Pith. sign in

REVIEW 2 major objections 2 minor 50 references

PFANC processes multiple consecutive noise frames with a CRNN to predict the control filter for the next frame rather than the current one.

Reviewed by Pith at T0; open to challenge. T0 means a machine referee read the full paper against a public rubric. the ladder, T0–T4 →

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

T0 review · grok-4.3

2026-06-27 19:21 UTC pith:NZZ53H54

load-bearing objection PFANC adds a CRNN predictor for the next-frame filter to cut GFANC's lag, but the Markov justification needs checking against the actual test noises. the 2 major comments →

arxiv 2606.08171 v1 pith:NZZ53H54 submitted 2026-06-06 eess.AS

Predictive Fixed-Filter Active Noise Control (PFANC) Using Convolutional Recurrent Neural Networks for Dynamic Noises

Zhengding Luo , Haowen Li , Haozhe Ma , Dongyuan Shi , Wen Zhang , Woon-Seng Gan This is my paper
classification eess.AS
keywords active noise controlpredictive filterconvolutional recurrent neural networkdynamic noisefixed-filter ANCnoise trackingCRNN
verification ladder T0 review T1 audit T2 compute T3 formal T4 reserved

The pith

A machine-rendered reading of the paper's core claim, the machinery that carries it, and where it could break.

The paper introduces PFANC as a proactive alternative to the reactive GFANC approach in active noise control. GFANC estimates a filter suited to the present noise frame and therefore incurs a tracking delay when noises change quickly. PFANC instead feeds several prior frames into a convolutional recurrent neural network to forecast the filter needed for the upcoming frame. A supporting analysis treats the noise sequence as a high-order Markov chain and shows that longer histories improve filter prediction accuracy. Simulations on chirp signals and recorded dynamic noises confirm lower residual noise and better cross-path transfer than GFANC variants.

Core claim

The PFANC method uses a CRNN to map a sequence of consecutive noise frames to the control filter that will be optimal for the immediate future frame; the high-order Markov chain analysis establishes that the added temporal context reduces prediction error relative to single-frame estimation, and numerical tests on linear and logarithmic chirps plus real-world recordings demonstrate lower error power and improved adaptability across acoustic paths.

What carries the argument

The PFANC predictor: a convolutional recurrent neural network that ingests multiple past noise frames and outputs the control filter for the next frame.

Load-bearing premise

That the statistical dependence between successive noise frames is strong enough to let a learned model forecast the filter that will actually minimize noise in the next interval.

What would settle it

A controlled test in which PFANC produces higher residual noise power than GFANC on the same set of rapidly changing recorded noises would falsify the performance claim.

Watch this falsifier. Get emailed when new claim-graph text bears on it.

If this is right

  • Dynamic noises that vary faster than the filter update rate can be tracked with less lag.
  • The high-order Markov analysis implies that longer input sequences yield better filter predictions up to the point where additional frames add no new information.
  • The learned mapping transfers across different primary and secondary paths without retraining.
  • Real-time implementations can pre-compute the filter before the noise arrives rather than reacting after it is measured.

Where Pith is reading between the lines

These are editorial extensions of the paper, not claims the author makes directly.

  • The same sequence-to-filter idea could be tested in other adaptive systems such as echo cancellation or vibration control where future disturbances are partially predictable from recent history.
  • If the CRNN is replaced by a lighter recurrent model, the computational cost per frame might drop while retaining most of the predictive gain.
  • Combining PFANC with occasional full adaptive updates could handle sudden noise-type switches that the learned predictor has not seen.

Editorial analysis

A structured set of objections, weighed in public.

Desk editor's note, referee report, simulated authors' rebuttal, and a circularity audit.

Referee Report

2 major / 2 minor

Summary. The paper proposes Predictive Fixed-Filter Active Noise Control (PFANC), which uses a Convolutional Recurrent Neural Network (CRNN) to process multiple consecutive noise frames and predict the optimal control filter for the next frame, addressing the tracking lag inherent in the reactive Generative Fixed-Filter Active Noise Control (GFANC) approach. A theoretical analysis based on a high-order Markov chain is presented to show that multiple prior frames improve filter prediction. The method is validated through numerical simulations on linear and logarithmic chirp signals as well as real-world dynamic noises, with claims of superiority over GFANC variants and good transferability across acoustic paths.

Significance. If the results hold, PFANC represents a meaningful step toward proactive control in active noise cancellation for non-stationary environments, potentially enabling better performance in applications with rapidly changing noises. The integration of CRNN-based prediction with fixed-filter ANC is a clear technical contribution, and the provision of both theoretical justification and simulation evidence strengthens the work, though the practical impact hinges on the magnitude and robustness of the reported gains.

major comments (2)
  1. [Theoretical analysis] Theoretical analysis (high-order Markov chain section): the claim that multiple frames enhance prediction rests on the assumption that the underlying noise process is Markovian of finite order; however, no diagnostic is reported confirming that the chirp signals or real dynamic noises satisfy this property, which directly affects whether the theoretical justification applies to the experimental results.
  2. [Numerical simulations] Numerical simulations section: while superiority over GFANC is asserted for chirps and real noises, the manuscript does not detail the quantitative metrics (e.g., noise attenuation in dB, convergence speed), statistical controls, or ablation on frame count, making it impossible to evaluate whether the predictive advantage is load-bearing or sensitive to the specific training distribution.
minor comments (2)
  1. [Abstract / Experiments] The abstract states that PFANC 'exhibits good transferability across different acoustic paths,' but the corresponding experimental protocol (training vs. test paths, quantitative transfer metrics) should be clarified for reproducibility.
  2. [Method] Notation for the CRNN input (number of frames, feature dimensions) and the precise definition of the 'next-frame control filter' target could be made more explicit to aid readers unfamiliar with the GFANC baseline.

Simulated Author's Rebuttal

2 responses · 0 unresolved

We thank the referee for the thoughtful and constructive comments, which highlight important aspects of the theoretical justification and experimental reporting. We address each major comment below and will incorporate revisions to strengthen the manuscript.

read point-by-point responses

  1. Referee: [Theoretical analysis] Theoretical analysis (high-order Markov chain section): the claim that multiple frames enhance prediction rests on the assumption that the underlying noise process is Markovian of finite order; however, no diagnostic is reported confirming that the chirp signals or real dynamic noises satisfy this property, which directly affects whether the theoretical justification applies to the experimental results.

    Authors: We agree that explicitly linking the high-order Markov chain model to the experimental signals would strengthen the connection between theory and results. The Markov analysis is intended as a general framework demonstrating the benefit of multiple prior frames for prediction under temporal dependence, rather than a claim that all tested noises are exactly finite-order Markov processes. Chirp signals are deterministic but exhibit predictable temporal evolution, while real-world noises often display short-term correlations. In the revised manuscript, we will add a diagnostic subsection (e.g., autocorrelation analysis and estimated Markov order via AIC/BIC on the chirp and recorded noise datasets) to clarify the degree to which the assumption holds and to discuss any limitations. revision: yes

  2. Referee: [Numerical simulations] Numerical simulations section: while superiority over GFANC is asserted for chirps and real noises, the manuscript does not detail the quantitative metrics (e.g., noise attenuation in dB, convergence speed), statistical controls, or ablation on frame count, making it impossible to evaluate whether the predictive advantage is load-bearing or sensitive to the specific training distribution.

    Authors: We acknowledge that the current presentation of results would benefit from greater quantitative detail and controls to allow readers to assess the magnitude and robustness of the reported gains. The manuscript includes comparative simulations but does not provide the requested metrics or ablations in sufficient depth. In the revision, we will expand the numerical simulations section with: (i) explicit tables reporting average noise attenuation (dB), convergence speed, and standard deviations across multiple runs; (ii) statistical significance tests (e.g., paired t-tests or Wilcoxon tests) against GFANC variants; and (iii) an ablation study varying the number of input frames (1, 2, 4, 8) with corresponding performance curves. These additions will also include details on training/validation splits to address sensitivity to the training distribution. revision: yes

Circularity Check

0 steps flagged

No significant circularity detected

full rationale

The PFANC method trains a CRNN on noise frames to output a predicted control filter for the next frame; this is a learned mapping, not a quantity defined by construction from the same inputs. The high-order Markov chain analysis is presented as theoretical justification for using multiple frames and does not reduce the empirical prediction step to a tautology or self-fit. Simulations on chirp signals and real recordings provide external benchmarks. No self-citation load-bearing steps, ansatz smuggling, or renaming of known results appear in the provided text. The derivation chain remains independent of its own fitted outputs.

Axiom & Free-Parameter Ledger

0 free parameters · 1 axioms · 0 invented entities

Review performed on abstract only; no explicit free parameters, invented entities, or additional axioms beyond the domain assumption of temporal predictability are stated.

axioms (1)
  • domain assumption Temporal correlations across consecutive noise frames allow prediction of the optimal next-frame control filter.
    Invoked to justify the proactive paradigm and the high-order Markov chain analysis.

pith-pipeline@v0.9.1-grok · 5756 in / 1105 out tokens · 23277 ms · 2026-06-27T19:21:00.003444+00:00 · methodology

0 comments
read the original abstract

The existing Generative Fixed-Filter Active Noise Control (GFANC) method generates a suitable control filter based on the current noise frame. This reactive design aims to estimate a control filter that is optimal for the present frame rather than the upcoming one. Consequently, it suffers from an inherent tracking lag and lacks the predictive capability to handle rapidly varying noises. To address this limitation, we propose the Predictive Fixed-Filter Active Noise Control (PFANC) method with a proactive control paradigm in this paper. In the PFANC method, multiple consecutive noise frames are processed by a Convolutional Recurrent Neural Network (CRNN) to predict the next-frame control filter. By utilizing temporal correlations across noise frames to anticipate the control filter in advance, the PFANC method can effectively track dynamic noise changes. Furthermore, the theoretical analysis based on a high-order Markov chain shows that incorporating multiple noise frames enhances the prediction of the control filter. Numerical simulations with linear and logarithmic chirp signals, as well as real-world dynamic noises, validate the effectiveness of the PFANC method and its superiority over GFANC and its variations. The PFANC method also exhibits good transferability across different acoustic paths.

Figures

Figures reproduced from arXiv: 2606.08171 by Dongyuan Shi, Haowen Li, Haozhe Ma, Wen Zhang, Woon-Seng Gan, Zhengding Luo.

Figure 1
Figure 1. Figure 1: Frequency spectrum of (a) the pre-trained broadband control filter and [PITH_FULL_IMAGE:figures/full_fig_p006_1.png] view at source ↗
Figure 2
Figure 2. Figure 2: Schematic of the PFANC method, where the CRNN runs on a co [PITH_FULL_IMAGE:figures/full_fig_p009_2.png] view at source ↗
Figure 3
Figure 3. Figure 3: Block diagram of the CRNN model, which learn dynamics from [PITH_FULL_IMAGE:figures/full_fig_p010_3.png] view at source ↗
Figure 4
Figure 4. Figure 4: Equivalent high-order Markov chain of ANC progress. [PITH_FULL_IMAGE:figures/full_fig_p015_4.png] view at source ↗
Figure 5
Figure 5. Figure 5: Training and validation loss evolution during CRNN model training. [PITH_FULL_IMAGE:figures/full_fig_p019_5.png] view at source ↗
Figure 6
Figure 6. Figure 6: Spectrograms of the error signal obtained (a) without ANC, and with [PITH_FULL_IMAGE:figures/full_fig_p021_6.png] view at source ↗
Figure 7
Figure 7. Figure 7: Spectrograms of the error signal obtained (a) without ANC, and with [PITH_FULL_IMAGE:figures/full_fig_p022_7.png] view at source ↗
Figure 8
Figure 8. Figure 8: Spectral flux of four types of noises: (a) 20–1 [PITH_FULL_IMAGE:figures/full_fig_p023_8.png] view at source ↗
Figure 9
Figure 9. Figure 9: Error signals (a)–(c) and the averaged noise reduction level per second [PITH_FULL_IMAGE:figures/full_fig_p025_9.png] view at source ↗
Figure 10
Figure 10. Figure 10: Error signals (a)–(c) and the averaged noise reduction level per second [PITH_FULL_IMAGE:figures/full_fig_p026_10.png] view at source ↗
Figure 11
Figure 11. Figure 11: Spectrograms of the error signal for the 20–1 [PITH_FULL_IMAGE:figures/full_fig_p027_11.png] view at source ↗
Figure 12
Figure 12. Figure 12: Spectrograms of the error signal for the 20–1 [PITH_FULL_IMAGE:figures/full_fig_p028_12.png] view at source ↗
Figure 7
Figure 7. Figure 7: Furthermore, the overall NR values in System-B are lower than those in [PITH_FULL_IMAGE:figures/full_fig_p028_7.png] view at source ↗

discussion (0)

Sign in with ORCID, Apple, or X to comment. Anyone can read and Pith papers without signing in.

Reference graph

Works this paper leans on

50 extracted references

  1. [1]

    S. M. Kuo, D. R. Morgan, Active noise control: a tutorial review, Proceed- ings of the IEEE 87 (6) (1999) 943–973

    1999

  2. [2]

    C. N. Hansen, Understanding active noise cancellation, CRC Press, 2002

    2002

  3. [3]

    Kajikawa, W.-S

    Y . Kajikawa, W.-S. Gan, S. M. Kuo, Recent advances on active noise control: open issues and innovative applications, APSIPA Transactions on Signal and Information Processing 1 (2012) e3

    2012

  4. [4]

    F. Yang, J. Guo, J. Yang, Stochastic analysis of the filtered-x lms algorithm for active noise control, IEEE /ACM Transactions on Audio, Speech, and Language Processing 28 (2020) 2252–2266

    2020

  5. [5]

    Huang, J

    G. Huang, J. R. Jensen, J. Chen, J. Benesty, M. G. Christensen, A. Sugiyama, G. Elko, T. Gaensler, Advances in microphone array processing and multi- channel speech enhancement, in: ICASSP 2025-2025 IEEE International Conference on Acoustics, Speech and Signal Processing (ICASSP), IEEE, 2025, pp. 1–5

    2025

  6. [6]

    Zheng, X.-L

    N. Zheng, X.-L. Zhang, Phase-aware speech enhancement based on deep neural networks, IEEE/ACM Transactions on Audio, Speech, and Language Processing 27 (1) (2018) 63–76

    2018

  7. [7]

    Cheer, S

    J. Cheer, S. J. Elliott, Multichannel control systems for the attenuation of interior road noise in vehicles, Mechanical Systems and Signal Processing 60 (2015) 753–769

    2015

  8. [8]

    S. J. Elliott, P. A. Nelson, Active noise control, IEEE signal processing mag- azine 10 (4) (1993) 12–35

    1993

  9. [9]

    Y . Zhou, H. Zhao, D. Liu, H. Zhang, Adaptive narrowband active noise con- trol trained by genetic algorithm, IEEE Transactions on Instrumentation and Measurement (2024)

    2024

  10. [10]

    Aboutiman, K

    A. Aboutiman, K. S. A. Maamoun, H. R. Karimi, F. Ripamonti, Hybrid deep learning–based active noise control for encapsulated structures with open- ings, Expert Systems with Applications (2026) 131247. 31

    2026

  11. [11]

    Zhang, T

    J. Zhang, T. D. Abhayapala, W. Zhang, P. N. Samarasinghe, S. Jiang, Active noise control over space: A wave domain approach, IEEE /ACM Transac- tions on audio, speech, and language processing 26 (4) (2018) 774–786

    2018

  12. [12]

    L. Bai, S. Lian, M. Li, Y . He, L. Rao, X. Zeng, R. Sun, K. Chen, J. Lu, Wavenet-volterra neural network for active noise control: A fully causal ap- proach, Mechanical Systems and Signal Processing 241 (2025) 113486

    2025

  13. [13]

    Liang, F

    C. Liang, F. Ripamonti, H. R. Karimi, S. Wrona, M. Pawełczyk, A stepwise simultaneous perturbation stochastic approximation algorithm for stability improvement of active noise control systems, Mechanical Systems and Sig- nal Processing 237 (2025) 112915

    2025

  14. [14]

    Iotov, S

    Y . Iotov, S. M. Nørholm, V . Belyi, M. Dyrholm, M. G. Christensen, Compu- tationally efficient fixed-filter anc for speech based on long-term prediction for headphone applications, in: ICASSP 2022-2022 IEEE International Con- ference on Acoustics, Speech and Signal Processing (ICASSP), IEEE, 2022, pp. 761–765

    2022

  15. [15]

    Q. Xu, S. Wang, J. Tao, H. Zou, X. Qiu, Creating personal sound zones in car cabins with active noise control, Applied Acoustics 235 (2025) 110670

    2025

  16. [16]

    Chang, C.-T

    C.-Y . Chang, C.-T. Chuang, S. M. Kuo, C.-H. Lin, Multi-functional active noise control system on headrest of airplane seat, Mechanical Systems and Signal Processing 167 (2022) 108552

    2022

  17. [17]

    C. R. Fuller, A. H. von Flotow, Active control of sound and vibration, IEEE Control Systems Magazine 15 (6) (1995) 9–19

    1995

  18. [18]

    Schumacher, H

    T. Schumacher, H. Kr ¨uger, M. Jeub, P. Vary, C. Beaugeant, Active noise control in headsets: A new approach for broadband feedback anc, in: 2011 IEEE International conference on acoustics, speech and signal processing (ICASSP), IEEE, 2011, pp. 417–420

    2011

  19. [19]

    Z. Luo, D. Shi, W.-S. Gan, A hybrid sfanc-fxnlms algorithm for active noise control based on deep learning, IEEE Signal Processing Letters 29 (2022) 1102–1106

    2022

  20. [20]

    J. Luo, K. Chen, L. Yang, J. Zhang, F. Tian, L. Wang, J. Qiao, A multi- strategy enhanced algorithm for active impulsive noise control, Mechanical Systems and Signal Processing 247 (2026) 113946. 32

    2026

  21. [21]

    Lu, K.-L

    L. Lu, K.-L. Yin, R. C. de Lamare, Z. Zheng, Y . Yu, X. Yang, B. Chen, A survey on active noise control in the past decade–part ii: Nonlinear systems, Signal Processing 181 (2021) 107929

    2021

  22. [22]

    D. Gao, H. Yao, X. Peng, S. Liang, J. Yang, J. Li, Active and passive noise control based on helmholtz resonator acoustic metamaterials and edgf- hafxlms algorithm, Mechanical Systems and Signal Processing 244 (2026) 113798

    2026

  23. [23]

    B. Wang, D. Shi, Z. Luo, X. Shen, J. Ji, W.-S. Gan, Transferable selective virtual sensing active noise control technique based on metric learning, in: ICASSP 2025 - 2025 IEEE International Conference on Acoustics, Speech and Signal Processing (ICASSP), 2025, pp. 1–5

    2025

  24. [24]

    L. Bhan, T. Murao, C. Shi, W.-S. Gan, S. Elliott, Feasibility of the full-rank fixed-filter approach in the active control of noise through open windows, in: INTER-NOISE and NOISE-CON Congress and Conference Proceedings, V ol. 253, Institute of Noise Control Engineering, 2016, pp. 3548–3555

    2016

  25. [25]

    Iotov, S

    Y . Iotov, S. M. Nørholm, V . Belyi, M. G. Christensen, Adaptive sparse lin- ear prediction in fixed-filter anc headphone applications for multi-speaker speech reduction, in: 2023 IEEE Workshop on Applications of Signal Pro- cessing to Audio and Acoustics (W ASPAA), IEEE, 2023, pp. 1–5

    2023

  26. [26]

    Miyoshi, Y

    M. Miyoshi, Y . Kaneda, Inverse filtering of room acoustics, IEEE Transac- tions on acoustics, speech, and signal processing 36 (2) (1988) 145–152

    1988

  27. [27]

    K. Doi, Y . Kajikawa, Sfanc with compensation filter based on mefxdctlms algorithm, in: 2023 Asia Pacific Signal and Information Processing Asso- ciation Annual Summit and Conference (APSIPA ASC), IEEE, 2023, pp. 1240–1244

    2023

  28. [28]

    X. Su, D. Shi, Z. Zhu, W.-S. Gan, L. Ye, Spatial-frequency-based selec- tive fixed-filter algorithm for multichannel active noise control, IEEE Signal Processing Letters 31 (2024) 2635–2639

    2024

  29. [29]

    S. Park, D. Park, Low-power fpga realization of lightweight active noise cancellation with cnn noise classification, Electronics 12 (11) (2023) 2511. 33

    2023

  30. [30]

    L. Yin, Z. Zhang, M. Wu, S. Zhou, J. Guo, J. Yang, J. Zhang, Selective fixed-filter active noise control based on frequency response matching in headphones, Applied Acoustics 211 (2023) 109505

    2023

  31. [31]

    D. Shi, B. Lam, K. Ooi, X. Shen, W.-S. Gan, Selective fixed-filter active noise control based on convolutional neural network, Signal Processing 190 (2022) 108317

    2022

  32. [32]

    Z. Luo, H. Ma, B. Wang, D. Shi, W.-S. Gan, Reinforcement learning-based selective fixed-filter active noise control (rl-sfanc): From theory to real-time headphone implementation, Signal Processing (2026) 110695

    2026

  33. [33]

    Aboutiman, Z

    A. Aboutiman, Z. Rachman, T. Oberman, F. Alletta, J. Kang, H. R. Karimi, F. Ripamonti, Subjective perception analysis of active noise control algo- rithms in an encapsulated structure: An experimental study, Applied Acous- tics 239 (2025) 110823

    2025

  34. [34]

    Z. Luo, J. Ji, B. Wang, D. Shi, H. Ma, W.-S. Gan, Deep learning-based gen- erative fixed-filter active noise control: Transferability and implementation, Mechanical Systems and Signal Processing 238 (2025) 113207

    2025

  35. [35]

    Z. Luo, D. Shi, W.-S. Gan, Q. Huang, Delayless generative fixed-filter active noise control based on deep learning and bayesian filter, IEEE/ACM Trans- actions on Audio, Speech, and Language Processing 32 (2024) 1048–1060

    2024

  36. [36]

    Z. Luo, D. Shi, X. Shen, J. Ji, W.-S. Gan, Gfanc-kalman: Generative fixed- filter active noise control with cnn-kalman filtering, IEEE Signal Processing Letters 31 (2024) 276–280

    2024

  37. [37]

    Zhang, A

    H. Zhang, A. Pandey, et al., Low-latency active noise control using attentive recurrent network, IEEE /ACM Transactions on Audio, Speech, and Lan- guage Processing 31 (2023) 1114–1123

    2023

  38. [38]

    Mostafavi, Y .-J

    A. Mostafavi, Y .-J. Cha, Deep learning-based active noise control on con- struction sites, Automation in Construction 151 (2023) 104885

    2023

  39. [39]

    Y .-J. Cha, A. Mostafavi, S. S. Benipal, Dnoisenet: Deep learning-based feed- back active noise control in various noisy environments, Engineering Appli- cations of Artificial Intelligence 121 (2023) 105971. 34

    2023

  40. [40]

    Zhang, D

    H. Zhang, D. Wang, Deep anc: A deep learning approach to active noise control, Neural Networks 141 (2021) 1–10

    2021

  41. [41]

    Singh, R

    D. Singh, R. Gupta, A. Kumar, R. Bahl, Enhancing active noise control through stacked autoencoders: Training strategies, comparative analysis, and evaluation with practical setup, Engineering Applications of Artificial Intel- ligence 135 (2024) 108811

    2024

  42. [42]

    J. Y . Oh, H. W. Jung, M. H. Lee, K. H. Lee, Y . J. Kang, Enhancing active noise control of road noise using deep neural network to update secondary path estimate in real time, Mechanical Systems and Signal Processing 206 (2024) 110940

    2024

  43. [43]

    Xiang, Y

    Q. Xiang, Y . Chu, M. Wu, G. Yu, A combined method for multi-channel active noise control based on online adaptive estimation and o ffline convo- lutional neural network, The Journal of the Acoustical Society of America 154 (4 supplement) (2023) A79–A79

    2023

  44. [44]

    Jozefowicz, W

    R. Jozefowicz, W. Zaremba, I. Sutskever, An empirical exploration of recur- rent network architectures, in: International conference on machine learning, PMLR, 2015, pp. 2342–2350

    2015

  45. [45]

    Y . Yu, X. Si, C. Hu, J. Zhang, A review of recurrent neural networks: Lstm cells and network architectures, Neural computation 31 (7) (2019) 1235– 1270

    2019

  46. [46]

    H. Li, W. Zhang, L. Zhang, Doa estimation of room reflections using nn- based music algorithm, in: 2023 Asia Pacific Signal and Information Pro- cessing Association Annual Summit and Conference (APSIPA ASC), IEEE, 2023, pp. 1960–1965

    2023

  47. [47]

    T. M. Cover, Elements of information theory, John Wiley & Sons, 1999

    1999

  48. [48]

    Y . A. Zhang, F. Ma, T. D. Abhayapala, P. N. Samarasinghe, A. Bastine, An active noise control system based on soundfield interpolation using a physics-informed neural network, in: ICASSP 2024-2024 IEEE Interna- tional Conference on Acoustics, Speech and Signal Processing (ICASSP), IEEE, 2024, pp. 506–510. 35

    2024

  49. [49]

    D. Chen, L. Cheng, D. Yao, J. Li, Y . Yan, A secondary path-decoupled ac- tive noise control algorithm based on deep learning, IEEE Signal Processing Letters 29 (2021) 234–238

    2021

  50. [50]

    S. O. Sadjadi, J. H. Hansen, Unsupervised speech activity detection using voicing measures and perceptual spectral flux, IEEE signal processing letters 20 (3) (2013) 197–200. 36

    2013