Pith. sign in

REVIEW 3 major objections 5 minor 1 cited by

DNN based HRIRs Identification with a Continuously Rotating Speaker Array

T0 review · 3 major / 5 minor · reviewed 2026-08-16 · deepseek-v4-flash

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

Pith's one-line read This paper claims that a sequence-to-sequence DNN can identify time-varying head-related impulse responses from a continuously rotating speaker array, achieving normalized misalignment of -25.58 dB in simulation and -23.21 dB in…

desk verdict A genuinely new DNN-based HRIR identification method with solid but single-trial evidence; the per-recording training scheme deserves closer scrutiny. read the letter →

arxiv 2504.14817 v1 pith:PXCYKHTL submitted 2025-04-21 eess.AS

Byeong-Yun Ko , Deokki Min , Hyeonuk Nam , Yong-Hwa Park This is my paper · ORCID
classification eess.AS
keywords head-relatedimpulseresponseHRIRidentificationdynamicHRTFmeasurementrotatingspeakerarraysequence-to-sequencelearninggatedrecurrentunitadaptivesystemlearnablenormalization
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 aims to make individualized head-related impulse response (HRIR) measurement much faster by letting the speaker array rotate continuously while all speakers play simultaneously, then recovering the time-varying HRIRs from the single microphone output. It claims that a DNN built from gated recurrent units and fully connected layers can do this by replacing the fixed update rule of an adaptive filter: at each time step the network reads the gradient of the instantaneous squared error, the excitation power, and a hidden state, and outputs a coefficient update for the HRIR vector. Because the update rule is optimized on the measurement itself, the method does not need a training dataset of impulse responses, unlike existing DNN-based adaptive filters. The paper reports that this scheme outperforms NLMS, Kalman filtering, JO-NLMS, and Meta-AF, with normalized misalignment of -25.58 dB in simulation and -23.21 dB in experiment, and preserves interaural time differences below the just-noticeable limit.

What carries the argument

The load-bearing object is the DNN update rule in Eq. (6), $\tilde{\mathbf{h}}_{n+1,\mathrm{ele}} = \tilde{\mathbf{h}}_{n,\mathrm{ele}} + g_\phi\{\nabla \mathrm{ISE}(n), \mathbf{x}_{n,\mathrm{ele}}\mathbf{x}_{n,\mathrm{ele}}^\top, \mathbf{c}_n\}$, in which a gated recurrent network turns the gradient of the instantaneous squared error into a coefficient update for the whole HRIR vector. The reset and update gates of the gated recurrent unit carry state across long sequences, the three-layer fully connected block models the nonlinear reflection and diffraction transitions of HRIRs, and the learnable normalization, an input derived from the excitation power, controls the update scale adaptively. The accompanying whole-sequence updating and optimization scheme alternates between rolling the HRIR estimate forward over all time steps with the network frozen and then optimizing the network with backpropagation through time on the accumulated error, which lets the system train without any labeled impulse-response dataset.

What would settle it

Re-run the whole-sequence update on the same binaural recording from different random initializations of the DNN weights; if the identified HRIR sequences differ by more than the reported normalized-misalignment or spectral-distortion margin, then the underdetermined system is being settled by hidden regularization rather than by the data. Equally direct: hold out one azimuth segment of the recording, optimize only on the rest, and ask whether the held-out HRIRs are recovered; failure would mean the identification is fitting the recording instead of discovering a transferable head response.

Watch

Extended reading notes

Core claim

The central claim is that the underdetermined problem of identifying a whole elevation plane of HRIRs from a single scalar microphone signal becomes tractable when the identification rule itself is learned rather than fixed. On a continuously rotating array, the HRIR vector changes smoothly with azimuth, and the proposed network exploits that continuity through a gated recurrent structure: reset and update gates let the update process remember past HRIR states over sequences of hundreds of thousands of samples, while three dense layers convert the instantaneous-error gradient into the next coefficient update. A learnable normalization driven by the speaker excitation power stabilizes the otherwise runaway scale of the gradient. The paper shows that with this update rule, identified HRIRs match ground truth to -25.58 dB normalized misalignment and 1.74 dB log spectral distortion in simulation, and -23.21 dB and 1.98 dB in experiment, and that localization cues, including interaural time differences below 12 microseconds and spectral peaks and notches in the 4-16 kHz range, survive the identification.

Load-bearing premise

The load-bearing premise is that a single microphone output per sample, which mathematically leaves the HRIR coefficients underdetermined, still fixes the true HRIR sequence once the DNN's smoothness and continuity priors are applied, so the method's accuracy could be an effect of those priors rather than of information in the recording.

Editorial extensions

If this is right

  • A full spherical HRIR set for an individual can in principle be captured in a few seconds of continuous rotation rather than minutes of stepped repositioning.
  • The method removes the need for a pre-measured corpus of impulse responses for DNN-based identification, so it can adapt to a new subject or a new measurement rig from the single recording itself.
  • Perceptually relevant localization cues survive the fast-rotation identification: simulation interaural time differences stay below 12 microseconds, and spectral notches and peaks in the 4-16 kHz band are reconstructed closely enough to support azimuth and elevation cues.
  • At 45 degrees per second the proposed method beats analytic adaptive filters by more than 7 dB in normalized misalignment in simulation, showing the neural update rule tracks faster rotations than the analytical models tolerate.

Reading between the lines

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

  • Editorial inference: the same DNN-in-the-loop coefficient update is not ear-specific, so it could transfer to other time-varying acoustic system identification problems, such as moving sources or changing room responses, wherever one scalar output must resolve many coefficients.
  • Editorial inference: because the underlying equation is underdetermined, the reported accuracy probably reflects the network's implicit bias toward smooth HRIR sequences as much as the data; a comparison against a smoothness-constrained solver with the same priors would separate identification from regularization.
  • Editorial inference: the per-measurement training cost, multi-hour and multi-GPU for one sequence, is the practical barrier; amortizing or warm-starting the optimization across subjects or angles is a natural next step that the paper leaves open.
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

3 major / 5 minor

Summary. The paper proposes a deep-neural-network-based method for identifying time-varying head-related impulse responses from a continuously rotating speaker array. The method replaces the fixed step-size update of an adaptive filter with a GRU-plus-FC network that takes the gradient of the instantaneous squared error, the excitation-signal power, and a hidden state as inputs, and outputs an additive update to the HRIR vector. A learnable normalization and a whole-sequence updating/optimization scheme are introduced, and training is performed without a labeled HRIR dataset by minimizing the logarithm of the mean squared prediction error over the recording. The method is evaluated in simulation with interpolated FABIAN HRIRs at 45°/s and in a KEMAR experiment with a custom nine-speaker arc array, reporting NM of -25.58 dB and -23.21 dB and LSD of 1.74 dB and 1.98 dB, respectively, outperforming NLMS, Kalman filter, JO-NLMS, and Meta-AF, with ITD errors below 12 μs.

Significance. If the results are reproducible, this is a valuable contribution: it offers a route to fast individualized HRIR measurement that avoids static repositioning and remains accurate at high rotation speeds, with objective validation against external ground truth in both simulation and hardware. Strengths of the paper are the direct time-domain/frequency-domain comparison to measured or interpolated HRIRs, the inclusion of an ablation study, public source code, and a real KEMAR experiment. The main limitations are that the evaluation consists of single-point estimates from one recording per condition and that the relation between the implemented segment-wise processing and the claimed whole-sequence optimization is not reconciled. These issues affect the strength of the central generalization claim but are addressable with additional experiments and clarification.

major comments (3)
  1. [V-A, VI-A, and Table 1] Section V-A reports that the full 176,400-sample sequence was divided into 10 segments of 17,640 samples and processed in parallel on 4 GPUs, and Section VI-A says the same was done for the 170,000-sample experimental recording. This is difficult to reconcile with the ablation in Table 1, where the 'N/10 sequence updating and optimization' condition yields NM of 75.29 dB and LSD of 550.71 dB while the proposed method yields -25.58 dB and 1.74 dB. If the reported results are obtained from 10 independent segments, the training loss in Eq. (8) is minimized over 17,640 samples per segment rather than over the whole sequence, so the 'whole-sequence' scheme is not what is evaluated. Please clarify whether the segments share DNN parameters and hidden state, how segment boundaries are handled, and why the N/10 ablation is catastrophic while the parallel segmentation works.
  2. [Tables I and III, Eq. (8)] The NM and LSD numbers in Tables I and III are point estimates from a single simulated noise realization and a single experimental measurement, respectively. Because the DNN is optimized by minimizing the prediction error in Eq. (8) on the same recording that is later used to compute the metrics, the 53-million-parameter model has ample capacity to memorize recording-specific noise or reflections; the reported values therefore do not yet demonstrate that the identified HRIRs generalize to a second recording of the same subject under the same rotation protocol. I request repeated independent noise draws in simulation and repeated rotations in the experiment, reported as mean ± standard deviation, or an explicit out-of-sample evaluation.
  3. [Section II, Eq. (2), Eq. (6), Eq. (8)] Section II correctly acknowledges that Eq. (2) is underdetermined at each time index, and Section III adds only informal assumptions of slow variation and spatial continuity. The paper never states under what conditions the update rule in Eq. (6), trained with the loss in Eq. (8), recovers the true time-varying HRIR rather than an arbitrary sequence that reproduces y(n). Since the training objective constrains only the microphone output, the favorable NM/LSD values could reflect implicit regularization rather than a guaranteed identification. Please provide either a formal identifiability/convergence analysis for the DNN update or a sensitivity study (varying initialization, noise level, rotation speed, and excitation signal) that supports the claim of accurate HRIR identification beyond the specific tested instance.
minor comments (5)
  1. [V-A and V-B] In Section V-A the estimated HRIR length is set to K~=192, but Section V-B states zero-padding matches K~=196 to K=256; please correct the value and the corresponding zero-padding description.
  2. [IV-B] Section IV-B's example uses a sampling rate of 48 kHz, whereas the simulation uses 44.1 kHz and the experiment uses 34 kHz; please unify the notation.
  3. [Eq. (11)] Equation (11) as printed does not show the dependence of the correlation sum on τ; please provide the full cross-correlation formula and the range of τ searched.
  4. [Section II] The phrase 'sampling period in the millisecond range' is likely a typo; for a 44.1 kHz rate the period is about 22.7 μs, not milliseconds.
  5. [Fig. 8] The notation (θ, φ) in the Fig. 8 titles is used without an explicit definition of azimuth and elevation conventions in the text; please add a sentence defining them.

Circularity Check

0 steps flagged · score 0.0 of 10

No significant circularity: the proposed DNN update rule is trained on the observed recording but evaluated against external ground-truth HRIRs, so the reported NM/LSD numbers are not forced by construction.

full rationale

The paper's derivation chain is self-contained with respect to circularity. The core update rule of Eq. (6) uses the DNN g_phi to map the ISE gradient, excitation-signal power, and hidden state into a coefficient update for the estimated HRIR vector h_tilde. The training loss in Eq. (8) is the logarithm of the mean squared prediction error of the observed microphone signal y(n), not the NM or LSD evaluation metrics. Evaluation is performed against external ground truth: interpolated FABIAN HRIRs in simulation and stepwise MESM measurements of the KEMAR HATS in experiment. Thus the reported NM of -25.58 dB and LSD of 1.74 dB in simulation, and -23.21 dB and 1.98 dB in experiment, are genuine comparisons of the identified impulse responses with independently measured or database-derived references, not quantities that reduce by construction to the training objective. The underdetermined nature of Eq. (2) is acknowledged explicitly, and the paper's contribution is precisely an empirical claim that the DNN resolves this ill-posed problem well in the tested configuration. That claim could be weakened by the absence of out-of-sample or repeated-trial validation, since the same recording is used for both optimization and evaluation; however, this is a generalization and overfitting risk, not a circularity in the sense of a prediction that is equivalent to its input by definition. The small number of self-citations by the authors (e.g., refs. [8] and [14]) appear only in background or method-motivation contexts and are not load-bearing for the central result; no uniqueness theorem or fitted constant is imported from those citations. Therefore no specific circular step can be exhibited, and the appropriate finding is no significant circularity.

Assumptions & free parameters 5 free parameters · 4 assumptions · 0 invented entities

The central claim inherits standard adaptive-filtering assumptions: slowly varying HRIRs, perfect-sweep orthogonality, and spatial continuity. It also implicitly assumes that the underdetermined identification problem is resolvable by the DNN update rule, which is not proven. No new physical entities are introduced.

free parameters (5)
  • Estimated HRIR length K-tilde = 192 (simulation), 340 (experiment)
    Chosen by hand to capture head and torso effects; changes the update-vector dimension and the perfect-sweep period.
  • DNN architecture size = 53 million parameters
    No ablation of width or depth is reported; the architecture is chosen ad hoc.
  • Training hyperparameters = learning rate 1e-4, momentum 0.9, max 300 epochs, Adam
    Set to avoid exploding gradients; not optimized per scenario.
  • Sequence segmentation = N/10 = 17,640 (simulation), 17,000 (experiment)
    Chosen for computational feasibility; the effect of broken continuity at segment boundaries is not analyzed.
  • Simulation noise variance = 0.01 (SNR 30 dB)
    Set to represent a typical measurement environment; sensitivity of results to this value is not reported.
assumptions (4)
  • domain assumption HRIR duration is about 23 ms and spatial sampling at 5 degree intervals yields continuous HRIR variation, so rotation speeds up to 217 deg/s are valid.
    Section III; sets the validity range for dynamic identification, and 45 deg/s is assumed to lie within it.
  • standard math Perfect sweep sequences have zero autocorrelation except at zero lag, and circular shifts cancel cross-correlation between speakers.
    Section V-A and Section IV; this property is used to make the one-equation-per-sample system practically identifiable.
  • domain assumption The linear time-varying model y(n) = x^T h_n + v(n) with slowly varying h_n accurately describes the binaural recording.
    Eq. (1)-(2), Section II; the entire identification procedure rests on this model.
  • ad hoc to paper The DNN update rule in Eq. (6) can recover the true time-varying HRIR despite the underdetermined one-output-per-sample problem.
    Section II acknowledges non-unique solutions; the paper provides no identifiability proof, only empirical validation.

how reviews work

0 comments
Cite this review

Pith. "Pith review of DNN based HRIRs Identification with a Continuously Rotating Speaker Array." pith.science (2026). https://pith.science/paper/PXCYKHTL

@misc{pith2026250414817,
  author       = {Pith},
  title        = {Pith review of: DNN based HRIRs Identification with a Continuously Rotating Speaker Array},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/PXCYKHTL}},
  note         = {Machine review of arXiv:2504.14817}
}
read the original abstract

Conventional static measurement of head-related impulse responses (HRIRs) is time-consuming due to the need for repositioning a speaker array for each azimuth angle. Dynamic approaches using analytical models with a continuously rotating speaker array have been proposed, but their accuracy is significantly reduced at high rotational speeds. To address this limitation, we propose a DNN-based HRIRs identification using sequence-to-sequence learning. The proposed DNN model incorporates fully connected (FC) networks to effectively capture HRIR transitions and includes reset and update gates to identify HRIRs over a whole sequence. The model updates the HRIRs vector coefficients based on the gradient of the instantaneous square error (ISE). Additionally, we introduce a learnable normalization process based on the speaker excitation signals to stabilize the gradient scale of ISE across time. A training scheme, referred to as whole-sequence updating and optimization scheme, is also introduced to prevent overfitting. We evaluated the proposed method through simulations and experiments. Simulation results using the FABIAN database show that the proposed method outperforms previous analytic models, achieving over 7 dB improvement in normalized misalignment (NM) and maintaining log spectral distortion (LSD) below 2 dB at a rotational speed of 45{\deg}/s. Experimental results with a custom-built speaker array confirm that the proposed method successfully preserved accurate sound localization cues, consistent with those from static measurement. Source code is available at https://github.com/byko0810/DNN-based-HRIRs-identification

Figures

Figures reproduced from arXiv: 2504.14817 by the authors.

Figure 1
Figure 1. Overview of the HRIRs identification with an azimuth-rotating speaker array, where the loudspeakers are arranged in a circular configuration along the elevation plane [PITH_FULL_IMAGE:figures/full_fig_p003_1.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

  1. Auditory Intelligence: Understanding the World Through Sound

    eess.AS 2025-08 conditional novelty 4.0 of 10

    A position paper proposing four cognitively inspired audio tasks (ASPIRE, SODA, AUX, AUGMENT) to push machine hearing beyond recognition toward explanation, reasoning, and interaction.

Reference graph

Works this paper leans on

53 extracted references · 50 canonical work pages · cited by 1 Pith paper

  1. [1]

    Rendering localized spatial audio in a virtual auditory space,

    D. N. Zotkin, R. Duraiswami, and L. S. Davis, “Rendering localized spatial audio in a virtual auditory space,” IEEE Trans. Multimedia, vol. 6, no. 4, pp. 553-564, Aug. 2004

    work page 2004

  2. [2]

    Auditory and visual 3D virtual reality therapy for chronic subjective tinnitus: Theoretical framework,

    A. Londero et al. , “Auditory and visual 3D virtual reality therapy for chronic subjective tinnitus: Theoretical framework,” Virtual Reality, vol. 14, no. 2, pp. 143–151, Jun. 2010

    work page 2010

  3. [3]

    Exploring 2D, 3D and spatial audio user interfaces in VR for reminiscence therapy,

    A. Uribe-Quevedo, C. Arevalo, J. Villegas, and W. Sun, “Exploring 2D, 3D and spatial audio user interfaces in VR for reminiscence therapy, ” in Proc. IEEE 12th Global Conf. Consum. Electron. (GCCE), Osaka, Japan, Oct. 2023, pp. 703–707

    work page 2023

  4. [4]

    Feasibility of virtual reality -based auditory localizatio n training with binaurally recorded auditory stimuli for patients with single-sided deafness,

    L. Shim et al., “Feasibility of virtual reality -based auditory localizatio n training with binaurally recorded auditory stimuli for patients with single-sided deafness,” Clin. Exp. Otorhinolaryngol., vol. 16, no. 3, pp. 217–224, Aug. 2023

    work page 2023

  5. [5]

    Headphone-based immersive audio for virtual reality headsets,

    S. N. Yao, “Headphone-based immersive audio for virtual reality headsets,” IEEE Trans. Consum. Electron., vol. 63, no. 3, pp. 300 –308, Aug. 2017

    work page 2017

  6. [6]

    StereoPilot: A wearable target location system for blind and visually impaired using spatial audio rendering,

    X. Hu, A. Song, Z. Wei, and H. Zeng, “StereoPilot: A wearable target location system for blind and visually impaired using spatial audio rendering,” IEEE Trans. Neural Syst. Rehabil. Eng. , vol. 30, pp. 1621– 1630, 2022

    work page 2022

  7. [7]

    Instantaneous binaural target PSD estimation for hearing aid noise reduction in complex acoustic environments,

    A. H. Kamkar -Parsi and M. Bouchard, “Instantaneous binaural target PSD estimation for hearing aid noise reduction in complex acoustic environments,” IEEE Trans. Instrum. Meas. , vol. 60, no. 4, pp. 1141 – 1154

    work page

  8. [8]

    Lee, S.-M

    G.-T. Lee, S.-M. Choi, B.-Y. Ko, and Y.-H. Park, ‘‘HRTF measurement for accurate sound localization cues,’’ 2022, arXiv:2203.03166

    arXiv 2022

Show all 53 references
  1. [9]

    Advanced binaural sound localization in 3-D for humanoid robots,

    F. Keyrouz, “Advanced binaural sound localization in 3-D for humanoid robots,” IEEE Trans. Instrum. Meas. , vol. 63, no. 9, p p. 2098 –2107, 2014

    2014

  2. [10]

    Localization using nonindividualized head-related transfer functions,

    E. M. Wenzel, M. Arruda, D. J. Kistler, and F. L. Wightman, “Localization using nonindividualized head-related transfer functions,” J. Acoust. Soc. Am., vol. 94, no. 1, pp. 111–123, 1993

    1993

  3. [11]

    Usability of individual ized head -related transfer functions in virtual reality: Empirical study with perceptual attributes in sagittal plane sound localization,

    C. Jenny and C. Reuter, “Usability of individual ized head -related transfer functions in virtual reality: Empirical study with perceptual attributes in sagittal plane sound localization,” JMIR Serious Games, vol. 8, no. 3, Jul. 2020

    2020

  4. [12]

    Simulation of the transfer function for a head-and- torso model over the entire audible frequency range,

    T. Huttunen et al., “Simulation of the transfer function for a head-and- torso model over the entire audible frequency range, ” presented at the Proc. 2007 Int. Congr. Acoustics (ICA), Madrid, Spain, Sep. 2007

    2007

  5. [13]

    Numerical modelling of the spatial acoustic response of the human pinna,

    Y. Kahana and P. A. Nelson, “Numerical modelling of the spatial acoustic response of the human pinna,” J. Sound Vib., vol. 292, no. 1–2, pp. 148–178, Apr. 2006

    2006

  6. [14]

    PRTFNet: HRTF individualization for accurate spectral cues using a compact PRTF,

    B.-Y. Ko, G. -T. Lee, H. Nam, and Y. -H. Park, “PRTFNet: HRTF individualization for accurate spectral cues using a compact PRTF, ” IEEE Access, vol. 11, pp. 96119–96130, 2023

    2023

  7. [15]

    Autoencoding HRTFs for DNN based HRTF personalization using anthropometric features,

    T. Chen, T. Kuo, and T. Chi, “Autoencoding HRTFs for DNN based HRTF personalization using anthropometric features,” in Proc. IEEE Int. Conf. Acoustics, Speech, Signal Process. (ICASSP) , Singapore, May 2019

    2019

  8. [16]

    On the influence of continuous subject rotation duri ng high -resolution head -related transfer function measurements,

    J. G. Richter and J. Fels, “On the influence of continuous subject rotation duri ng high -resolution head -related transfer function measurements,” IEEE/ACM Trans. Audio Speech Lang. Process. , vol. 27, no. 4, pp. 730–741, Apr. 2019

    2019

  9. [17]

    Multiple exponential sweep method for fast measurement of head-related transfer functions,

    P. Majdak, P. Balazs, and B. Laback, “Multiple exponential sweep method for fast measurement of head-related transfer functions, ” J. Audio Eng. Soc., vol. 55, no. 11, pp. 623–637, Nov. 2007

    2007

  10. [18]

    Fukudome, T

    K. Fukudome, T. Suetsugu, T. Ueshin, R. Idegami, and K. Takeya, “The 14 > REPLACE THIS LINE WITH YOUR MANUSCRIPT ID NUMBER (DOUBLE-CLICK HERE TO EDIT) < fast measurement of head -related impulse responses for all azimuthal directions using the co ntinuous measurement method wi...

    2007

  11. [19]

    Fast measurement of individual head -related transfer functions,

    J.-G. Richter, “Fast measurement of individual head -related transfer functions,” Ph.D dissertation, RWTH Aachen, Aachen, Germany, 2019

    2019

  12. [20]

    HRTF measurements with a continuously moving loudspeaker and swept sines,

    V. Pulk ki, “HRTF measurements with a continuously moving loudspeaker and swept sines, ” in Proc. 128th Audio Eng. Soc. Conv., 2010

    2010

  13. [21]

    3D-continuous-azimuth acquisition of head -related impulse responses using multi -channel adaptive filtering,

    G. Enzner, “3D-continuous-azimuth acquisition of head -related impulse responses using multi -channel adaptive filtering, ” in Proc. IE EE Workshop Appl. Signal Process. Audio Acoust., 2009

    2009

  14. [22]

    Simulation and analysis of measurement techniques for the fast acquisition of head -related transfer functions,

    M. Fallahi, F. Brinkmann, and S. Weinzierl, “Simulation and analysis of measurement techniques for the fast acquisition of head -related transfer functions,” in Proc. German Annu. Conf. Acoust. (DAGA) , Nuremberg, Germany, Sep. 7–12, 2015

    2015

  15. [23]

    Measurement of head-related transfer functions: A review,

    S. Li and J. Peissig, “Measurement of head-related transfer functions: A review,” Appl. Sci., vol. 10, no. 14, p. 5014, Jul. 2020

    2020

  16. [24]

    Implementing continuous HRTF measurement in near-field,

    E. L. Tan, S. Peksi, and W. S. Gan, “Implementing continuous HRTF measurement in near-field,” in Proc. IEEE Int. Conf. Acoust., Speech Signal Process. (ICASSP), 2023

    2023

  17. [25]

    Analysis and optimal control of LMS-type adaptive filtering for continuous-azimuth acquisition of head -related impulse responses,

    G. Enzner, “Analysis and optimal control of LMS-type adaptive filtering for continuous-azimuth acquisition of head -related impulse responses, ” in Proc. IEEE Int. Conf. Acoust., Speech Signal Process. (ICASSP), Mar. 2008, pp. 393–396

    2008

  18. [26]

    Towards mobile 3D HRTF measurement,

    S. Li, A. Tobbala, and J. Peissig, “Towards mobile 3D HRTF measurement,” in Proc. 148th Audio Eng. Soc. Conv., May 2020

    2020

  19. [27]

    Simulation and analysis of measurement techniques for the fast acquisition of individual head -related transfer functions ,

    M. Fallahi, “Simulation and analysis of measurement techniques for the fast acquisition of individual head -related transfer functions ,” M.S. thesis, Tech. Univ. Berlin, Berlin, Germany, 2014

    2014

  20. [28]

    A flexible framework for expectation maximization-based MIMO system identification for time -variant linear acoustic systems,

    T. Kabzinski and P. Jax, “A flexible framework for expectation maximization-based MIMO system identification for time -variant linear acoustic systems,” IEEE Open J. Signal Process. , vol. 5, pp. 112 –121, 2023

    2023

  21. [29]

    Towards faster continuous multi -channel HRTF measurements based on learning system models,

    T. Kabzinski and P. Jax, “Towards faster continuous multi -channel HRTF measurements based on learning system models, ” in Proc. IEEE ICASSP, 2022, pp. 436–440

    2022

  22. [30]

    A high -resolution head -related transfer function database including different orientations of head above the torso,

    F. Brinkmann, A. Lind au, S. Weinzierl, G. Geissler, and S. van de Par, “A high -resolution head -related transfer function database including different orientations of head above the torso, ” Tech. Univ. Berlin, Berlin, Germany, Tech. Rep., 2019

    2019

  23. [31]

    Moving in a moving world: A review on vestibular motion sickness,

    G. Bertolini and D. Straumann, “Moving in a moving world: A review on vestibular motion sickness,” Front. Neurol., Feb. 2016

    2016

  24. [32]

    The effects of whole body rotations in the pitch and yaw planes on postural stability,

    F. Carrick, G. Pagnacco, E. Oggero, and D. Barton, “The effects of whole body rotations in the pitch and yaw planes on postural stability, ” Funct. Neurol., 2011

    2011

  25. [33]

    Introduction to digital speech processing,

    L. R. Rabiner and R. W. Schafer, “Introduction to digital speech processing,” Found. Trends Signal Process., vol. 1, no. 1 –2, pp. 1–194, 2007

    2007

  26. [34]

    Sequential and direct access of head -related transfer functions (HRTFs) for quasi -continuous angular positions,

    C. Antweiler and P. Vary, “Sequential and direct access of head -related transfer functions (HRTFs) for quasi -continuous angular positions, ” in Proc. 19th Eur. Signal Process. Conf. (EUSIPCO) , Aug. 2011, pp. 1648–1652

    2011

  27. [35]

    Perfect-sweep NLMS for time-variant acoustic system identification,

    C. Antweiler, A. Telle, P. Vary, and G. Enzner, “Perfect-sweep NLMS for time-variant acoustic system identification,” in Proc. IEEE Int. Conf. Acoust., Speech Signal Process. (ICASSP), Mar. 2012, pp. 517–520

    2012

  28. [36]

    Xie, Head-Related Transfer Function and Virtual Auditory Display

    B. Xie, Head-Related Transfer Function and Virtual Auditory Display . Plantation, FL, USA: J. Ross Publ., 2013

    2013

  29. [37]

    A cross -evaluated database of measured and simulated HRTFs includ ing 3D head meshes, anthropometric features, and headphone impulse responses,

    F. Brinkmann et al ., “A cross -evaluated database of measured and simulated HRTFs includ ing 3D head meshes, anthropometric features, and headphone impulse responses,” J. Audio Eng. Soc., vol. 67, no. 9, pp. 705–718, 2019

    2019

  30. [38]

    Insights into head-related transfer function: Spatial dimen sionality and continuous representation,

    W. Zhang, T. D. Abhayapala, R. A. Kennedy, and R. Duraiswami, “Insights into head-related transfer function: Spatial dimen sionality and continuous representation, ” J. Acoust. Soc. Am. , vol. 127, no. 4, pp. 2347–2357, Apr. 2010

    2010

  31. [39]

    Efficient real spherical harmonic representation of head-related transfer functions,

    G. D. Romigh, D. S. Brungart, R. M. Stern, and B. D. Simpson, “Efficient real spherical harmonic representation of head-related transfer functions,” IEEE J. Sel. Topics Signal Process. , vol. 9, no. 5, pp. 921 – 930, Aug. 2015

    2015

  32. [40]

    Posture detection using sounds and temperature: LMS -based approach to enable sensory substitution,

    L. Russell, R. Goubran, and F. Kwamena, “Posture detection using sounds and temperature: LMS -based approach to enable sensory substitution,” IEEE Trans. Instrum. Meas., vol. 67, no. 7, pp. 1543–1554, Jul. 2018

    2018

  33. [41]

    Adaptive filtering improved apnea detection performance using tracheal sounds in noisy environment: A simulation study,

    Y. Wu et al., “Adaptive filtering improved apnea detection performance using tracheal sounds in noisy environment: A simulation study, ” Biomed. Res. Int., vol. 2020, 2020

    2020

  34. [42]

    On the identifica tion of electrical equivalent circuit models based on noisy measurements,

    B. Balasingam and K. R. Pattipati, “On the identifica tion of electrical equivalent circuit models based on noisy measurements, ” IEEE Trans. Instrum. Meas., vol. 70, 2021

    2021

  35. [43]

    Analysis and comparison of different adaptive filtering algorithms for fast continuous HRTF measurement,

    C. K. Correa, S. Li, and J. Peissig, “Analysis and comparison of different adaptive filtering algorithms for fast continuous HRTF measurement,” in Proc. Fortschr. Akust. (DAGA) , Kiel, Germany, Mar. 6–9, 2017

    2017

  36. [44]

    Meta-AF: Meta-learning for adaptive filters,

    J. Casebeer, N. J. Bryan, and P. Smaragdis, “Meta-AF: Meta-learning for adaptive filters, ” IEEE/ACM Trans. Audio Speech Lang. Process. , vol. 31, pp. 355–370, 2023

    2023

  37. [45]

    An optimized NLMS algorithm for system identification,

    S. Ciochină, C. Paleo logu, and J. Benesty, “An optimized NLMS algorithm for system identification,” Signal Process., vol. 118, pp. 115– 121, Jan. 2016

    2016

  38. [46]

    An overview of gradient descent optimization algorithms,

    S. Ruder, “An overview of gradient descent optimization algorithms, ” 2016, arXiv:1609.04747

    2016 arXiv

  39. [47]

    Adam: A method for stochastic optimization,

    D. P. Kingma and J. Ba, “Adam: A method for stochastic optimization,” 2014, arXiv:1412.6980

    2014 arXiv

  40. [48]

    Audibility and interpolation of head -above-torso orientation in binaural technology,

    F. Brinkmann, R. Roden, A. Lindau, and S. Weinzierl, “Audibility and interpolation of head -above-torso orientation in binaural technology, ” IEEE J. Sel. Topics Signal Process. , vol. 9, no. 5, pp. 931–942, Aug. 2015

    2015

  41. [49]

    A nonparametric VSS NLMS algorithm,

    J. Benesty, H. Rey, L. R. Vega, and S. Tressens, “A nonparametric VSS NLMS algorithm,” IEEE Signal Process. Lett., vol. 13, no. 10, pp. 581 – 584, Oct. 2006

    2006

  42. [50]

    Spectral cues used in the localization of sound sources on the median plane,

    J. Hebrank and D. Wright, “Spectral cues used in the localization of sound sources on the median plane, ” J. Acoust. Soc. Am., vol. 56, no. 6, pp. 1829–1834, 1974

    1974

  43. [51]

    XII, on our perception of sound direction,

    L. Rayleigh, “XII, on our perception of sound direction, ” London, Edinburgh, Dublin Philos. Mag. J. Sci., vol. 13, no. 74, pp. 214 –232, 1907

    1907

  44. [52]

    Head-related transfer function interpolation in azimuth, elevation, and distance,

    H. Gamper, “Head-related transfer function interpolation in azimuth, elevation, and distance, ” J. Acoust. Soc. Am. , vol. 134, no. 6, pp. EL547–EL553, Dec. 2013

    2013

  45. [53]

    Median plane localization using a parametric model of the head -related t ransfer function based on spectral cues,

    K. Iida, M. Itoh, A. Itagaki, and M. Morimoto, “Median plane localization using a parametric model of the head -related t ransfer function based on spectral cues, ” Appl. Acoust., vol. 68, no. 8, pp. 835 – 850, Aug. 2007. BYEONG-YUN KO received the B.S. degree in ship architec...

    2007

Pith tools

Reviewed August 16, 2026 · model on record in the stance chip above.