Pith. sign in

REVIEW 3 major objections 5 minor 53 references

Evaluation of Seismic Artificial Intelligence with Uncertainty

T0 review · 3 major / 5 minor · reviewed 2026-08-10 · deepseek-v4-flash

Pith's one-line read Seismic AI's best training approach depends on the amount of available data, not a single global winner.

desk verdict A useful evaluation framework for seismic deep learning models, but the split design lets station identity leak across train and test, and without code or a station-disjoint baseline the central learning-efficiency claims are not fully supported. read the letter →

arxiv 2501.14809 v2 pith:U53ZBMU5 submitted 2025-01-15 cs.LG cs.AIphysics.geo-ph

classification cs.LGcs.AIphysics.geo-ph
keywords uncertaintyquantificationmodelbenchmarkingdeeplearningefficiencyseismicphasepickingtransferdataleakagestatisticalexperimentaldesign
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 argues that standard evaluations of seismic deep learning models can mislead because they ignore two sources of variability: random data sampling and stochastic training, which jointly make performance uncertain. It proposes an evaluation framework that measures both sources and links them to learning efficiency, defined as performance gain per amount of diverse training data. The paper demonstrates the framework on PhaseNet, a widely used phase-picking model, comparing three training approaches across five data budgets. The central finding is that no training approach wins across all budgets: transfer learning helps most at the smallest data size, while training from scratch overtakes it once roughly 30,000 waveforms are available, so model choice should depend on a practitioner's data budget.

What carries the argument

The load-bearing mechanism is a three-part evaluation design. First, INSTANCE earthquake sources are clustered into 20 groups by latitude and longitude with k-means; 12 central clusters supply training and validation data, while 8 northern and southern clusters form a balanced test set, so that data budgets add whole geographic clusters rather than random waveforms and leakage from similar signals is reduced. Second, each model is trained on 12 random selections of 1, 3, 6, 9, or 12 clusters, with 4 deep-ensemble initializations per selection, yielding 720 model instances. Third, a mixed-effects statistical model, $y_{madi} = \gamma + \mu_m + \alpha_a + \theta_{ma} + \epsilon^{\mathrm{data}}_{mad} + \epsilon^{\mathrm{train}}_{madi}$, estimates the effect of training approach, data budget, and their interaction while splitting total variance into data-sampling and training components, and produces confidence intervals for every metric used in comparisons.

What would settle it

Compute pairwise waveform similarities between clusters: if waveforms drawn from different source-location clusters are on average as similar as waveforms within a cluster, then location clustering has not controlled leakage and the training budgets are not adding independent information, which would invalidate the learning-efficiency comparison.

Watch

Extended reading notes

Core claim

The paper's central claim is that a fair comparison of seismic AI models requires jointly estimating performance uncertainty and learning efficiency, and that once these are accounted for, no model is a global winner. Using a mixed-effects statistical model applied to 720 PhaseNet instances, the authors separate data uncertainty (which cluster subset of training sources is drawn) from training uncertainty (which of 4 initializations is used). They find that the transfer-learning model with free weights has the best mean recall at one training cluster, while the from-scratch model is best at three or more clusters, and that rankings from a single model instance carry substantial mis-ranking risk, such as a 2% chance that the worst average model appears to be the best on recall. The paper concludes that expending effort on very large labeled datasets may be wasteful, since performance gains diminish beyond roughly 30,000 to 60,000 waveforms.

Load-bearing premise

The framework's conclusions rest on the assumption that k-means clustering by source location sufficiently reduces data leakage and captures the diversity that matters for learning efficiency; if waveforms from different clusters still share substantial information through common stations, networks, or propagation paths, the learning-efficiency curves and uncertainty estimates would be biased.

Editorial extensions

If this is right

  • Practitioners can choose a training approach by matching their labeled-data budget to the estimated performance curves instead of trusting a single benchmark ranking.
  • Developers know where to spend effort: when data uncertainty dominates, label or acquire more diverse data; when training uncertainty dominates, run more initializations or ensembles.
  • Reported state-of-the-art claims for seismic pickers are unreliable unless they come with variance estimates; a single run can put the worst average model first.
  • Acquiring ever-larger training sets has diminishing returns; roughly 30,000 to 60,000 waveforms may suffice for reasonable PhaseNet performance even from scratch.
  • Transfer learning is not automatically better than training from scratch; its advantage is limited to small data budgets in this setting, which matters for foundation-model fine-tuning.

Reading between the lines

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

  • Because the framework treats the data split as part of the experiment, the same design could benchmark seismic foundation models against simpler pickers, testing whether pretraining on global data beats from-scratch training at each fine-tuning budget.
  • The choice of source-location clustering is one of many possible stratifications; a natural extension is to test whether clustering on time, network, magnitude, or signal-to-noise ratio changes the learning-efficiency conclusions, since the authors note that a single feature captures diversity only partially.
  • The frequent negative estimates of data variance at large budgets suggest the design under-samples cluster combinations; increasing the number of random cluster sets beyond 12, or sampling with replacement, could give sharper uncertainty estimates where the paper currently reports near-zero data variance.
  • Outside seismology, the same template could apply to any benchmark where geographic or domain structure creates leakage, such as medical imaging from different hospitals, by replacing source-location clusters with site or device clusters.
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 an evaluation framework for seismic deep learning models that jointly quantifies performance uncertainty (from stochastic training and random data sampling) and learning efficiency as a function of training data budget. The authors construct train/validation/test splits from the INSTANCE dataset by k-means clustering on source locations, train 720 PhaseNet instances under three training approaches (standard, transfer learning with free weights, and transfer learning with frozen encoder) across five data-budget levels (1, 3, 6, 9, and 12 clusters), and summarize classification and regression metrics using a mixed-effects model (Eq. 1). They report that no training approach is globally best across budgets, that uncertainty decreases with more data, and that ignoring uncertainty creates a substantial risk of mis-ranking models. The paper also discusses negative variance estimates as a known limitation and suggests future improvements such as more cluster sets or a Bayesian framework.

Significance. If the framework's leakage controls are adequate, this paper makes a useful contribution to seismic AI benchmarking by drawing attention to learning efficiency and uncertainty, and by providing a concrete experimental design (720 model instances on public data) that others can emulate. The statistical model is straightforward and its Gaussian assumptions are checked with QQ plots. The authors are transparent about limitations, including negative variance estimates and the single-feature stratification. The main practical claim—that model rankings depend on the amount of available training data—is nontrivial and relevant to practitioners choosing between training from scratch and transfer learning, and to the seismic foundation-model community. However, the validity of the learning-efficiency and uncertainty conclusions depends on the split adequately preventing data leakage, which is the central issue I raise in the major comments.

major comments (3)
  1. [Section 3.1, Figure 2, Table 1] The split construction in Section 3.1 assigns earthquake waveforms to clusters by source location and noise waveforms to clusters by station location, but it never blocks on station identity. Because a fixed station records events from many source clusters, test waveforms from northern and southern sources can be recorded at stations whose other waveforms—from central sources, or noise waveforms assigned by station location—appear in the training pool. The manuscript itself lists 'nearby stations' as a source of data leakage in Section 2, and Figures A1-A5 show cluster-dependent signal properties, so station-level site effects are likely shared across the split. Table 1 reports only source and waveform counts, not station overlap, and Figure 2 (bottom) does not identify which stations are in training versus test. This is load-bearing because the learning-efficiency curves in Figures 6-7 and the uncertainty decomposition in Figure 8 could all be optimistically biased if test waveforms are more similar to training waveforms than a station-disjoint split would permit. I request (i) reporting the number and proportion of stations shared between the training and test waveform sets, and (ii) a sensitivity analysis using a station-disjoint split, or an alternative demonstration that station sharing does not affect the ranking conclusions.
  2. [Section 7, Figure 8] The paper acknowledges that data variance is often estimated as negative for larger training budgets and is not displayed. Because the framework's stated contribution includes segregating performance uncertainty arising from stochastic training and random data sampling (Section 5.2), the current design does not support reliable statements about the relative contribution of data variance for budgets of 9 and 12 clusters, where the number of distinct cluster draws is small (and exactly one for 12 clusters). The suppression of negative estimates in Figure 8 makes the relative-contribution bars potentially misleading. Please report the frequency of negative variance estimates across all metrics and budgets, and state explicitly which budget ranges the variance decomposition is trustworthy for.
  3. [Section 6, Figures 6-7] The test set uses only the 4 northernmost and 4 southernmost clusters, while all training and validation data come from the 12 central clusters (Section 3.1). The paper positions this as a semi-out-of-distribution evaluation, which is a reasonable choice, but the central conclusion 'no model is a global winner in terms of learning efficiency' (Section 8) is therefore about performance on a particular semi-OOD test distribution. The ranking could differ for an in-distribution test set sampled from the central clusters, and the current paper provides no evidence about that case. Adding an in-distribution test set, or explicitly re-scoping the conclusion to semi-OOD generalization, would clarify the claim.
minor comments (5)
  1. [Author affiliations] The affiliations use 'V A' with a space; this should be 'VA', both in the header and anywhere else it appears.
  2. [Section 3.1] The validation split description says 'Maintaining a ∼11.4% earthquake to noise ratio', but the text immediately before refers to the 'noise to earthquake waveform ratio' in INSTANCE; the ratio should be stated consistently as noise-to-earthquake.
  3. [Figure 8 caption] The caption says 'the opaque and transparent bars are the contributions of data and training variances' and then says 'the training contribution is transparent and the data contribution is opaque'; this is internally inconsistent and should be clarified.
  4. [Section 7] The word 'wholistic' should be 'holistic'.
  5. [General] The manuscript does not include a data or code availability statement; since the framework is meant to be reusable, stating whether code and trained model weights will be released would aid reproducibility.

Circularity Check

0 steps flagged · score 0.0 of 10

No circularity: the evaluation is empirical, and the reported comparisons are not forced by construction or by self-citation.

full rationale

This is an empirical evaluation paper, not a derivation of model behavior from the framework's assumptions. The central quantities, recall, noise percent correct, cumulative RMSR, and variance components, are computed directly from held-out test waveforms after training 720 model instances; no fitted constant is subsequently renamed as a prediction, and no model-superiority conclusion is built into the loss, threshold rule, or statistical decomposition. Equation (1) is a descriptive mixed-effects model used to summarize measured outcomes, not a source of those outcomes. The contested experimental choices, such as k-means source-location splitting, manual selection of northern and southern test clusters, and validation-set threshold selection, affect external validity and could bias results if station-level leakage exists, but they are not circular: the test set is not used to fit thresholds or to select clusters by performance, and the paper's acknowledged limitations explicitly state that conclusions may change under different data choices. There is no load-bearing self-citation and no imported uniqueness theorem. Therefore no circular step is identified.

Assumptions & free parameters 6 free parameters · 5 assumptions · 0 invented entities

The central claim is an empirical evaluation framework, so the ledger mostly records experimental design choices rather than fitted physical constants. The conclusions depend on the cluster split design, the number of repetitions, and the statistical error model, all of which are reasonable but not independently validated.

free parameters (6)
  • Number of k-means clusters k = 20
    Chosen by hand to keep clusters large enough for regional activity but small enough for low-data settings; affects split granularity and all downstream uncertainty estimates.
  • Number of random cluster sets = 12
    Chosen to estimate data sampling uncertainty; the paper notes in Section 7 that more cluster sets would improve precision, so this is a practical design choice.
  • Deep ensemble size = 4
    Four initializations per cluster set are used to estimate training uncertainty; this choice affects the stability of the variance decomposition.
  • Per-instance pick threshold = Varies; chosen to maximize mean of F1 and noise percent correct on validation
    The threshold is tuned on validation data for each model instance and affects all reported classification metrics.
  • True positive window = ±0.3 seconds
    A customary choice in phase picking evaluation that determines true positive, false positive, and false negative counts and the cumulative RMSR domain.
  • Gaussian target width = 0.1 seconds standard deviation
    Used to construct P and S training targets following PhaseNet conventions; influences the labels the models learn from.
assumptions (5)
  • domain assumption K-means clustering by source location separates waveforms into groups with reduced information overlap.
    Section 3.1 states the split is 'meant to reduce data leakage', but no quantitative validation is provided to show that leakage is actually reduced relative to random splitting.
  • domain assumption Noise waveforms can be assigned to spatial clusters using station location as a proxy.
    Section 3.1 assigns noise waveforms, which have no source locations, to clusters based on station location. If station and source distributions differ, this could misassign noise data.
  • standard math Training and data uncertainty terms are zero-mean independent Gaussians for each metric.
    Section 5.2 introduces Eq. 1 with independent Gaussian errors. The paper verifies this visually with QQ plots for some metrics, but bounded metrics and negative variance estimates show the approximation has limits.
  • ad hoc to paper Twelve cluster sets and four initializations suffice to estimate the variance components.
    Section 7 acknowledges that more cluster sets would improve data variance estimates; the chosen sizes are practical rather than derived from a power analysis.
  • ad hoc to paper The northern and southern test clusters provide a meaningful semi-out-of-distribution evaluation.
    Section 3.1 manually selects the 4 northernmost and 4 southernmost clusters for testing, and the intended deployment distribution is not established.

how reviews work

0 comments
Cite this review

Pith. "Pith review of Evaluation of Seismic Artificial Intelligence with Uncertainty." pith.science (2026). https://pith.science/paper/U53ZBMU5

@misc{pith2026250114809,
  author       = {Pith},
  title        = {Pith review of: Evaluation of Seismic Artificial Intelligence with Uncertainty},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/U53ZBMU5}},
  note         = {Machine review of arXiv:2501.14809}
}
read the original abstract

Artificial intelligence has transformed the seismic community with deep learning models (DLMs) that are trained to complete specific tasks within workflows. However, there is still lack of robust evaluation frameworks for evaluating and comparing DLMs. We address this gap by designing an evaluation framework that jointly incorporates two crucial aspects: performance uncertainty and learning efficiency. To target these aspects, we meticulously construct the training, validation, and test splits using a clustering method tailored to seismic data and enact an expansive training design to segregate performance uncertainty arising from stochastic training processes and random data sampling. The framework's ability to guard against misleading declarations of model superiority is demonstrated through evaluation of PhaseNet [1], a popular seismic phase picking DLM, under 3 training approaches. Our framework helps practitioners choose the best model for their problem and set performance expectations by explicitly analyzing model performance with uncertainty at varying budgets of training data.

Figures

Figures reproduced from arXiv: 2501.14809 by the authors.

Figure 1
Figure 1. An example waveform from the INSTANCE dataset showing the 3-component response in counts (Z - top, N [PITH_FULL_IMAGE:figures/full_fig_p003_1.png] view at source ↗
Figure 2
Figure 2. INSTANCE data with 20 colored clusters for the sources (top, circles) and the stations (bottom, triangles). [PITH_FULL_IMAGE:figures/full_fig_p004_2.png] view at source ↗
Figure 3
Figure 3. Distributions of selected data features (log epicentral distance, magnitude, log depth, SNR averaged over the [PITH_FULL_IMAGE:figures/full_fig_p006_3.png] view at source ↗
Figures from the paper (6 more)
Figure 4
Figure 4. Figure 4: An example of a model output for a test waveform showing the predicted probability of the P-arrival at each [PITH_FULL_IMAGE:figures/full_fig_p008_4.png]
Figure 5
Figure 5. Figure 5: The Gaussian assumptions are visually verified through quantile-quantile plots. As an example, the sorted [PITH_FULL_IMAGE:figures/full_fig_p009_5.png]
Figure 6
Figure 6. Figure 6: The statistical framework is used to summarize classification results including earthquake recall (left) and [PITH_FULL_IMAGE:figures/full_fig_p010_6.png]
Figure 7
Figure 7. Figure 7: The statistical framework is used to summarize the regression-based functional metric of cumulative RMSR [PITH_FULL_IMAGE:figures/full_fig_p011_7.png]
Figure 8
Figure 8. Figure 8: Training variance (left) and data sampling variance (middle) are estimated with 90% confidence intervals [PITH_FULL_IMAGE:figures/full_fig_p011_8.png]
Figure 9
Figure 9. Figure 9: To demonstrate the importance of considering uncertainty in the evaluation framework, the probability of [PITH_FULL_IMAGE:figures/full_fig_p012_9.png]

Discussion (0). Continue with ORCID to comment.

Reference graph

Works this paper leans on

53 extracted references · 40 canonical work pages

  1. [1]

    PhaseNet: A Deep-Neural-Network-Based Seismic Arrival Time Picking Method.Geophys

    Weiqiang Zhu and Gregory C Beroza. PhaseNet: A Deep-Neural-Network-Based Seismic Arrival Time Picking Method.Geophys. J. Int., October 2018

  2. [2]

    S. J. Arrowsmith, D. T. Trugman, J. MacCarthy, K. J. Bergen, D. Lumley, and M. B. Magnani. Big Data Seismology.Rev. Geophys., 60(2):e2021RG000769, June 2022

  3. [3]

    Deep Learning for Geophysics: Current and Future Trends.Rev

    Siwei Yu and Jianwei Ma. Deep Learning for Geophysics: Current and Future Trends.Rev. Geophys., 59(3):e2021RG000742, September 2021

  4. [4]

    Mostafa Mousavi and Gregory C

    S. Mostafa Mousavi and Gregory C. Beroza. Deep-learning seismology.Science, 377(6607):eabm4470, August 2022

  5. [5]

    Mostafa Mousavi, Gregory C

    S. Mostafa Mousavi, Gregory C. Beroza, Tapan Mukerji, and Majid Rasht-Behesht. Applications of deep neural networks in exploration seismology: A technical survey.Geophysics, 89(1):W A95–W A115, January 2024

  6. [6]

    SeisBench—A Toolbox for Machine Learning in Seismology.Seismol

    Jack Woollam, Jannes Münchmeyer, Frederik Tilmann, Andreas Rietbrock, Dietrich Lange, Thomas Bornstein, Tobias Diehl, Carlo Giunchi, Florian Haslinger, Dario Jozinovi´c, Alberto Michelini, Joachim Saul, and Hugo Soto. SeisBench—A Toolbox for Machine Learning in Seismology.Seismol. Res. Lett., 93(3):1695–1709, May 2022

  7. [7]

    Which Picker Fits My Data? A Quantitative Evaluation of Deep Learning Based Seismic Pickers.J

    Jannes Münchmeyer, Jack Woollam, Andreas Rietbrock, Frederik Tilmann, Dietrich Lange, Thomas Bornstein, Tobias Diehl, Carlo Giunchi, Florian Haslinger, Dario Jozinovi´c, Alberto Michelini, Joachim Saul, and Hugo Soto. Which Picker Fits My Data? A Quantitative Evaluation of Deep Learning Based Seismic Pickers.J. Geophys. Res.: Solid Earth, 127(1):e2021JB02...

  8. [8]

    Accounting for Variance in Machine Learning Benchmarks, March 2021

    Xavier Bouthillier, Pierre Delaunay, Mirko Bronzi, Assya Trofimov, Brennan Nichyporuk, Justin Szeto, Naz Sepah, Edward Raff, Kanika Madan, Vikram V oleti, Samira Ebrahimi Kahou, Vincent Michalski, Dmitriy Serdyuk, Tal Arbel, Chris Pal, Gaël Varoquaux, and Pascal Vincent. Accounting for Variance in Machine Learning Benchmarks, March 2021. arXiv:2103.03098

Show all 53 references
  1. [9]

    Reporting Score Distributions Makes a Difference: Performance Study of LSTM-networks for Sequence Tagging

    Nils Reimers and Iryna Gurevych. Reporting Score Distributions Makes a Difference: Performance Study of LSTM-networks for Sequence Tagging. InProceedings of the 2017 Conference on Empirical Methods in Natural Language Processing, volume Proceedings of the 2017 Conference on Em...

  2. [10]

    Spiegelhalter

    Harvey Goldstein and David J. Spiegelhalter. League Tables and Their Limitations: Statistical Issues in Compar- isons of Institutional Performance.J. R. Stat. Soc. A, 159(3):385, 1996

  3. [11]

    Dietterich

    Thomas G. Dietterich. Approximate Statistical Tests for Comparing Supervised Classification Learning Algorithms. Neural Comput., 10(7):1895–1923, October 1998

  4. [12]

    Gritsenko, Zhe Zhao, Neil Houlsby, Fernando Diaz, Donald Metzler, and Oriol Vinyals

    Mostafa Dehghani, Yi Tay, Alexey A. Gritsenko, Zhe Zhao, Neil Houlsby, Fernando Diaz, Donald Metzler, and Oriol Vinyals. The Benchmark Lottery, July 2021. arXiv:2107.07002

  5. [13]

    Follow the leader(board) with confidence: Estimating p-values from a single test set with item and response variance

    Shira Wein, Christopher Homan, Lora Aroyo, and Chris Welty. Follow the leader(board) with confidence: Estimating p-values from a single test set with item and response variance. InFindings of the Association for Computational Linguistics: ACL 2023, pages 3138–3161, Toronto, Ca...

  6. [14]

    Escalante-B, and Laurenz Wiskott

    Hlynur Davíð Hlynsson, Alberto N. Escalante-B, and Laurenz Wiskott. Measuring the Data Efficiency of Deep Learning Methods. InProceedings of the 8th International Conference on Pattern Recognition Applications and Methods, pages 691–698, 2019. arXiv:1907.02549 [cs]

  7. [15]

    Laith Alzubaidi, Jinshuai Bai, Aiman Al-Sabaawi, Jose Santamaría, A. S. Albahri, Bashar Sami Nayyef Al- dabbagh, Mohammed A. Fadhel, Mohamed Manoufali, Jinglan Zhang, Ali H. Al-Timemy, Ye Duan, Amjed Abdullah, Laith Farhan, Yi Lu, Ashish Gupta, Felix Albu, Amin Abbosh, and Yua...

  8. [16]

    Bornstein, D

    T. Bornstein, D. Lange, J. Münchmeyer, J. Woollam, A. Rietbrock, G. Barcheck, I. Grevemeyer, and F. Tilmann. PickBlue: Seismic Phase Picking for Ocean Bottom Seismometers With Deep Learning.Earth Space Sci., 11(1):e2023EA003332, January 2024

  9. [17]

    SpanSeq: similarity-based sequence data splitting method for improved development and assessment of deep learning projects.NAR Genomics Bioinf., 6(3):lqae106, July 2024

    Alfred Ferrer Florensa, Jose Juan Almagro Armenteros, Henrik Nielsen, Frank Møller Aarestrup, and Philip Thomas Lanken Conradsen Clausen. SpanSeq: similarity-based sequence data splitting method for improved development and assessment of deep learning projects.NAR Genomics Bio...

  10. [18]

    Ross, Men-Andrin Meier, Egill Hauksson, and Thomas H

    Zachary E. Ross, Men-Andrin Meier, Egill Hauksson, and Thomas H. Heaton. Generalized Seismic Phase Detection with Deep Learning.Bull. Seismol. Soc. Am., 108(5A):2894–2901, October 2018. 18 Evaluation of Seismic Artificial Intelligence with UncertaintyA PREPRINT

  11. [19]

    Mostafa Mousavi, William L

    S. Mostafa Mousavi, William L. Ellsworth, Weiqiang Zhu, Lindsay Y . Chuang, and Gregory C. Beroza. Earthquake transformer—an attentive deep-learning model for simultaneous earthquake detection and phase picking.Nat. Commun., 11(1):3952, August 2020

  12. [20]

    CubeNet: Array-Based Seismic Phase Picking with Deep Learning.Seismol

    Guoyi Chen and Junlun Li. CubeNet: Array-Based Seismic Phase Picking with Deep Learning.Seismol. Res. Lett., 93(5):2554–2569, September 2022

  13. [21]

    Deep learning for seismic phase detection and picking in the aftershock zone of 2008 M7.9 Wenchuan Earthquake

    Lijun Zhu, Zhigang Peng, James McClellan, Chenyu Li, Dongdong Yao, Zefeng Li, and Lihua Fang. Deep learning for seismic phase detection and picking in the aftershock zone of 2008 M7.9 Wenchuan Earthquake. Phys. Earth Planet. Inter ., 293:106261, August 2019

  14. [22]

    Mostafa Mousavi, Peter Bailis, and Gregory C

    Weiqiang Zhu, Kai Sheng Tai, S. Mostafa Mousavi, Peter Bailis, and Gregory C. Beroza. An End-To-End Earthquake Detection Method for Joint Phase Picking and Association Using Deep Learning.J. Geophys. Res.: Solid Earth, 127(3):e2021JB023283, March 2022

  15. [23]

    Seismic-phase detection using multiple deep learning models for global and local representations of waveforms.Geophys

    Tomoki Tokuda and Hiromichi Nagao. Seismic-phase detection using multiple deep learning models for global and local representations of waveforms.Geophys. J. Int., 235(2):1163–1182, July 2023

  16. [24]

    OBSTransformer: a deep-learning seismic phase picker for OBS data using automated labelling and transfer learning.Geophys

    Alireza Niksejel and Miao Zhang. OBSTransformer: a deep-learning seismic phase picker for OBS data using automated labelling and transfer learning.Geophys. J. Int., 237(1):485–505, February 2024

  17. [25]

    Cianetti, R

    S. Cianetti, R. Bruni, S. Gaviano, D. Keir, D. Piccinini, G. Saccorotti, and C. Giunchi. Comparison of Deep Learning Techniques for the Investigation of a Seismic Sequence: An Application to the 2019, Mw 4.5 Mugello (Italy) Earthquake.J. Geophys. Res.: Solid Earth, 126(12):e20...

  18. [26]

    Beroza, and William L

    Yongsoo Park, Gregory C. Beroza, and William L. Ellsworth. Basement Fault Activation before Larger Earthquakes in Oklahoma and Kansas.Seism. Rec., 2(3):197–206, July 2022

  19. [27]

    Yoon, Elizabeth S

    Clara E. Yoon, Elizabeth S. Cochran, Elizabeth A. Vanacore, Victor Huerfano, Gisela Báez-Sánchez, John D. Wilding, and Jonathan Smith. A Detailed View of the 2020–2023 Southwestern Puerto Rico Seismic Sequence with Deep Learning.Bull. Seismol. Soc. Am., 113(6):2377–2415, December 2023

  20. [28]

    Armstrong, Zachary Claerhout, Ben Baker, and Keith D

    Alysha D. Armstrong, Zachary Claerhout, Ben Baker, and Keith D. Koper. A Deep-Learning Phase Picker with Calibrated Bayesian-Derived Uncertainties for Earthquakes in the Yellowstone V olcanic Region.Bull. Seismol. Soc. Am., 113(6):2323–2344, December 2023

  21. [29]

    Mostafa Mousavi, Yixiao Sheng, Weiqiang Zhu, and Gregory C

    S. Mostafa Mousavi, Yixiao Sheng, Weiqiang Zhu, and Gregory C. Beroza. STanford EArthquake Dataset (STEAD): A Global Data Set of Seismic Signals for AI.IEEE Access, 7:179464–179476, 2019

  22. [30]

    INSTANCE – the Italian seismic dataset for machine learning.Earth Syst

    Alberto Michelini, Spina Cianetti, Sonja Gaviano, Carlo Giunchi, Dario Jozinovi ´c, and Valentino Lauciani. INSTANCE – the Italian seismic dataset for machine learning.Earth Syst. Sci. Data, 13(12):5509–5544, November 2021

  23. [31]

    Scikit-learn: Machine Learning in Python, June 2018

    Fabian Pedregosa, Gaël Varoquaux, Alexandre Gramfort, Vincent Michel, Bertrand Thirion, Olivier Grisel, Math- ieu Blondel, Andreas Müller, Joel Nothman, Gilles Louppe, Peter Prettenhofer, Ron Weiss, Vincent Dubourg, Jake Vanderplas, Alexandre Passos, David Cournapeau, Matthieu...

  24. [32]

    Development of a high-performance seismic phase picker using deep learning in the Hakone volcanic area.Earth Planets Space, 75(1):85, May 2023

    Ahyi Kim, Yuji Nakamura, Yohei Yukutake, Hiroki Uematsu, and Yuki Abe. Development of a high-performance seismic phase picker using deep learning in the Hakone volcanic area.Earth Planets Space, 75(1):85, May 2023

  25. [33]

    U-Net: Convolutional Networks for Biomedical Image Segmentation

    Olaf Ronneberger, Philipp Fischer, and Thomas Brox. U-Net: Convolutional Networks for Biomedical Image Segmentation. InMedical Image Computing and Computer-Assisted Intervention – MICCAI 2015, volume 9351, pages 234–241, Cham, 2015. Springer International Publishing. Series Ti...

  26. [34]

    Chamberlain, John Townend, and Emily Warren-Smith

    Olivia Pita-Sllim, Calum J. Chamberlain, John Townend, and Emily Warren-Smith. Parametric Testing of EQTransformer’s Performance against a High-Quality, Manually Picked Catalog for Reliable and Accurate Seismic Phase Picking.Seism. Rec., 3(4):332–341, October 2023

  27. [35]

    Delbridge, and David R

    Yongsoo Park, Brent G. Delbridge, and David R. Shelly. Making Phase-Picking Neural Networks More Consistent and Interpretable.Seism. Rec., 4(1):72–80, January 2024

  28. [36]

    Santos-Villalobos, Singanallur V

    Chengping Chai, Monica Maceira, Hector J. Santos-Villalobos, Singanallur V . Venkatakrishnan, Martin Schoenball, Weiqiang Zhu, Gregory C. Beroza, Clifford Thurber, and EGS Collab Team. Using a Deep Neural Network and Transfer Learning to Bridge Scales for Seismic Phase Picking...

  29. [37]

    Fine-Tuning U-Net for Ultrasound Image Segmentation: Different Layers, Different Outcomes.IEEE Trans

    Mina Amiri, Rupert Brooks, and Hassan Rivaz. Fine-Tuning U-Net for Ultrasound Image Segmentation: Different Layers, Different Outcomes.IEEE Trans. Ultrason., Ferroelect., Freq. Contr ., 67(12):2510–2518, December 2020. 19 Evaluation of Seismic Artificial Intelligence with Unce...

  30. [38]

    Simple and Scalable Predictive Uncertainty Estimation using Deep Ensembles, November 2017

    Balaji Lakshminarayanan, Alexander Pritzel, and Charles Blundell. Simple and Scalable Predictive Uncertainty Estimation using Deep Ensembles, November 2017. arXiv:1612.01474

  31. [39]

    Hopcroft, and Kilian Q

    Gao Huang, Yixuan Li, Geoff Pleiss, Zhuang Liu, John E. Hopcroft, and Kilian Q. Weinberger. Snapshot Ensembles: Train 1, get M for free, March 2017. arXiv:1704.00109

  32. [40]

    Ganaie, Minghui Hu, A.K

    M.A. Ganaie, Minghui Hu, A.K. Malik, M. Tanveer, and P.N. Suganthan. Ensemble deep learning: A review.Eng. Appl. Artif. Intell., 115:105151, October 2022

  33. [41]

    PyTorch: An Imperative Style, High-Performance Deep Learning Library, December 2019

    Adam Paszke, Sam Gross, Francisco Massa, Adam Lerer, James Bradbury, Gregory Chanan, Trevor Killeen, Zeming Lin, Natalia Gimelshein, Luca Antiga, Alban Desmaison, Andreas Köpf, Edward Yang, Zach DeVito, Martin Raison, Alykhan Tejani, Sasank Chilamkurthy, Benoit Steiner, Lu Fan...

  34. [42]

    Kingma and Jimmy Ba

    Diederik P. Kingma and Jimmy Ba. Adam: A Method for Stochastic Optimization, January 2017. arXiv:1412.6980

  35. [43]

    Benchmark on the accuracy and efficiency of several neural network based phase pickers using datasets from China Seismic Network.Earthquake Sci., 36(2):113–131, April 2023

    Ziye Yu, Weitao Wang, and Yini Chen. Benchmark on the accuracy and efficiency of several neural network based phase pickers using datasets from China Seismic Network.Earthquake Sci., 36(2):113–131, April 2023

  36. [44]

    Bayesian approach for neural networks—review and case studies.Neural Networks, 14(3):257–274, April 2001

    Jouko Lampinen and Aki Vehtari. Bayesian approach for neural networks—review and case studies.Neural Networks, 14(3):257–274, April 2001

  37. [45]

    Dropout as a Bayesian Approximation: Representing Model Uncertainty in Deep Learning, October 2016

    Yarin Gal and Zoubin Ghahramani. Dropout as a Bayesian Approximation: Representing Model Uncertainty in Deep Learning, October 2016. arXiv:1506.02142

  38. [46]

    A Simple Baseline for Bayesian Uncertainty in Deep Learning, December 2019

    Wesley Maddox, Timur Garipov, Pavel Izmailov, Dmitry Vetrov, and Andrew Gordon Wilson. A Simple Baseline for Bayesian Uncertainty in Deep Learning, December 2019. arXiv:1902.02476

  39. [47]

    A Survey of Uncertainty in Deep Neural Networks, January 2022

    Jakob Gawlikowski, Cedrique Rovile Njieutcheu Tassi, Mohsin Ali, Jongseok Lee, Matthias Humt, Jianxiang Feng, Anna Kruspe, Rudolph Triebel, Peter Jung, Ribana Roscher, Muhammad Shahzad, Wen Yang, Richard Bamler, and Xiao Xiang Zhu. A Survey of Uncertainty in Deep Neural Networ...

  40. [48]

    Bayesian Deep Learning and a Probabilistic Perspective of General- ization, March 2022

    Andrew Gordon Wilson and Pavel Izmailov. Bayesian Deep Learning and a Probabilistic Perspective of General- ization, March 2022. arXiv:2002.08791

  41. [49]

    Hudson, Ehsan Adeli, Russ Altman, Simran Arora, Sydney von Arx, Michael S

    Rishi Bommasani, Drew A. Hudson, Ehsan Adeli, Russ Altman, Simran Arora, Sydney von Arx, Michael S. Bernstein, Jeannette Bohg, Antoine Bosselut, Emma Brunskill, Erik Brynjolfsson, Shyamal Buch, Dallas Card, Rodrigo Castellon, Niladri Chatterji, Annie Chen, Kathleen Creel, Jare...

  42. [50]

    Seismic foundation model: A next generation deep-learning model in geophysics.Geophysics, 90(2):IM59–IM79, March 2025

    Hanlin Sheng, Xinming Wu, Xu Si, Jintao Li, Sibo Zhang, and Xudong Duan. Seismic foundation model: A next generation deep-learning model in geophysics.Geophysics, 90(2):IM59–IM79, March 2025

  43. [51]

    A foundation model enpowered by a multi-modal prompt engine for universal seismic geobody interpretation across surveys, September

    Hang Gao, Xinming Wu, Luming Liang, Hanlin Sheng, Xu Si, Gao Hui, and Yaxing Li. A foundation model enpowered by a multi-modal prompt engine for universal seismic geobody interpretation across surveys, September

  44. [52]

    de Hoop, and Ivan Dokmani ´c

    Tianlin Liu, Jannes Münchmeyer, Laura Laurenti, Chris Marone, Maarten V . de Hoop, and Ivan Dokmani ´c. SeisLM: a Foundation Model for Seismic Waveforms, October 2024. arXiv:2410.15765

  45. [53]

    SeisCLIP: A Seismology Foundation Model Pre-Trained by Multimodal Data for Multipurpose Seismic Feature Extraction.IEEE Trans

    Xu Si, Xinming Wu, Hanlin Sheng, Jun Zhu, and Zefeng Li. SeisCLIP: A Seismology Foundation Model Pre-Trained by Multimodal Data for Multipurpose Seismic Feature Extraction.IEEE Trans. Geosci. Remote Sens., 62:1–13, 2024. 20

Pith tools

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