Pith. sign in

REVIEW 4 major objections 6 minor 68 references

Compact Binary Coalescence Gravitational Wave Signals Counting and Separation

T0 review · 4 major / 6 minor · reviewed 2026-08-11 · deepseek-v4-flash

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

Pith's one-line read UnMixFormer achieves 99.89% counting accuracy and 0.9831 mean overlap in separating up to five overlapping compact binary coalescence signals.

desk verdict Worth reading as a proof-of-concept, but the headline numbers are in-sample and the absence of baselines means the real advance is still unquantified. read the letter →

arxiv 2412.18259 v2 pith:5VANXAP4 submitted 2024-12-24 gr-qc astro-ph.HEastro-ph.IM

Tianyu Zhao , Yue Zhou , Ruijun Shi , Peng Xu , Zhoujian Cao , Zhixiang Ren This is my paper · ORCID
classification gr-qcastro-ph.HEastro-ph.IM
keywords gravitationalwavescompactbinarycoalescenceoverlappingsignalssignalseparationsourcecountingdeeplearningtransformerCosmicExplorer
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 introduces UnMixFormer, a neural network designed for next-generation gravitational-wave detectors where overlapping signals from multiple compact binary mergers will be common. The central claim is that one architecture can both count how many signals are present (between two and five) and separate the individual waveforms, reaching 99.89% counting accuracy and 0.9831 mean overlap on synthetic data embedded in Cosmic Explorer Gaussian noise. The authors argue this moves beyond existing methods that handle only one or two concurrent signals and limited waveform families. A sympathetic reader would care because overlapping signals will bias parameter estimation if treated as single events, so a working count-and-separate tool would directly improve astrophysical inference in the third-generation era.

What carries the argument

The central mechanism is a dual-path transformer with intra- and inter-segment attention, augmented by Fourier Analysis Networks (FAN) in place of MLP feed-forward layers, paired with a multi-decoder selector. The counting head outputs a probability over signal multiplicity and activates the matching decoder; separation is trained with a permutation-invariant SI-SNR loss plus cross-entropy on the count, so the same model both estimates the number of sources and reconstructs each coherent waveform from the whitened one-second, 16,384-sample input.

What would settle it

Run the trained model on a segment of real or simulated non-Gaussian detector noise containing glitches and non-stationary transients with injected overlapping signals, and check whether counting accuracy and mean overlap remain close to 99.89% and 0.9831; a substantial drop would show the reported numbers depend on the Gaussian-noise assumption.

Watch

Extended reading notes

Core claim

The discovery is that jointly optimizing a counting head and a bank of decoders within a dual-path attention architecture yields accurate source counting and waveform separation for up to five overlapping compact binary coalescence signals, across BBH, BNS, and NS-BH systems, at SNRs between 10 and 50. On held-out synthetic data, the model achieves 99.89% counting accuracy (with 100% accuracy within ±1 signal), AUC values above 0.9999 for counting each multiplicity, and a mean overlap of 0.9831 between reconstructed and template waveforms; mismatch degrades gracefully as the number of signals increases. The model also separates inspiral-only segments and generalizes to waveforms with spin precession, orbital eccentricity, and higher-order modes, which were not present in its training set.

Load-bearing premise

The load-bearing premise is that Gaussian noise generated from the Cosmic Explorer 40 km power spectral density accurately represents real next-generation detector noise, including its non-stationarity and glitches, so that the reported counting and separation accuracy would transfer to actual observations.

Editorial extensions

If this is right

  • Third-generation detectors could use a single trained model to count and separate overlapping BBH, BNS, and NS-BH signals in about 1.5 ms per sample, making real-time analysis feasible.
  • Because the model separates waveforms rather than fitting one template to the mixture, it could reduce the parameter-estimation bias that overlapping events introduce when treated as single signals.
  • The approach extends to inspiral-only long-duration data, relevant to the longer signals expected in next-generation detectors.
  • Generalization to precessing, eccentric, and higher-mode waveforms suggests the representation learned on simpler aligned-spin templates captures enough structure to handle more complex physics without retraining.

Reading between the lines

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

  • Editors' inference: if the Gaussian-noise assumption holds, the same architecture could be adapted to multi-detector networks by treating each detector as a channel, which the authors note but do not implement; spatial diversity would likely improve localization as well as separation.
  • Editors' inference: the counting head is trained only for 2–5 signals, so on real data a single loud event would need to be handled separately or the model retrained with a '0 or 1' class; the denoising example hints at capability but does not test counting at multiplicity 1.
  • Editors' inference: a direct stress test would be to run the model on LIGO-Virgo O4 data with glitches and non-stationary noise; the resulting drop in counting accuracy would quantify how much of the reported performance depends on ideal Gaussian noise.
Share X Bluesky LinkedIn Reddit HN

Signed reviews

No signed human review yet.

Editorial analysis

A structured set of objections, weighed in public.

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

Referee Report

4 major / 6 minor

Summary. The paper presents UnMixFormer, an attention-based neural network that combines a counting head with multiple waveform decoders to count and separate 2-5 overlapping compact binary coalescence signals in simulated Cosmic Explorer (CE-40km) noise. The authors report 99.89% counting accuracy and a mean overlap of 0.9831 between separated and target waveforms on a held-out synthetic test set with SNR 10-50. They also show qualitative examples of separation for five overlapping BBH/BNS/NS-BH signals, including cases with spin precession, orbital eccentricity, and higher-order modes, and a single-signal denoising example.

Significance. The problem addressed here is important: next-generation detectors will see overlapping CBC signals, and current matched-filtering and deep-learning methods are largely limited to one or two concurrent sources. The architecture is coherent and the use of a permutation-invariant SI-SNR objective for multi-source separation is well matched to the task. The paper provides a held-out synthetic test set of 20,000 samples, confusion matrices, ROC curves, and per-SNR mismatch analysis, which are useful. However, the evaluation is entirely within a single synthetic Gaussian-noise setting, no baseline comparison is made, and the headline overlap number is reported in an ambiguous way. If the results are confirmed under more realistic noise and compared with existing methods, this would be a valuable contribution to GW data analysis.

major comments (4)
  1. [Abstract and Section III.B] The reported mean overlap of 0.9831 is the average of the five per-signal means in the five-signal case (Fig. 4d: 0.9965, 0.9940, 0.9896, 0.9783, 0.9573), not an aggregate over all 2-5 signal test samples. Weighting the per-signal means in Fig. 4(a)-(d) by the number of signals gives an overall mean of approximately 0.989, so the abstract and introduction should either report the correct aggregate or explicitly state that 0.9831 refers only to the five-signal subset.
  2. [Section II.A and Sections III.A-III.C] All experiments use Gaussian noise generated from the CE-40km PSD, and the same PSD is used for whitening and for the overlap metric through the inner product in Eq. (3). The evaluation is therefore carried out under exactly the noise statistics assumed by the matched-filter overlap. The paper provides no test with non-stationary noise, glitches, spectral lines, or a mismodeled PSD, so the conclusion that this is a working method for next-generation detectors is not supported by the experiments. Please add out-of-distribution noise tests and report counting accuracy and overlap under those conditions.
  3. [Section III.C, Figs. 5, 8, 9] The generalization experiments are not sufficiently specified. Fig. 5 is a single qualitative example; Fig. 9 reports boxplots of mismatch versus precession and eccentricity but does not state the number of test samples, the number of noise realizations, or whether these waveforms were included in training. The single-signal denoising test in Fig. 8 is also ambiguous because the training data only contains 2-5 signals and the counting head has never been trained on one-signal examples; it is not explained how decoder selection works in that test. Similarly, the inspiral-only result in Fig. 6 refers to a retrained model whose dataset composition is not described. Please specify the exact evaluation protocols, sample sizes, and whether the same trained weights are used.
  4. [Section III (overall)] The paper claims a substantial advance over existing methods that 'can typically handle only one or two concurrent signals,' but no quantitative baseline is implemented or compared. Without at least one matched-filtering baseline or a previously published deep-learning separation method evaluated on the same test set, the reported counting accuracy and overlap values do not establish the claimed improvement. Please add a baseline comparison on the same data, or temper the claim accordingly.
minor comments (6)
  1. [Abstract and Introduction] The phrase 'unknown number of concurrent signals' should be qualified as 'between 2 and 5'; the model is not designed for zero, one, or more than five signals, as stated later in Section II.B.c.
  2. [Section II.A, Table I] The declination δ is sampled uniformly from 0 to π, but physical declination is usually defined in [-π/2, π/2]; this should be clarified or corrected, as it affects the antenna-pattern coverage of the simulated sky.
  3. [Section II.B.d and Section II.C] The value of the loss trade-off parameter λ in Eq. (10) is not reported in the implementation details; please provide the value and any sensitivity analysis.
  4. [Section II.B.b] The description of the FAN layer in Eq. (11) is incomplete: the dimensions and roles of W_p, W_{\bar p}, and B_{\bar p} are not stated.
  5. [Section III.B and Fig. 4] The histograms in Fig. 4 show per-signal overlaps sorted in descending order, but the text does not state how many test samples each panel contains or whether the same samples are used for all signal counts; adding this information would improve interpretability.
  6. [Throughout] Please correct typographical issues: 'burried' in the Fig. 5 caption, inconsistent spacing in 'F AN', and the use of 'CBSs' in the Introduction should be made consistent.

Circularity Check

0 steps flagged · score 0.0 of 10

No significant circularity: the central claims are empirical results on a held-out synthetic test set, not derivations from fitted inputs or self-citations.

full rationale

The paper's central claims—99.89% counting accuracy and 0.9831 mean overlap—are empirical measurements on a 20,000-sample test set generated independently of the 100,000-sample training set under the same simulation pipeline described in Section II.A. The counting head is trained with a cross-entropy loss and the decoders with a permutation-invariant SI-SNR loss; the reported metrics are evaluated on unseen samples, so the reported performance does not reduce to the training objective by construction. The overlap metric (Eq. 5) is mathematically related to the SI-SNR objective (Eqs. 8-9) in that both reward waveform shape similarity, but evaluation on held-out data provides independent statistical evidence rather than a tautology. The shared CE-40km Gaussian PSD in noise generation, whitening, and the SNR/overlap inner product is a distributional assumption about detector noise, not a logical circularity; it affects external validity and real-detector transferability but does not make the measured in-distribution accuracy a restatement of the inputs. Architectural components (dual-path transformer, FAN, multi-decoder DPRNN) are cited from external or non-load-bearing sources, and no uniqueness theorem or fitted parameter is invoked to force the main result. The single-signal denoising and precession/eccentricity tests are out-of-distribution probes reported as generalization checks, not as in-distribution predictions. No specific equation or fitted parameter could be exhibited that reduces a claimed prediction to its input, so per the hard rules the appropriate finding is no significant circularity.

Assumptions & free parameters 2 free parameters · 5 assumptions · 0 invented entities

No new physical entities are introduced. The central claims rest on simulated data and standard waveform models. The main free parameters are the unspecified loss weight lambda and the fixed maximum signal count K=5.

free parameters (2)
  • lambda (loss trade-off weight) = not specified
    In Eq. (10), L = lambda * L_SI-SNR + (1-lambda) * L_CE, but the paper never states lambda's value. It controls the balance between counting and separation during training.
  • K (maximum number of concurrent signals) = 5
    The model assumes a known upper limit K on the number of sources; training only includes 2-5 signals. This is a hand-chosen architecture constraint that limits real-world applicability.
assumptions (5)
  • domain assumption Linear superposition of signals and noise: d(t) = sum_i s_i(t) + n(t) (Eq. 1)
    Standard in GW data analysis, but assumes negligible non-linear interaction between overlapping signals and ignores correlated noise.
  • domain assumption SEOBNRv4, IMRPhenomT, and TaylorF2 waveforms are accurate models for BBH, NS-BH, and BNS signals respectively
    The model is trained and evaluated on these templates; real signals may deviate, especially for NS-BH with tides or higher modes.
  • domain assumption Gaussian noise with CE-40km PSD approximates real detector noise
    Real detector noise is non-Gaussian and non-stationary; the paper's results are only demonstrated on this simulated noise.
  • ad hoc to paper The number of concurrent signals in any input is between 2 and 5 (with one denoising test)
    The counting head is a 4-class classifier over {2,3,4,5}; the model cannot correctly process 1 signal in general or more than 5 signals.
  • domain assumption SI-SNR and overlap are appropriate fidelity metrics for waveform reconstruction
    These metrics measure reconstruction quality in a scale-invariant way, but are not validated against the scientific requirements for parameter estimation.

how reviews work

0 comments
Cite this review

Pith. "Pith review of Compact Binary Coalescence Gravitational Wave Signals Counting and Separation." pith.science (2026). https://pith.science/paper/5VANXAP4

@misc{pith2026241218259,
  author       = {Pith},
  title        = {Pith review of: Compact Binary Coalescence Gravitational Wave Signals Counting and Separation},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/5VANXAP4}},
  note         = {Machine review of arXiv:2412.18259}
}
read the original abstract

As next-generation gravitational-wave (GW) observatories approach unprecedented sensitivities, the need for robust methods to analyze increasingly complex, overlapping signals becomes ever more pressing. Existing matched-filtering approaches and deep-learning techniques can typically handle only one or two concurrent signals, offering limited adaptability to more varied and intricate superimposed waveforms. To overcome these constraints, we present the UnMixFormer, an attention-based architecture that not only identifies the unknown number of concurrent compact binary coalescence GW events but also disentangles their individual waveforms through a multi-decoder architecture, even when confronted with five overlapping signals. Our UnMixFormer is capable of capturing both short- and long-range dependencies by modeling them in a dual-path manner, while also enhancing periodic feature representation by incorporating Fourier Analysis Networks. Our approach adeptly processes binary black hole, binary neutron star, and neutron star-black hole systems over extended time series data (16,384 samples). When evaluating on synthetic data with signal-to-noise ratios (SNR) ranging from 10 to 50, our method achieves 99.89% counting accuracy, a mean overlap of 0.9831 between separated waveforms and templates, and robust generalization ability to waveforms with spin precession, orbital eccentricity, and higher modes, marking a substantial advance in the precision and versatility of GW data analysis.

Figures

Figures reproduced from arXiv: 2412.18259 by the authors.

Figure 1
Figure 1. Illustration of training data with overlapping signals. This figure demonstrates the overlapping of multiple GW signals, including two BBH (BBH-1, BBH-2), two BNS (BNS-1, BNS-2), and one NS-BH. The individual signals are projected onto the detector with SNRs of 12, 15, 30, 20, and 15, respectively. The corresponding chirp masses are 45.14, 24.44, 1.63, 1.38, and 4.97 M⊙. The bottom waveform depicts the combined data… view at source ↗
Figure 2
Figure 2. UnMixFormer Architecture. (a). The overall framework for counting and separating overlapping GW signals. We firstly employ CNN-based encoders to extract data embeddings, which are then fused and passed into UnMixFormer blocks. The counting head predicts the number of sources and activates the appropriate decoder to reconstruct individual waveforms. (b). The core UnMixFormer block operates with intra- and inter-atten… view at source ↗
Figure 3
Figure 3. Counting performance of overlapping CBC signals. (a). The normalized confusion matrix shows the high accuracy of predicting the number of overlapping signals, with correct predictions dominating the diagonal entries (2 to 5 signals). (b). ROC curves illustrating the performance of signal counting for varying numbers of signals. The curves demonstrate near-perfect detection across all cases. Table I. Summary of param… view at source ↗
Figures from the paper (5 more)
Figure 4
Figure 4. Figure 4: Overlap distribution for separated signals. Histograms of the overlap between separated waveforms and their corresponding target templates for cases with (a) 2 signals, (b) 3 signals, (c) 4 signals, and (d) 5 signals. The overlap measures the similarity between the sep…
Figure 5
Figure 5. Figure 5 [PITH_FULL_IMAGE:figures/full_fig_p008_5.png]
Figure 6
Figure 6. Figure 6: Showcase of inspiral-only signal separation performance. The top panel depicts mixed data containing five target inspiral-only signals buried in noise. The subsequent panels display the separated individual signals alongside their corresponding target templates, the ov…
Figure 8
Figure 8. Figure 8: Generalization ability to single signal de￾noising scenario. This figure illustrates our model’s capac￾ity to denoise and reconstruct a single GW signal from noisy data. The high overlap demonstrating robust generalization ability to denoising scenarios. include 3 BBH …
Figure 9
Figure 9. Figure 9: Generalization performance across different precession and orbital eccentricity. (a) The mismatch (log10 M) as a function of the spin precession parameter (p (sx) 2 + (sy) 2) for varying numbers of signals (2 to 5), demonstrating the robustness of our model under diffe…

Discussion (0). Continue with ORCID to comment.

Reference graph

Works this paper leans on

68 extracted references · 54 canonical work pages

  1. [1]

    B. P. Abbott, R. Abbott, T. D. Abbott, M. R. Aber- nathy, F. Acernese, et al., Phys. Rev. Lett. 116, 061102 (2016)

    work page 2016

  2. [2]

    Cahillane and G

    C. Cahillane and G. Mansell, Galaxies 10, 36 (2022)

    2022

  3. [3]

    D. V. Martynov, E. D. Hall, B. P. Abbott, R. Abbott, T. D. Abbott, et al., Phys. Rev. D 93, 112004 (2016)

    work page 2016

  4. [4]

    Maggiore, C

    M. Maggiore, C. V. D. Broeck, N. Bartolo, E. Belgacem, D. Bertacca, et al., J. Cosmol. Astropart. Phys. 2020, 050 (2020)

    work page 2020

  5. [5]

    A Horizon Study for Cosmic Ex- plorer: Science, Observatories, and Community,

    M. Evans, R. X. Adhikari, C. Afle, S. W. Ballmer, S. Biscoveanu, et al., “A Horizon Study for Cosmic Ex- plorer: Science, Observatories, and Community,” (2021), arXiv:2109.09882

    arXiv 2021

  6. [6]

    Laser Interferometer Space Antenna,

    P. Amaro-Seoane, H. Audley, S. Babak, J. Baker, E. Ba- rausse, et al., “Laser Interferometer Space Antenna,” (2017), arXiv:1702.00786

    arXiv 2017

  7. [7]

    Hu and Y.-L

    W.-R. Hu and Y.-L. Wu, Natl. Sci. Rev. 4, 685 (2017)

    2017

  8. [8]

    Z. Ren, T. Zhao, Z. Cao, Z.-K. Guo, W.-B. Han, et al., Front. Phys. 18, 64302 (2023)

    work page 2023

Show all 68 references
  1. [9]

    Luo, L.-S

    J. Luo, L.-S. Chen, H.-Z. Duan, Y.-G. Gong, S. Hu, et al., Class. Quantum Gravity 33, 035010 (2016), arXiv:1512.02076

    2016 arXiv

  2. [10]

    Bailes, B

    M. Bailes, B. K. Berger, P. R. Brady, M. Branchesi, K. Danzmann, et al., Nat. Rev. Phys. 3, 344 (2021)

    2021

  3. [11]

    Rigorous analytic solu- tion to the gravitational-wave overlapping event rates,

    Z. Wang, Z. Hu, and L. Shao, “Rigorous analytic solu- tion to the gravitational-wave overlapping event rates,” (2025), arXiv:2501.05218

    2025

  4. [12]

    Iacovelli, M

    F. Iacovelli, M. Mancarella, S. Foffa, and M. Maggiore, Astrophys. J. 941, 208 (2022)

    2022

  5. [13]

    Pieroni, A

    M. Pieroni, A. Ricciardone, and E. Barausse, Sci. Rep. 12, 17940 (2022)

    2022

  6. [14]

    M. C. Digman and N. J. Cornish, Astrophys. J. 940, 10 (2022)

    2022

  7. [15]

    Arca Sedda, P

    M. Arca Sedda, P. Amaro Seoane, and X. Chen, Astron. Astrophys. 652, A54 (2021)

    2021

  8. [16]

    The LISA Data Challenges,

    Q. Baghi, “The LISA Data Challenges,” (2022), arXiv:2204.12142

    2022 arXiv

  9. [17]

    Himemoto, A

    Y. Himemoto, A. Nishizawa, and A. Taruya, Phys. Rev. D 104, 044010 (2021)

    2021

  10. [18]

    Chamberlain and N

    K. Chamberlain and N. Yunes, Phys. Rev. D 96, 084039 (2017)

    2017

  11. [19]

    A. D. Johnson, K. Chatziioannou, and W. M. Farr, Phys. Rev. D 109, 084015 (2024)

    2024

  12. [20]

    International Partnership Program of the Chinese Academy of Sciences

    In the case where the SNR is equal to 10, some sam- ples exhibit an overlap of 0. This occurs because the estimated number of signals exceeds the true number of signals, leaving no corresponding target for the extra es- timated signal, resulting in an overlap of 0. As the num-...

  13. [21]

    Regimbau and S

    T. Regimbau and S. A. Hughes, Phys. Rev. D 79, 062002 (2009)

    2009

  14. [22]

    Hourihane, K

    S. Hourihane, K. Chatziioannou, M. Wijngaarden, D. Davis, T. Littenberg, et al., Phys. Rev. D 106, 042006 (2022)

    2022

  15. [23]

    Badaracco, B

    F. Badaracco, B. Banerjee, M. Branchesi, and A. Chin- carini, New Astron. Rev. 99, 101707 (2024)

    2024

  16. [24]

    Wu and A

    S. Wu and A. H. Nitz, Phys. Rev. D 107, 063022 (2023)

    2023

  17. [25]

    Samajdar, J

    A. Samajdar, J. Janquart, C. Van Den Broeck, and T. Di- etrich, Phys. Rev. D 104, 044003 (2021)

    2021

  18. [26]

    Relton and V

    P. Relton and V. Raymond, Phys. Rev. D 104, 084039 (2021)

    2021

  19. [27]

    Janquart, T

    J. Janquart, T. Baka, A. Samajdar, T. Dietrich, and C. Van Den Broeck, Mon. Not. R. Astron. Soc.523, 1699 (2023)

    2023

  20. [28]

    Pizzati, S

    E. Pizzati, S. Sachdev, A. Gupta, and B. S. Sathyaprakash, Phys. Rev. D 105, 104016 (2022)

    2022

  21. [29]

    Z. Wang, D. Liang, J. Zhao, C. Liu, and L. Shao, Class. Quantum Gravity 41, 055011 (2024)

    2024

  22. [30]

    Antonelli, O

    A. Antonelli, O. Burke, and J. R. Gair, Mon. Not. R. Astron. Soc. 507, 5069 (2021)

    2021

  23. [31]

    Kerachian, S

    M. Kerachian, S. Mukherjee, G. Lukes-Gerakopoulos, and S. Mitra, Astron. Astrophys. 684, A17 (2024)

    2024

  24. [32]

    Relton, A

    P. Relton, A. Virtuoso, S. Bini, V. Raymond, I. Harry, et al., Phys. Rev. D 106, 104045 (2022)

    2022

  25. [33]

    Y. Dang, Z. Wang, D. Liang, and L. Shao, Astrophys. J. 964, 194 (2024)

    2024

  26. [34]

    Hu and J

    Q. Hu and J. Veitch, Astrophys. J. 945, 103 (2023)

    2023

  27. [35]

    R. Shi, Y. Zhou, T. Zhao, Z. Cao, and Z. Ren, Phys. Rev. D 109, 084017 (2024)

    2024

  28. [36]

    Dawning of a new era in gravitational wave data analysis: Unveil- ing cosmic mysteries via artificial intelligence – a system- atic review,

    T. Zhao, R. Shi, Y. Zhou, Z. Cao, and Z. Ren, “Dawning of a new era in gravitational wave data analysis: Unveil- ing cosmic mysteries via artificial intelligence – a system- atic review,” (2023), arXiv:2311.15585

    2023 arXiv

  29. [37]

    George and E

    D. George and E. A. Huerta, Phys. Rev. D 97, 044039 (2018)

    2018

  30. [38]

    R. Shi, Y. Zhou, T. Zhao, Z. Wang, Z. Ren, et al., Phys. Rev. D 111, 044016 (2025)

    2025

  31. [39]

    T. Zhao, Y. Zhou, R. Shi, Z. Cao, and Z. Ren, Phys. Rev. D 109, 084054 (2024)

    2024

  32. [40]

    Gabbard, M

    H. Gabbard, M. Williams, F. Hayes, and C. Messenger, Phys. Rev. Lett. 120, 141103 (2018)

    2018

  33. [41]

    T. Zhao, R. Lyu, H. Wang, Z. Cao, and Z. Ren, Commun. Phys. 6, 212 (2023)

    2023

  34. [42]

    E. A. Huerta, A. Khan, X. Huang, M. Tian, M. Levental, et al., Nat. Astron. 5, 1062 (2021)

    2021

  35. [43]

    Wei, Applications of Deep Learning for Gravita- tional Wave Physics, Ph.D

    W. Wei, Applications of Deep Learning for Gravita- tional Wave Physics, Ph.D. thesis, University of Illinois Urbana-Champaign, Champaign, IL, US (2021)

    2021

  36. [44]

    Wei and E

    W. Wei and E. A. Huerta, Phys. Lett. B 800, 135081 (2020)

    2020

  37. [45]

    M. Dax, S. R. Green, J. Gair, J. H. Macke, A. Buonanno, et al., Phys. Rev. Lett. 127, 241103 (2021)

    2021

  38. [46]

    H. Wang, Y. Zhou, Z. Cao, Z. Guo, and Z. Ren, Mach. Learn.: Sci. Technol. 5, 015046 (2024)

    2024

  39. [47]

    Wildberger, M

    J. Wildberger, M. Dax, S. R. Green, J. Gair, M. P¨ urrer, et al., Phys. Rev. D 107, 084046 (2023)

    2023

  40. [48]

    M. Dax, S. R. Green, J. Gair, M. P¨ urrer, J. Wildberger, et al., Phys. Rev. Lett. 130, 171403 (2023)

    2023

  41. [49]

    What to do when things get crowded? Scalable joint analysis of overlapping gravitational wave signals,

    J. Alvey, U. Bhardwaj, S. Nissanke, and C. Weniger, “What to do when things get crowded? Scalable joint analysis of overlapping gravitational wave signals,” (2023), arXiv:2308.06318

    2023 arXiv

  42. [50]

    Langendorff, A

    J. Langendorff, A. Kolmus, J. Janquart, and C. Van Den Broeck, Phys. Rev. Lett. 130, 171402 (2023)

    2023

  43. [51]

    Dal Canton, A

    T. Dal Canton, A. H. Nitz, B. Gadre, G. S. Cabourn Davies, V. Villa-Ortega, et al., Astrophys. J. 923, 254 (2021)

    2021

  44. [52]

    These parameters are then used to generate the corre- sponding waveforms for each source

    waveform template is used; for NS-BH systems, the IMRPhenomT [53] waveform template is employed; and for BNS, the TaylorF2 [54] waveform template is utilized. These parameters are then used to generate the corre- sponding waveforms for each source. b. Detector response After s...

  45. [53]

    Gravitational Wave Mixture Separation for Future Gravitational Wave Observatories Utilizing Deep Learning,

    C. Ma, W. Zhou, and Z. Cao, “Gravitational Wave Mixture Separation for Future Gravitational Wave Observatories Utilizing Deep Learning,” (2024), arXiv:2407.13239

    2024 arXiv

  46. [54]

    Boh´ e, L

    A. Boh´ e, L. Shao, A. Taracchini, A. Buonanno, S. Babak, 13 et al., Phys. Rev. D 95, 044028 (2017)

    2017

  47. [55]

    Estell´ es, M

    H. Estell´ es, M. Colleoni, C. Garc ´ ıa-Quir´ os, S. Husa, D. Keitel, et al., Phys. Rev. D 105, 084040 (2022)

    2022

  48. [56]

    Messina, R

    F. Messina, R. Dudi, A. Nagar, and S. Bernuzzi, Phys. Rev. D 99, 124051 (2019)

    2019

  49. [57]

    Whelan, Visualization of Antenna Pattern Factors via Projected Detector Tensors, Tech

    J. Whelan, Visualization of Antenna Pattern Factors via Projected Detector Tensors, Tech. Rep. LIGO-T1100431- v2 (2012)

    2012

  50. [58]

    J. Chen, Q. Mao, and D. Liu, in Interspeech 2020(ISCA,

    2020

  51. [59]

    Y. Luo, Z. Chen, and T. Yoshioka, in ICASSP 2020 - 2020 IEEE International Conference on Acoustics, Speech and Signal Processing (ICASSP)(2020) pp. 46– 50

    2020

  52. [60]

    Zhu, Multi-Decoder DPRNN High Accuracy Source Counting and Separation , Senior, University of Illi- nois at Urbana-Champaign, Champaign, IL, US (2020), arXiv:2011.12022

    J. Zhu, Multi-Decoder DPRNN High Accuracy Source Counting and Separation , Senior, University of Illi- nois at Urbana-Champaign, Champaign, IL, US (2020), arXiv:2011.12022

    2020 arXiv

  53. [61]

    F AN: Fourier Analysis Networks,

    Y. Dong, G. Li, Y. Tao, X. Jiang, K. Zhang, et al., “F AN: Fourier Analysis Networks,” (2024), arXiv:2410.02675

    2024

  54. [62]

    Vincent, R

    E. Vincent, R. Gribonval, and C. Fevotte, IEEE Trans. Audio Speech Lang. Process. 14, 1462 (2006)

    2006

  55. [63]

    Pariente, S

    M. Pariente, S. Cornell, J. Cosentino, S. Sivasankaran, E. Tzinis, et al., in Interspeech 2020 (ISCA, 2020) pp. 2637–2641

    2020

  56. [64]

    H. Wu, T. Hu, Y. Liu, H. Zhou, J. Wang, et al., in The Eleventh International Conference on Learning Rep- resentations (2022)

    2022

  57. [65]

    Deep Time Series Models: A Comprehensive Survey and Benchmark,

    Y. Wang, H. Wu, J. Dong, Y. Liu, M. Long, et al., “Deep Time Series Models: A Comprehensive Survey and Benchmark,” (2024), arXiv:2407.13278

    2024 arXiv

  58. [66]

    D. P. Kingma and J. Ba, in 3rd International Conference on Learning Representations, ICLR 2015, San Diego, CA, USA, edited by Y. Bengio and Y. LeCun (2015) arXiv:1412.6980

    2015 arXiv

  59. [67]

    X. Liu, Z. Cao, and Z.-H. Zhu, Class. Quantum Gravity 41, 195019 (2024)

    2024

  60. [68]

    Cao and W.-B

    Z. Cao and W.-B. Han, Phys. Rev. D 96, 044028 (2017)

    2017

Pith tools

Reviewed August 11, 2026 · model on record in the stance chip above.