HARP: A Large-Scale Higher-Order Ambisonic Room Impulse Response Dataset

J\"urgen Peissig; Shivam Saini

REVIEW 4 major objections 4 minor 1 cited by

HARP: A Large-Scale Higher-Order Ambisonic Room Impulse Response Dataset

T0 review · 4 major / 4 minor · reviewed 2026-08-12 · deepseek-v4-flash

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

Pith's one-line read This paper introduces HARP, a dataset of 100,000 simulated 7th-order Ambisonic room impulse responses, claimed to be the first large-scale HOA-RIR dataset.

desk verdict A large-scale 7th-order HOA RIR dataset would be a real contribution, but this version doesn't yet show that the 64-microphone encoding works, so it's a promising announcement rather than a usable dataset. read the letter →

arxiv 2411.14207 v2 pith:VHV2WNAU submitted 2024-11-21 cs.SD cs.AIcs.LGcs.MMeess.AS

Shivam Saini , Jürgen Peissig This is my paper · ORCID

classification cs.SDcs.AIcs.LGcs.MMeess.AS

keywords Higher-OrderAmbisonicsRoomImpulseResponsedatasetImageSourceMethodSphericalHarmonicsSpatialaudioMachinelearningacousticsReverberationHOA-RIR

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

This paper introduces HARP, a simulated dataset of 100,000 room impulse responses encoded as 7th-order Ambisonics (HOA-RIRs). It is offered as the first large-scale resource of its kind, filling a gap left by existing datasets that stop at lower orders or cover only a few rooms. The responses are generated by the Image Source Method across a variety of room geometries and materials, using a 64-microphone array whose directivities are the real spherical harmonics, so the array captures HOA coefficients directly. If the encoding is correct, HARP supplies machine-learning researchers with the scale and spatial resolution needed for tasks such as room parameter estimation, dereverberation, and immersive rendering.

What carries the argument

The load-bearing object is the 64-microphone superposition array. Each capsule's directivity is one real spherical harmonic $Y_{n,m}(\theta,\phi)$ derived from the complex spherical harmonics $Y_n^m(\theta,\phi)$ with N3D normalization; the signal at position $\mathbf{r}_m$ is written as $$p(\mathbf{r}_m,t)=\sum_{n=0}^{7}\sum_{m=-n}^{n}c_n^m(t)Y_n^m(\theta_m,\phi_m),$$ so the weights $c_n^m(t)$ are the HOA coefficients up to order 7. Because a 7th-order field has $(7+1)^2=64$ coefficients, 64 suitably arranged capsules match the degrees of freedom. These directivities are inserted into an image-source room simulator that computes, for each room configuration, impulse responses to all 64 microphones; the 0th-order response doubles as the omnidirectional channel for RT60 analysis.

What would settle it

Re-simulate one room configuration from HARP in an independent image-source or boundary-element solver, place the same source and receiver, sample the sound field with the described 64-microphone array, and compare the resulting 7th-order coefficients against a reference computed by analytic projection of the same field onto spherical harmonics; if the error exceeds the sampling discretization error of a 7th-order grid, the dataset's core encoding claim fails.

Watch

Extended reading notes

Core claim

The paper's central claim is that a 64-microphone superposition arrangement, each microphone having a directivity given by one real spherical harmonic up to order 7, can sample a simulated room sound field directly in the spherical-harmonics domain, and that applying this configuration inside an image-source room simulator yields 100,000 valid 7th-order Ambisonic RIRs with varied room geometries, materials, and source-receiver positions. It reports that these are, to its knowledge, the first large-scale HOA-RIRs, stored in AmbiX (ACN/SN3D) format with metadata, and it argues that this scale and order is what distinguishes HARP from earlier datasets with only a few hundred samples or lower-order encoding.

Load-bearing premise

The whole dataset stands on the unstated and unvalidated assumption that the 64-microphone superposition array recovers the true 7th-order spherical-harmonic coefficients of each simulated sound field.

Editorial extensions

If this is right

HARP gives machine-learning researchers 100,000 labelled HOA-RIRs to train or pre-train models for room parameter estimation, dereverberation, source localization, and spatial upsampling.
The AmbiX/SN3D format lets the responses feed directly into standard Ambisonic renderers without conversion.
The RT60 distribution concentrated between 0.4 and 0.8 seconds means most samples match typical living rooms and offices, with edge cases included for more extreme acoustics.
The inclusion of multiple room geometries and material lookup tables supports experiments on generalization across acoustic environments.
If the encoding works as claimed, this is the first large-scale HOA-RIR dataset, providing a reference scale for other synthetic and measured spatial-audio datasets.

Reading between the lines

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

If the array is validated, the same construction can be carried to even higher orders or to other simulation engines, since the count 64 only reflects the $(N+1)^2$ harmonic count for $N=7$.
A natural test the paper does not report is comparing the dataset's 0th-order RIRs against measured RIRs in similar rooms; a systematic mismatch would quantify how much of the real-world gap comes from the image-source approximation.
The lack of diffraction and scattering in the image-source method suggests models trained on HARP may need fine-tuning on measured data before use in rooms with furniture or other obstacles.
The claim of being the first large-scale HOA-RIR dataset rests on the comparison table; if a comparable 7th-order dataset appears, the contribution would shift from firstness to scale and methodology.

Share X Bluesky LinkedIn Reddit HN

Editorial analysis

A structured set of objections, weighed in public.

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

Desk Editor's Note

read the letter

The headline is simple: if HARP is valid, it's the first large-scale HOA-RIR dataset at 7th order, and that matters for the ML side of spatial audio. Existing datasets top out at 4th order and a few hundred samples (TAU-SRIR, MOTUS), so 100,000 simulated 7th-order RIRs is a genuine step up in scale. The idea of using spherical-harmonic directivity patterns for the microphones inside Pyroomacoustics is also reasonable, and extending that library is a useful, reproducible contribution if the code is released.

But the paper doesn't yet support its central claim. The 64-microphone configuration that supposedly captures HOA coefficients directly is never fully specified. I don't need psychoanalysis here, just the math: the paper gives equations for spherical harmonics and says the signal is a weighted sum, but it never states whether the microphones are coincident with different SH patterns or spread out, and it doesn't give coordinates, weights, or an inversion. That's a load-bearing gap. If the microphones are simply pressure sensors at different locations, you need a 64x64 linear system to get coefficients. If they are directional SH patterns, you need to explain how the integral in the directivity maps to the coefficient in Eq. (3). Neither is done.

There's also a normalization inconsistency: Eq. (1) uses N3D, but the stored format is ACN/SN3D, and no conversion factor is provided. That could silently scale every coefficient. The only validation, the RT60 histogram from the 0th-order channel, can't catch errors in orders 1–7. And there is no dataset link, no free-field benchmark, no comparison against analytic or measured fields.

The paper's own conclusion is honest about known limitations (ISM misses diffraction etc.), which is good. But the missing validation is not a minor omission; it's the difference between a dataset paper and a dataset announcement.

For whom is this useful? Researchers who want a large synthetic HOA training resource. After a revision that spells out the microphone setup, validates the encoding (even just a plane-wave test), fixes the normalization, and releases the data, this could become a standard resource. Right now I wouldn't cite it or train on it.

My recommendation: send it to peer review, because the artifact could be important and the flaws are fixable, but the referee should demand the missing details and a validation section before acceptance.

Referee Report

4 major / 4 minor

Summary. The paper introduces HARP, a proposed dataset of 100,000 simulated 7th-order Higher-Order Ambisonic Room Impulse Responses (HOA-RIRs) generated with the Image Source Method using the Pyroomacoustics library. The authors describe a 64-microphone configuration intended to capture RIRs directly in the spherical harmonics domain, a room-simulation setup with randomized room geometries, absorption materials from a lookup table, and source-receiver positions, and they compare HARP with existing ambisonic RIR datasets. They also note a planned contribution to Pyroomacoustics through a new SphericalHarmonicDirectivity class.

Significance. If the generation method is correct and the dataset is released, HARP would be a valuable large-scale resource for spatial audio and machine-learning research, offering far more 7th-order HOA-RIRs than existing datasets. The paper also claims a first large-scale HOA-RIR dataset and a new directivity implementation in a widely used simulation library. However, the central technical claim—that the 64-microphone configuration directly captures 7th-order HOA coefficients—is not derived, specified, or validated, and the paper currently provides no dataset link. The practical significance therefore depends entirely on unverified assertions.

major comments (4)

[Section III-A, Eqs. (1)-(4)] The manuscript never specifies whether the 64 microphones are coincident directional sensors or spatially distributed transducers, nor does it give their positions, polar patterns, or the combination weights that map the 64 outputs to the HOA coefficients c_n^m(t). Eq. (3) is the spherical harmonic expansion of the pressure at a point, but the output of a directional microphone is a weighted integral of the incident field against the directivity pattern, not the point pressure. If the microphones are spatially separated, a 64x64 linear system or a spherical sampling scheme such as a t-design is required, and none is described. The central claim that the configuration 'captures RIRs directly in the Spherical Harmonics domain' is therefore unsupported as written.
[Section III-A vs. Section III-D] Section III-A states that N3D normalization is used for the spherical harmonic coefficients, while Section III-D states that each RIR is stored in the AmbiX format (ACN/SN3D). N3D and SN3D are different normalizations, and no conversion factors between them are provided. Without this conversion, the stored coefficients may be scaled incorrectly relative to the simulated microphone outputs, corrupting the dataset for downstream Ambisonics rendering and analysis.
[Section IV, Figure 4] The only quantitative validation presented is the RT60 distribution computed from the 0th-order (omnidirectional) channel. This cannot detect errors in the higher-order channels (orders 1-7), which are the entire point of the dataset. Moreover, the caption states 'Histogram to be updated in camera-ready version,' indicating that the figure is a placeholder. There is no analytic free-field benchmark, no comparison to measured HOA-RIRs, no comparison to an independently computed sound field, and no error metric for the spherical harmonic coefficients. The dataset's central premise is therefore unverified.
[Dataset availability] The paper claims to introduce a dataset and to provide it as a resource, but it gives no URL, repository, or access instructions. For a dataset paper, the availability of the data and metadata is a load-bearing component; without it, the contribution cannot be evaluated or used by the community. A revision should include a working link and a data availability statement.

minor comments (4)

[Throughout] The text contains several typographical errors, such as 'FOr instance,' 'capturedd,' and 'f ¨ur'; these should be corrected in a revision.
[Section III-B] The phrase 'a very image source order (40)' should read 'a very high image source order (40)' or similar.
[Figure 2] The 'Free field measurement' shown in Figure 2 is not described quantitatively in the text; the authors should specify what is plotted and how it verifies the directivity of the proposed configuration.
[Section IV, Table I] The comparison table would be more useful if it included the number of RIRs for ARNI SRIR (currently listed as '-') and the exact room configurations for HARP, rather than the qualitative 'Wide variety.'

Circularity Check

0 steps flagged · score 1.0 of 10

No significant circularity: the HOA coefficients are generated by directivity-defined sensors rather than fitted or predicted, though the method is under-specified and unvalidated.

full rationale

Walking the derivation chain: Section III-A defines 64 microphone directivities from the spherical-harmonic basis of order <= 7 (Eqs. 1 and 2). Because 64 = (7+1)^2, the channels correspond one-to-one to the HOA coefficients c_n^m(t); a simulated sensor whose directivity is Y_n^m outputs the projection of the incident field onto that basis function. That is the standard definition of ideal HOA encoding, not a fitted parameter renamed as a prediction. Eq. (3) is only the SH expansion statement, and no parameter is optimized against a target or held-out quantity. The sole quantitative check, the RT60 distribution in Figure 4, uses the 0th-order channel and therefore validates the room simulation's reverberation characteristics but not the higher-order coefficients; this is a validation gap, not circularity. The one self-citation [2] appears in a list of downstream applications and is not load-bearing. The paper is under-specified (no microphone coordinates or combination weights, no N3D-to-SN3D conversion factor) and unvalidated against external benchmarks, but no load-bearing step reduces to its own input by construction. Hence no significant circularity is found.

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

The paper's central contribution rests on a custom microphone configuration and a modified simulator, neither of which is released. The only numbers the claim depends on are simulation hyperparameters, not fitted data.

free parameters (3)

maximum image source order = 40
Hand-chosen truncation of the image source expansion; no convergence analysis is provided, yet RT60 and RIR accuracy depend on it.
number of source-receiver combinations per room = 20
Chosen for dataset size; no rationale or variability analysis is given.
lookup table absorption coefficients = material-dependent values from standard references
Assigned from a material table; the table itself is not included in the paper.

assumptions (4)

domain assumption Linear acoustics and superposition hold for the simulated rooms.
The 64-microphone direct SH encoding relies on linear superposition of image sources; no nonlinear effects are modeled.
domain assumption The image source method with 40th-order reflections produces RIRs representative of real rooms.
Authors acknowledge ISM does not capture diffraction and scattering; real-world validity is assumed.
ad hoc to paper A set of 64 spherical harmonic directional microphones can directly measure HOA coefficients up to 7th order.
This is the central methodological assumption in Section III-A; no proof, coordinates, or validation is given.
domain assumption N3D normalized real spherical harmonics are orthonormal and correctly implemented in the modified Pyroomacoustics directivity class.
Eqs. (1) to (4) assume orthonormality; no tests of the custom implementation are shown.

how reviews work

0 comments

Cite this review

Pith. "Pith review of HARP: A Large-Scale Higher-Order Ambisonic Room Impulse Response Dataset." pith.science (2026). https://pith.science/paper/VHV2WNAU

@misc{pith2026241114207,
  author       = {Pith},
  title        = {Pith review of: HARP: A Large-Scale Higher-Order Ambisonic Room Impulse Response Dataset},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/VHV2WNAU}},
  note         = {Machine review of arXiv:2411.14207}
}

read the original abstract

This contribution introduces a dataset of 7th-order Ambisonic Room Impulse Responses (HOA-RIRs), created using the Image Source Method. By employing higher-order Ambisonics, our dataset enables precise spatial audio reproduction, a critical requirement for realistic immersive audio applications. Leveraging the virtual simulation, we present a unique microphone configuration, based on the superposition principle, designed to optimize sound field coverage while addressing the limitations of traditional microphone arrays. The presented 64-microphone configuration allows us to capture RIRs directly in the Spherical Harmonics domain. The dataset features a wide range of room configurations, encompassing variations in room geometry, acoustic absorption materials, and source-receiver distances. A detailed description of the simulation setup is provided alongside for an accurate reproduction. The dataset serves as a vital resource for researchers working on spatial audio, particularly in applications involving machine learning to improve room acoustics modeling and sound field synthesis. It further provides a very high level of spatial resolution and realism crucial for tasks such as source localization, reverberation prediction, and immersive sound reproduction.

Figures

Figures reproduced from arXiv: 2411.14207 by the authors.

**Figure 1.** Spherical harmonics up to 4th order A. Microphone Configuration To capture the sound field with a high spatial resolution, we propose a 64-channel 7th-order HOA framework. Traditional grid-based microphone arrays often suffer from spatial aliasing as mentioned earlier. We mitigate this by placing the microphones using a superposition arrangement, which ensures even sampling of the sound field. The simulation strate… view at source ↗

**Figure 2.** Free field point source a meter away from the microphone array. [PITH_FULL_IMAGE:figures/full_fig_p002_2.png] view at source ↗

**Figure 3.** Example room configuration with 4 point sources and the proposed [PITH_FULL_IMAGE:figures/full_fig_p003_3.png] view at source ↗

Figures from the paper (2 more)

**Figure 4.** Figure 4: RT60 Distribution calculated with 0 th-order microphone. Histogram to be updated in camera-ready version. Table I highlights key differences between HARP and other publicly available Spatial RIR datasets. HARP distinguishes itself by offering 7th-order HOA-RIRs, makin…

**Figure 5.** Figure 5: Distribution of the dataset in relation to RT60 [s], Room Volume [PITH_FULL_IMAGE:figures/full_fig_p004_5.png]

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

Physics-Informed Direction-Aware Neural Acoustic Fields
cs.SD 2025-07 conditional novelty 5.0 of 10

A physics-informed neural network enforces momentum and continuity equations on all four Ambisonic channels, outperforming data-only and W-channel-only baselines for room impulse response interpolation.

Reference graph

Works this paper leans on

24 extracted references · 19 canonical work pages · cited by 1 Pith paper

[1]

3d sound field recording with higher order ambisonics–objective measurements and validation of a 4th order spherical microphone,

S. Moreau, J. Daniel, and S. Bertet, “3d sound field recording with higher order ambisonics–objective measurements and validation of a 4th order spherical microphone,” in 120th Convention of the AES, pp. 20–23, 2006

work page 2006
[2]

Blind room acoustic parameters estimation using mobile audio transformer,

S. Saini and J. Peissig, “Blind room acoustic parameters estimation using mobile audio transformer,” in 2023 IEEE Workshop on Applications of Signal Processing to Audio and Acoustics (WASPAA) , 2023

work page 2023
[3]

Impulse response data augmentation and deep neural networks for blind room acoustic parameter estimation,

N. J. Bryan, “Impulse response data augmentation and deep neural networks for blind room acoustic parameter estimation,” in 2020 IEEE International Conference on Acoustics, Speech and Signal Processing (ICASSP), 2020

work page 2020
[4]

Online Blind Reverberation Time Estimation Using CRNNs,

S. Deng, W. Mack, and E. A. Habets, “Online Blind Reverberation Time Estimation Using CRNNs,” in Proc. Interspeech 2020 , pp. 5061–5065, 2020

work page 2020
[5]

Online reverberation time and clarity estimation in dynamic acoustic conditions,

P. G ¨otz, C. Tuna, A. Walther, and E. A. P. Habets, “Online reverberation time and clarity estimation in dynamic acoustic conditions,” The Journal of the Acoustical Society of America , vol. 153, pp. 3532–3542, 06 2023

work page 2023
[6]

Single-and multi-microphone speech dereverberation using spectral enhancement,

E. A. P. Habets, “Single-and multi-microphone speech dereverberation using spectral enhancement,” 2007

work page 2007
[7]

Metamgc: a music generation framework for concerts in metaverse,

C. Jin, F. Wu, J. Wang, Y . Liu, Z. Guan, and Z. Han, “Metamgc: a music generation framework for concerts in metaverse,” EURASIP journal on audio, speech, and music processing , vol. 2022, no. 1, p. 31, 2022. Submitted to ICASSP 2025 Workshop

work page 2022
[8]

A parametric spatial audio compression codec for higher-order ambisonics,

C. Hold, “A parametric spatial audio compression codec for higher-order ambisonics,” 2024

work page 2024

Show all 24 references

[9]

Parametric late rever- beration from broadband directional estimates,

N. Meyer-Kahlen, S. J. Schlecht, and T. Lokki, “Parametric late rever- beration from broadband directional estimates,” in 2021 Immersive and 3D Audio: from Architecture to Automotive (I3DA) , pp. 1–10, 2021

2021
[10]

Upmix b-format ambisonic room impulse responses using a generative model,

J. Xia and W. Zhang, “Upmix b-format ambisonic room impulse responses using a generative model,” Applied Sciences, vol. 13, no. 21, p. 11810, 2023

2023
[11]

Overview and evaluation of sound event localization and detection in dcase 2019,

A. Politis, A. Mesaros, S. Adavanne, T. Heittola, and T. Virtanen, “Overview and evaluation of sound event localization and detection in dcase 2019,” IEEE/ACM Transactions on Audio, Speech, and Language Processing, vol. 29, pp. 684–698, 2020

2019
[12]

Pyroomacoustics: A python package for audio room simulation and array processing algorithms,

R. Scheibler, E. Bezzam, and I. Dokmani ´c, “Pyroomacoustics: A python package for audio room simulation and array processing algorithms,” in 2018 IEEE international conference on acoustics, speech and signal processing (ICASSP), pp. 351–355, IEEE, 2018

2018
[13]

Zotter and M

F. Zotter and M. Frank, Ambisonics: A practical 3D audio theory for recording, studio production, sound reinforcement, and virtual reality . Springer Nature, 2019

2019
[14]

How to make ambisonics sound good,

M. Frank, “How to make ambisonics sound good,” in Forum Acus- ticum,(Krakow), 2014

2014
[15]

Investigation on localisation accuracy for first and higher order ambisonics reproduced sound sources,

S. Bertet, J. Daniel, E. Parizet, and O. Warusfel, “Investigation on localisation accuracy for first and higher order ambisonics reproduced sound sources,” Acta Acustica united with Acustica , vol. 99, no. 4, pp. 642–657, 2013

2013
[16]

Ambisonics encoding of other audio formats for multiple listening conditions,

J. Daniel, J.-B. Rault, and J.-D. Polack, “Ambisonics encoding of other audio formats for multiple listening conditions,” in Audio Engineering Society Convention 105 , Audio Engineering Society, 1998

1998
[17]

A comparative study of 3-d audio encoding and rendering techniques,

J.-M. Jot, V . Larcher, and J.-M. Pernaux, “A comparative study of 3-d audio encoding and rendering techniques,” in Audio engineering society conference: 16th international conference: Spatial sound reproduction , Audio Engineering Society, 1999

1999
[18]

Image method for efficiently simulating small-room acoustics,

J. B. Allen and D. A. Berkley, “Image method for efficiently simulating small-room acoustics,” The Journal of the Acoustical Society of America, vol. 65, no. 4, pp. 943–950, 1979

1979
[19]

Database of omnidirectional and b-format room impulse responses,

R. Stewart and M. Sandler, “Database of omnidirectional and b-format room impulse responses,” in 2010 IEEE International Conference on Acoustics, Speech and Signal Processing , pp. 165–168, 2010

2010
[20]

A dataset of reverberant spatial sound scenes with moving sources for sound event localization and detection,

A. Politis, S. Adavanne, and T. Virtanen, “A dataset of reverberant spatial sound scenes with moving sources for sound event localization and detection,” in Detection and Classification of Acoustic Scenes and Events Workshop, 2020

2020
[21]

A dataset of higher-order ambisonic room impulse responses and 3d models measured in a room with varying furniture,

G. G ¨otz, S. J. Schlecht, and V . Pulkki, “A dataset of higher-order ambisonic room impulse responses and 3d models measured in a room with varying furniture,” in 2021 Immersive and 3D Audio: from Architecture to Automotive (I3DA) , pp. 1–8, IEEE, 2021

2021
[22]

Openair: An interactive auralization web resource and database,

D. T. Murphy and S. Shelley, “Openair: An interactive auralization web resource and database,” in Audio Engineering Society Convention 129 , Audio Engineering Society, 2010

2010
[23]

Dataset of spatial room impulse responses in a variable acoustics room for six degrees-of- freedom rendering and analysis,

T. McKenzie, L. McCormack, and C. Hold, “Dataset of spatial room impulse responses in a variable acoustics room for six degrees-of- freedom rendering and analysis,” 2021

2021
[24]

Homula-rir: A room impulse response dataset for teleconferencing and spatial audio applications acquired through higher- order microphones and uniform linear microphone arrays,

F. Miotello, P. Ostan, M. Pezzoli, L. Comanducci, A. Bernardini, F. An- tonacci, and A. Sarti, “Homula-rir: A room impulse response dataset for teleconferencing and spatial audio applications acquired through higher- order microphones and uniform linear microphone arrays,” in ...

2024

Pith tools

Reviewed August 12, 2026 · model on record in the stance chip above.

[1] [1]

3d sound field recording with higher order ambisonics–objective measurements and validation of a 4th order spherical microphone,

S. Moreau, J. Daniel, and S. Bertet, “3d sound field recording with higher order ambisonics–objective measurements and validation of a 4th order spherical microphone,” in 120th Convention of the AES, pp. 20–23, 2006

work page 2006

[2] [2]

Blind room acoustic parameters estimation using mobile audio transformer,

S. Saini and J. Peissig, “Blind room acoustic parameters estimation using mobile audio transformer,” in 2023 IEEE Workshop on Applications of Signal Processing to Audio and Acoustics (WASPAA) , 2023

work page 2023

[3] [3]

Impulse response data augmentation and deep neural networks for blind room acoustic parameter estimation,

N. J. Bryan, “Impulse response data augmentation and deep neural networks for blind room acoustic parameter estimation,” in 2020 IEEE International Conference on Acoustics, Speech and Signal Processing (ICASSP), 2020

work page 2020

[4] [4]

Online Blind Reverberation Time Estimation Using CRNNs,

S. Deng, W. Mack, and E. A. Habets, “Online Blind Reverberation Time Estimation Using CRNNs,” in Proc. Interspeech 2020 , pp. 5061–5065, 2020

work page 2020

[5] [5]

Online reverberation time and clarity estimation in dynamic acoustic conditions,

P. G ¨otz, C. Tuna, A. Walther, and E. A. P. Habets, “Online reverberation time and clarity estimation in dynamic acoustic conditions,” The Journal of the Acoustical Society of America , vol. 153, pp. 3532–3542, 06 2023

work page 2023

[6] [6]

Single-and multi-microphone speech dereverberation using spectral enhancement,

E. A. P. Habets, “Single-and multi-microphone speech dereverberation using spectral enhancement,” 2007

work page 2007

[7] [7]

Metamgc: a music generation framework for concerts in metaverse,

C. Jin, F. Wu, J. Wang, Y . Liu, Z. Guan, and Z. Han, “Metamgc: a music generation framework for concerts in metaverse,” EURASIP journal on audio, speech, and music processing , vol. 2022, no. 1, p. 31, 2022. Submitted to ICASSP 2025 Workshop

work page 2022

[8] [8]

A parametric spatial audio compression codec for higher-order ambisonics,

C. Hold, “A parametric spatial audio compression codec for higher-order ambisonics,” 2024

work page 2024