocoders are rarely available.
  • The ADMM structure suggests that the same block-splitting strategy could be reused for other audio inverse problems that combine a redundancy constraint with a lossy perceptual transform.
  • The comparisons to cascaded methods indicate that the joint update's error mitigation, rather than the phase-only refinement, is what drives the quality gain.

Reading between the lines

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

  • The paper leaves implicit that the same ADMM scheme could be unfolded into a trainable network, similar to deep Griffin–Lim; the closed-form prox updates are differentiable, so a learned version is a natural next step.
  • The reported sensitivity to $\rho$ hints that an adaptive penalty schedule, which the authors do not test, could further reduce iteration counts and remove the manual tuning burden.
  • A rigorous convergence analysis for the nonconvex ADMM scheme would strengthen the empirical claim; until then, the method's reliability on out-of-distribution signals is an open question.
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 / 5 minor

Summary. The paper proposes an ADMM-based joint estimation method for mel-spectrogram inversion. It reformulates the joint full-band magnitude and phase reconstruction problem (7) as an ADMM problem (8), derives closed-form proximity-operator updates for X, W, Z, and Y (Section III-B, Eqs. (12)–(20)), and presents Algorithm 2. Experiments on TIMIT speech and DCASE2023 foley sounds compare ADMM-Joint with cascaded PG-GLA, ADMM-GLA, and the iPALM-based joint method [27], reporting improved SCM, PESQ, ESTOI, and full-band spectral convergence, with the claim that ADMM-Joint with 100 iterations is comparable to iPALM-Joint with 500 iterations (Section IV-B).

Significance. If the empirical claims hold, this is a useful training-free, signal-processing-based mel-spectrogram inversion method that reduces the iteration count of the prior joint estimation method by a factor of five while improving objective quality on speech and foley sounds. The derivation of the ADMM updates is largely correct, and the conditional-independence splitting is a sensible way to keep each subproblem simple. The experiments cover multiple metrics and two domains, which is a strength. However, the central comparison is weakened by the lack of any convergence or stationarity analysis for the nonconvex multi-block ADMM scheme, and by the fact that the primary metric SCM is essentially the objective being minimized; the external metrics (PESQ/ESTOI and full-band SC) partially mitigate this but do not remove the need for a rigorous assessment of whether the algorithm actually solves (7).

major comments (4)
  1. [Section III-B, Algorithm 2] The central claim that ADMM-Joint solves problem (7) and converges to a useful stationary point is not supported by any convergence analysis. The objective (7) is nonconvex, and the gradient of J(X,Y) with respect to X is not Lipschitz at |X|=0, as the paper itself notes in Section II-A when setting the derivative to zero. Standard nonconvex ADMM theorems cited in Section III-A do not directly apply to this multi-block, non-Lipschitz setting. The paper reports no primal residuals ||Z-X|| or ||Y-W||, no dual residuals, and no stationarity-violation measures over iterations. Without such evidence, the observed advantage over iPALM at 100 iterations could be an artifact of early stopping or of the particular choice of ρ rather than a genuine convergence property. Please add residual/stationarity plots or a convergence analysis for Algorithm 2.
  2. [Section IV-A, Fig. 1] The comparison with iPALM-Joint may not be fully fair with respect to hyperparameters. Figure 1 shows that the proposed method's performance varies substantially with ρ, and the pair (λ,ρ)=(5000,0.1) is selected on a 100-utterance validation set. The iPALM-Joint baseline uses λ=10 taken from the prior work [27] without re-tuning on the same validation set, and its inertial parameter α is not specified. To support the claim of superiority, both methods should be tuned on the same validation data, or the authors should show that ADMM-Joint remains competitive across a range of ρ and that the chosen ρ is not a single fragile point.
  3. [Section IV-B, Eqs. (7), (21)] The SCM metric defined in Eq. (21) is exactly the mel-reconstruction error ∥E|STFT(x̂)|−M∥/∥M∥, which is the term I(Y) minimized in problem (7). Thus SCM gains for ADMM-Joint partly reflect better minimization of the same objective, making SCM a circular measure of quality. The paper does report external quality measures (PESQ and ESTOI on speech, full-band SC on foley) that support the claim, but the main convergence-speed comparison in Fig. 2 is made in terms of SCM. Please clearly acknowledge this circularity, and preferably report the convergence comparison also in terms of an external metric or a stationarity measure.
  4. [Section IV-B, Fig. 2-3] For the speech experiments, the paper does not report any statistical significance test for the differences in SCM, PESQ, or ESTOI between ADMM-Joint and iPALM-Joint. The foley results include a paired t-test (Section IV-C), but the speech boxplots are interpreted qualitatively. Given the claim of consistent improvement, please add a significance test or confidence intervals for the speech metrics as well.
minor comments (5)
  1. [Section IV-B] The abbreviation 'iPLAM' appears in the text and in Figure legends; it should be 'iPALM' consistently.
  2. [Section IV-A] The phrase 'spectral convergence on mel-spectrogarm' contains a typo: 'mel-spectrogram'.
  3. [Section IV-A, Fig. 1] The top-left panel shows SCM versus λ only, while the other two panels show PESQ and ESTOI versus ρ; for completeness, please also show how PESQ and ESTOI vary with λ, since the selected λ=5000 is justified only by SCM.
  4. [Section IV-C] Please report the exact p-value or t-statistic for the paired t-test on foley sounds, not merely the statement that significance was confirmed.
  5. [Section IV-B] The initialization of variables for both algorithms is not specified. Please state how Z, Y, V, U are initialized for ADMM-Joint and how Z, Zold, Y are initialized for iPALM-Joint, since initialization can affect convergence in nonconvex problems.

Circularity Check

0 steps flagged · score 0.0 of 10

No significant circularity: ADMM updates are derived from first principles, and the claimed gains are supported by external objective metrics.

full rationale

The paper's derivation chain is self-contained. The optimization problem in Eq. (7) is clearly defined, and the ADMM reformulation in Eq. (8) with variable splitting is standard. Each subproblem is solved in closed form: the X-update is a simple proximity operator of the magnitude least-squares term, the W-update is a linear least-squares solution, the Z-update is the STFT-consistency projection, and the Y-update is a nonnegative projection of a weighted average. None of these steps imports the target result as an assumption. The SCM metric in Eq. (21) is related to the optimized objective, but it is an evaluation metric, not a fitted parameter or a predicted quantity; both the proposed and baseline methods optimize the same objective, so comparing SCM is a legitimate empirical check. More importantly, the paper's conclusions are also supported by PESQ, ESTOI, and full-band spectral convergence, which are not directly optimized. The self-citations to Refs. [8], [27], and [36] are either elementary derivable formulas, baseline comparisons, or motivational references; they are not load-bearing in a circular sense. The lack of a convergence proof for nonconvex ADMM is a rigor/robustness limitation, not a circularity issue. No specific circular step can be exhibited from the paper's own equations.

Assumptions & free parameters 2 free parameters · 4 assumptions · 0 invented entities

The central claim rests on two tuned hyperparameters (lambda, rho), the standard STFT consistency projection, the linear mel-filterbank model, and an unproven convergence assumption for nonconvex multi-block ADMM. No new physical or mathematical entities are introduced.

free parameters (2)
  • lambda (mel-fidelity weight) = 5000
    Balances mel-spectrogram fidelity against STFT consistency in the objective (7); tuned on 100 TIMIT utterances in Section IV-A, where the paper reports performance is stable over a wide range.
  • rho (ADMM penalty parameter) = 0.1
    Sets the step size for the ADMM dual updates; tuned on the same TIMIT validation set, and Fig. 1 shows SCM, PESQ, and ESTOI vary noticeably with rho.
assumptions (4)
  • standard math The image of STFT, C, is a nonempty set, and iSTFT with the canonical dual window is a left inverse on C; the projection P_C(X)=STFT(iSTFT(X)) is valid.
    Invoked in Section II-A and used in the Z-update (18); standard frame-theoretic property of STFT with a dual window.
  • ad hoc to paper Multi-block ADMM converges to a stationary point of the nonconvex problem (7).
    Assumed without proof in Section III; the paper references ADMM's success in other nonconvex problems but provides no specific convergence theorem for this formulation.
  • domain assumption The mel-filterbank E is a known, fixed linear operator, and the given mel-spectrogram M satisfies M = E A for some nonnegative full-band magnitude A.
    Standard assumption in mel-spectrogram inversion, stated in Section II-B; on the fitted signal, this holds by construction, but for arbitrary inputs M may be inconsistent with any STFT magnitude, complicating inversion.
  • standard math Proximity operator formulas (15) and (17) are correct for the complex least-squares magnitude cost and the linear least-squares cost.
    Used for the X- and W-updates; (15) is credited to [36] and (17) follows from the normal equations.

how reviews work

0 comments
Cite this review

Pith. "Pith review of Mel-Spectrogram Inversion via Alternating Direction Method of Multipliers." pith.science (2026). https://pith.science/paper/FBEXURN5

@misc{pith2026250105557,
  author       = {Pith},
  title        = {Pith review of: Mel-Spectrogram Inversion via Alternating Direction Method of Multipliers},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/FBEXURN5}},
  note         = {Machine review of arXiv:2501.05557}
}
read the original abstract

Signal reconstruction from its mel-spectrogram is known as mel-spectrogram inversion and has many applications, including speech and foley sound synthesis. In this paper, we propose a mel-spectrogram inversion method based on a rigorous optimization algorithm. To reconstruct a time-domain signal with inverse short-time Fourier transform (STFT), both full-band STFT magnitude and phase should be predicted from a given mel-spectrogram. Their joint estimation has outperformed the cascaded full-band magnitude prediction and phase reconstruction by preventing error accumulation. However, the existing joint estimation method requires many iterations, and there remains room for performance improvement. We present an alternating direction method of multipliers (ADMM)-based joint estimation method motivated by its success in various nonconvex optimization problems including phase reconstruction. An efficient update of each variable is derived by exploiting the conditional independence among the variables. Our experiments demonstrate the effectiveness of the proposed method on speech and foley sounds.

Figures

Figures reproduced from arXiv: 2501.05557 by the authors.

Figure 1
Figure 1. According to its top left panel, the proposed method performed [PITH_FULL_IMAGE:figures/full_fig_p003_1.png] view at source ↗
Figure 3
Figure 3. Boxplots of PESQ and ESTOI with 500 iterations. [PITH_FULL_IMAGE:figures/full_fig_p004_3.png] view at source ↗
Figure 2
Figure 2. Average SCM with respect to the number of iterations and the boxplot [PITH_FULL_IMAGE:figures/full_fig_p004_2.png] view at source ↗

Discussion (0). Continue with ORCID to comment.

Reference graph

Works this paper leans on

41 extracted references · 40 canonical work pages

  1. [27]

    Signal reconstruction from mel- spectrogram based on bi-level consistency of full-band magnitude and phase,

    Y . Masuyama, N. Ueno, and N. Ono, “Signal reconstruction from mel- spectrogram based on bi-level consistency of full-band magnitude and phase,” in Proc. IEEE Workshop Appl. Signal Process. Audio Acoust. (WASPAA), Oct. 2023

    work page 2023

  2. [1]

    Phase processing for single-channel speech enhancement: History and recent advances,

    T. Gerkmann, M. Krawczyk-Becker, and J. Le Roux, “Phase processing for single-channel speech enhancement: History and recent advances,” IEEE Signal Process. Mag. , vol. 32, pp. 55–66, Mar. 2015

    work page 2015

  3. [2]

    Advances in phase-aware sig- nal processing in speech communication,

    P. Mowlaee, R. Saeidi, and Y . Stylianou, “Advances in phase-aware sig- nal processing in speech communication,” Speech Commun., vol. 81, pp. 1–29, Jul. 2016

    work page 2016

  4. [3]

    Time-frequency phase re- trieval for audio—the effect of transform parameters,

    A. Marafioti, N. Holighaus, and P. Majdak, “Time-frequency phase re- trieval for audio—the effect of transform parameters,” IEEE Trans. Sig- nal Process., vol. 69, pp. 3585–3596, Jun. 2021

    work page 2021

  5. [4]

    A noniterative method for reconstruction of phase from STFT magnitude,

    Z. Pr ˚uˇsa, P. Balazs, and P. L. Søndergaard, “A noniterative method for reconstruction of phase from STFT magnitude,” IEEE/ACM Trans. Au- dio, Speech, Lang. Process. , vol. 25, pp. 1154–1164, May 2017

    work page 2017

  6. [5]

    Phase reconstruction from amplitude spectrograms based on directional- statistics deep neural networks,

    S. Takamichi, Y . Saito, N. Takamune, D. Kitamura, and H. Saruwatari, “Phase reconstruction from amplitude spectrograms based on directional- statistics deep neural networks,” Signal Process. , vol. 169, p. 107368, Apr. 2020

    work page 2020

  7. [6]

    Signal estimation from modified short-time Fourier transform,

    D. Griffin and J. Lim, “Signal estimation from modified short-time Fourier transform,” IEEE Trans. Acoust., Speech, Signal Process. , vol. 32, pp. 236–243, Apr. 1984

    work page 1984

  8. [7]

    A fast Griffin–Lim algorithm,

    N. Perraudin, P. Balazs, and P. L. Søndergaard, “A fast Griffin–Lim algorithm,” in Proc. IEEE Workshop Appl. Signal Process. Audio Acoust. (WASPAA), Oct. 2013

    work page 2013

Show all 41 references
  1. [8]

    Griffin–Lim like phase recov- ery via alternating direction method of multipliers,

    Y . Masuyama, K. Yatabe, and Y . Oikawa, “Griffin–Lim like phase recov- ery via alternating direction method of multipliers,” IEEE Signal Pro- cess. Lett. , vol. 26, pp. 184–188, Jan. 2019

    2019

  2. [9]

    Phase retrieval with Breg- man divergences and application to audio signal recovery,

    P. H. Vial, T. O. P. Magron, and C. F ´evotte, “Phase retrieval with Breg- man divergences and application to audio signal recovery,” IEEE J. Sel. Top. Signal Process. , vol. 15, no. 1, pp. 51–64, Jan. 2021

    2021

  3. [10]

    Beyond Griffin–Lim: Improved iterative phase retrieval for speech,

    T. Peer, S. Welker, and T. Gerkmann, “Beyond Griffin–Lim: Improved iterative phase retrieval for speech,” in Proc. Int. Workshop Acoust. Sig- nal Enhance. (IWAENC) , Sep. 2022

    2022

  4. [11]

    Faster than fast: Accelerating the Griffin-Lim algorithm,

    R. Nenov, D. K. Nguyen, and P. Balazs, “Faster than fast: Accelerating the Griffin-Lim algorithm,” in Proc. IEEE Int. Conf. Acoust., Speech, Signal Process. (ICASSP) , Jun. 2023

    2023

  5. [12]

    Deep Griffin–Lim iteration,

    Y . Masuyama, K. Yatabe, Y . Koizumi, Y . Oikawa, and N. Harada, “Deep Griffin–Lim iteration,” in Proc. IEEE Int. Conf. Acoust., Speech, Signal Process. (ICASSP), May 2019, pp. 61–65

    2019

  6. [13]

    Deep Griffin–Lim iteration: Trainable iterative phase reconstruc- tion using neural network,

    ——, “Deep Griffin–Lim iteration: Trainable iterative phase reconstruc- tion using neural network,” IEEE J. Sel. Top. Signal Process. , vol. 15, pp. 37–50, Jan. 2021

    2021

  7. [14]

    Learning the proxim- ity operator in unfolded ADMM for phase retrieval,

    P. H. Vial, P. Magron, T. Oberlin, and C. F ´evotte, “Learning the proxim- ity operator in unfolded ADMM for phase retrieval,” IEEE Signal Pro- cess. Lett. , vol. 29, pp. 1619–1623, Jul. 2022

    2022

  8. [15]

    Diffphase: Generative diffusion- based STFT phase retrieval,

    T. Peer, S. Welker, and T. Gerkmann, “Diffphase: Generative diffusion- based STFT phase retrieval,” in IEEE Int. Conf. Acoust., Speech, Signal Process. (ICASSP), May 2023

    2023

  9. [16]

    Natural TTS synthesis by conditioning wavenet on MEL spectrogram predictions,

    J. Shen, R. Pang, R. J. Weiss, M. Schuster, N. Jaitly, Z. Yang, Z. Chen, Y . Zhang, Y . Wang, R. Skerrv-Ryan, R. A. Saurous, Y . Agiomvrgian- nakis, and Y . Wu, “Natural TTS synthesis by conditioning wavenet on MEL spectrogram predictions,” in Proc. IEEE Int. Conf. Acoust., Sp...

    2018

  10. [17]

    Deep voice 3: Scaling text-to-speech with convolutional sequence learning,

    W. Ping, K. Peng, A. Gibiansky, S. O. Arik, A. Kannan, S. Narang, J. Raiman, and J. Miller, “Deep voice 3: Scaling text-to-speech with convolutional sequence learning,” in Proc. Int. Conf. Learn. Represent. (ICLR), Apr. 2018

    2018

  11. [18]

    Fast- speech 2: Fast and high-quality end-to-end text to speech,

    Y . Ren, C. Hu, X. Tan, T. Qin, S. Zhao, Z. Zhao, and T. Y . Liu, “Fast- speech 2: Fast and high-quality end-to-end text to speech,” in Proc. Int. Conf. Learn. Represent. (ICLR) , May 2021

    2021

  12. [19]

    CycleGAN-VC3: Examining and improving CycleGAN-VCs for mel-spectrogram conver- sion,

    T. Kaneko, H. Kameoka, K. Tanaka, and N. Hojo, “CycleGAN-VC3: Examining and improving CycleGAN-VCs for mel-spectrogram conver- sion,” in Proc. Interspeech, Oct. 2020, pp. 2017–2021

    2020

  13. [20]

    Non- autoregressive sequence-to-sequence voice conversion,

    T. Hayashi, W. C. Huang, K. Kobayashi, and T. Toda, “Non- autoregressive sequence-to-sequence voice conversion,” in Proc. IEEE Int. Conf. Acoust., Speech, Signal Process. (ICASSP) , Jun. 2021, pp. 7068–7072

    2021

  14. [21]

    Mel spectrogram inversion with stable pitch,

    B. D. Giorgi, M. Levy, and R. Sharp, “Mel spectrogram inversion with stable pitch,” in Proc. Int. Soc. Music Inf. Retr . Conf. (ISMIR) , Dec. 2022, pp. 233–239

    2022

  15. [22]

    Melgan: Generative adversarial networks for conditional waveform synthesis,

    K. Kumar, R. Kumar, T. De Boissiere, L. Gestin, W. Z. Teoh, J. Sotelo, A. de Br ´ebisson, Y . Bengio, and A. C. Courville, “Melgan: Generative adversarial networks for conditional waveform synthesis,” in Proc. Adv. Neural Inf. Process. Syst. (NeurIPS) , vol. 32, Dec. 2019

    2019

  16. [23]

    Hifi-gan: Generative adversarial networks for efficient and high fidelity speech synthesis,

    J. Kong, J. Kim, and J. Bae, “Hifi-gan: Generative adversarial networks for efficient and high fidelity speech synthesis,” in Proc. Adv. Neural Inf. Process. Syst. (NeurIPS) , Dec. 2020

    2020

  17. [24]

    BigVGAN: A universal neural vocoder with large-scale training,

    S. Lee, W. Ping, B. Ginsburg, B. Catanzaro, and S. Yoon, “BigVGAN: A universal neural vocoder with large-scale training,” in Int. Conf. Learn. Represent. (ICLR), May 2023

    2023

  18. [25]

    Librosa: Audio and music signal analysis in python,

    B. McFee, C. Raffel, D. Liang, D. P. W. Ellis, M. McVicar, E. Bat- tenberg, and O. Nieto, “Librosa: Audio and music signal analysis in python,” in Proc. Python Science Conf. , Jul. 2015, pp. 18–24

    2015

  19. [26]

    A speech recon- struction algorithm via iteratively reweighted ℓ1 minimization for MFCC codec,

    G. Min, X. Zhang, X. Liu, C. Zhang, and Y . Chen, “A speech recon- struction algorithm via iteratively reweighted ℓ1 minimization for MFCC codec,” in Proc. Eur . Signal Process. Conf. (EUSIPCO) , Nov. 2019

    2019

  20. [28]

    S. Boyd, N. Parikh, E. Chu, B. Peleato, and J. Eckstein, Distributed optimization and statistical learning via the alternating direction method of multipliers . Now Publishers Inc., Jan. 2010

    2010

  21. [29]

    Local saddles of relaxed averaged alternating reflections algo- rithms on phase retrieval,

    P. Chen, “Local saddles of relaxed averaged alternating reflections algo- rithms on phase retrieval,” Inverse Probl., vol. 38, no. 1, Dec. 2021

    2021

  22. [30]

    Proximal algorithms,

    N. Parikh and S. Boyd, “Proximal algorithms,” F ound. Trends Optim., vol. 1, pp. 127–239, Jan. 2014

    2014

  23. [31]

    Consistent Wiener filtering for audio source separation,

    J. Le Roux and E. Vincent, “Consistent Wiener filtering for audio source separation,” IEEE Signal Process. Lett., vol. 20, pp. 217–220, Mar. 2013

    2013

  24. [32]

    A nonconvex approach for phase retrieval: Reshaped wirtinger flow and incremental algorithms,

    H. Zhang, Y . Zhou, Y . Liang, and Y . Chi, “A nonconvex approach for phase retrieval: Reshaped wirtinger flow and incremental algorithms,” J. Mach. Learn. Represent. , vol. 18, pp. 1–35, Nov. 2017

    2017

  25. [33]

    Inertial proximal alternating linearized mini- mization (iPALM) for nonconvex and nonsmooth problems,

    T. Pock and S. Sabach, “Inertial proximal alternating linearized mini- mization (iPALM) for nonconvex and nonsmooth problems,” SIAM J. Imag. Sci., vol. 9, no. 4, pp. 1756–1787, 2016

    2016

  26. [34]

    Alternating direction method of multipliers for non-negative matrix factorization with the beta-divergence,

    D. L. Sun and C. F ´evotte, “Alternating direction method of multipliers for non-negative matrix factorization with the beta-divergence,” in Proc. IEEE Int. Conf. Acoust., Speech, Signal Process. (ICASSP) , May 2014, pp. 6201–6205

    2014

  27. [35]

    A proper version of synthesis-based sparse audio declipper,

    P. Z ´aviˇska, P. Rajmic, O. Mokr `y, and Z. Pr ˚uˇsa, “A proper version of synthesis-based sparse audio declipper,” inProc. IEEE Int. Conf. Acoust., Speech Signal Process. (ICASSP) , May 2019, pp. 591–595

    2019

  28. [36]

    Joint ampli- tude and phase refinement for monaural source separation,

    Y . Masuyama, K. Yatabe, K. Nagatomo, and Y . Oikawa, “Joint ampli- tude and phase refinement for monaural source separation,” IEEE Sig- nal Process. Lett. , vol. 27, pp. 1939–1943, Oct. 2020

    1939

  29. [37]

    P.862.2, 2007

    P .862.2: Wideband extension to Recommendation P .862 for the assess- ment of wideband telephone networks and speech codecs , ITU-T Std. P.862.2, 2007

    2007

  30. [38]

    An algorithm for predicting the intelligibil- ity of speech masked by modulated noise maskers,

    J. Jensen and C. H. Taal, “An algorithm for predicting the intelligibil- ity of speech masked by modulated noise maskers,” IEEE/ACM Trans. Audio, Speech, Lang. Process. , vol. 24, pp. 2009–2022, Aug. 2016

    2009

  31. [39]

    Foley sound synthesis at the dcase 2023 challenge,

    K. Choi, J. Im, L. Heller, B. McFee, K. Imoto, Y . Okamoto, M. Lagrange, and S. Takamichi, “Foley sound synthesis at the dcase 2023 challenge,” arXiv:2304.12521, 2023

    2023 arXiv

  32. [40]

    Signal reconstruction from STFT magnitude: a state of the art,

    N. Strumel and L. Daudet, “Signal reconstruction from STFT magnitude: a state of the art,” in Int. Conf. Digit. Audio Effects (DAFx) , Sep. 2011, pp. 375–386

    2011

  33. [41]

    GLA- Grad: A Griffin-Lim extended waveform generation diffusion model,

    H. Liu, T. Baoueb, M. Fontaine, J. Le Roux, and G. Richard, “GLA- Grad: A Griffin-Lim extended waveform generation diffusion model,” in IEEE Int. Conf. Acoust., Speech, Signal Process. (ICASSP) , Apr. 2024, pp. 11 611–11 615

    2024

Pith tools

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