Pith. sign in

REVIEW 2 major objections 5 minor 50 references

The Riemannian Means Field Classifier for EEG-Based BCI Data

T0 review · 2 major / 5 minor · reviewed 2026-08-16 · deepseek-v4-flash

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

Pith's one-line read A means field of power means, fed as squared distances to LDA, beats the standard MDM across 20 EEG-BCI databases while staying deterministic.

desk verdict A genuinely new Riemannian classifier with a careful 20-database benchmark and honest reporting; the main flaw is an unverified 'hyperparameter-free' claim that should be fixed by sensitivity analysis. read the letter →

arxiv 2504.17352 v1 pith:QE2AR7RO submitted 2025-04-24 cs.HC eess.SP

Anton Andreev (GIPSA-lab) , Grégoire Cattan , Marco Congedo (GIPSA-VIBS , GIPSA-lab) This is my paper · ORCID
classification cs.HCeess.SP
keywords brain-computerinterfaceEEGclassificationRiemanniangeometrysymmetricpositive-definitematricespowermeansminimumdistancetomeanclassifiermotorimageryP300
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 the Means Field (MF) classifier for EEG-based brain–computer interfaces. Instead of classifying a trial by its distance to one geometric mean per class, as the standard MDM does, MF computes the squared Riemannian distance from the trial to each of a family of power means spanning the harmonic-to-arithmetic range, and feeds this vector of distances to a linear discriminant analysis. The authors claim that on a benchmark of 20 public databases (587 subjects, motor imagery and P300), MF clearly outperforms both MDM and the previous MDMF and comes close to the tangent-space logistic regression state of the art, while remaining deterministic and effectively free of hyperparameters. If right, this gives BCI practitioners a simple, reproducible classifier that needs no tuning and no tangent-space projection.

What carries the argument

The central object is the means field: a family of power means $\mathcal{P}_h$ of SPD covariance matrices for exponents $h \in [-1,1]$, interpolating between the harmonic mean ($h=-1$) and arithmetic mean ($h=1$), with the geometric mean as the $h\to 0$ limit. The MF classifier takes the squared Riemannian distance from each trial to every mean in the field and passes that distance vector to LDA, which learns a linear combination of distances; this is what distinguishes it from MDM and MDMF, which each use only one or the minimum distance. Two supporting mechanisms carry the result: RPME iteratively removes trials whose standardized geometric distance exceeds a threshold, making mean estimates robust, and ADCSP reduces high-dimensional motor-imagery covariance matrices in two stages with a CSP-like joint diagonalization.

What would settle it

Re-run the 20-database benchmark while varying the means field (e.g., 5, 11, or 21 power means, and shifted values of $h$) and the RPME threshold $z_{th}$ from 1.5 to 4, and check whether MF still outperforms MDM and MDMF by similar margins. If the advantage over MDM shrinks or reverses for some plausible settings, the central claim of a robust, tuning-free classifier is contradicted.

Watch

Extended reading notes

Core claim

On the paper's own terms, the central discovery is that the soft information contained in a whole field of distances carries more classification signal than the single nearest-mean distance. The MF classifier estimates eleven power means per class, for exponents $h \in \{\pm 1, \pm 0.75, \pm 0.5, \pm 0.25, \pm 0.1, 0\}$, builds a feature vector from the squared affine-invariant distances of each trial to all of those means, and classifies that vector with LDA. Across 20 public EEG-BCI databases the authors report that this pipeline, combined with an adaptive double common spatial pattern filter for motor imagery and robust power-mean estimation for both paradigms, outperforms MDM and MDMF significantly on both paradigms, with meta-analytic effect sizes SMD = 1.122 and 1.057 for motor imagery and 0.522 and 0.681 for P300, all with p < 0.001. It remains slightly but significantly behind tangent-space logistic regression, with meta-effect sizes of -0.28 and -0.498, but the gap is small. The paper presents the result as evidence that manifold-native, deterministic classifiers can rival tangent-space methods.

Load-bearing premise

The paper's claim that the new settings (the 11 power means and their positions, the ADCSP dimension thresholds, and the RPME outlier threshold) are not critical is asserted without sensitivity analysis, so the load-bearing premise is that these reasonable a priori choices do not drive the reported gains.

Editorial extensions

If this is right

  • For motor imagery, MF with ADCSP outperforms MDM and MDMF on every database individually and overall, with meta-effect SMD = 1.122 and 1.057, both with p < 0.001.
  • For P300, MF outperforms MDM and MDMF overall, with meta-effect SMD = 0.522 and 0.681, both with p < 0.001, though TS+LR remains ahead in seven of ten databases.
  • MF approaches but does not surpass tangent-space logistic regression: TS+LR wins six of ten motor-imagery databases and seven of ten P300 databases.
  • RPME improves MF significantly on both paradigms, more so for P300, and ADCSP both improves accuracy and reduces runtime relative to standard CSP.
  • A fully deterministic, essentially tuning-free classifier working directly on the manifold can compete with state-of-the-art tangent-space methods.

Reading between the lines

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

  • Editorial inference: because LDA has a closed-form solution and the means are fixed once the exponents are chosen, the same means-field idea could replace the geometric mean in any MDM-based pipeline, including transfer-learning and calibration-free settings, at low computational cost.
  • Editorial inference: the means field need not be restricted to power means; other SPD mean families such as log-Euclidean or alpha-divergence means could enrich the distance vector, and the clustering variant suggested in the discussion could improve granularity.
  • Editorial inference: the larger RPME gains on P300 suggest that outlier handling matters most in less controlled recording conditions, so applying MF+RPME to out-of-the-lab data is a natural testable extension.
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

2 major / 5 minor

Summary. The paper proposes a Riemannian Means Field (MF) classifier that represents each trial by the squared affine-invariant distances to a field of power means of the class covariance matrices and feeds these features to LDA. The method is compared against the standard Riemannian MDM, the authors' earlier MDMF, and tangent-space logistic regression (TS+LR) on 20 public MOABB databases (10 motor imagery, 10 P300; 587 subjects total), using within-session 5-fold cross-validation with identical folds, per-database permutation/Wilcoxon tests, and a Liptak meta-analysis. The paper also introduces an adaptive double CSP (ADCSP) spatial filter for MI data and a robust power-mean estimation (RPME) procedure. The headline results are that MF outperforms MDM and MDMF with large meta-effect sizes (SMD = 1.122 and 1.057 for MI; 0.522 and 0.681 for P300), while being slightly but significantly worse than TS+LR on both paradigms. The authors characterize MF as deterministic and 'virtually hyperparameter-free'.

Significance. If the empirical results hold, the MF classifier is a practically valuable contribution: it is simple, deterministic, fast when combined with spatial filtering, and substantially closes the gap between manifold-based Riemannian classifiers and tangent-space state-of-the-art methods. The study has notable strengths: it uses 20 public databases, keeps folds identical across pipelines, applies exact permutation tests where possible, combines evidence with a weighted Liptak meta-analysis, and reports effect sizes. The code is made available. The main weakness is that the central 'virtually hyperparameter-free' claim is not supported by sensitivity analysis, and at least one of the introduced choices (RPME) demonstrably changes performance.

major comments (2)
  1. [Section 4 and Section 2.1] The claim that the newly introduced hyperparameters (the 11 power means and their positions, the ADCSP dimension thresholds 28 and 10, the RPME outlier threshold zth=2.5, I=4, tolerance, and iteration limits) are 'not critical' and that 'reasonable a priori choices give results that are statistically indistinguishable' is unsupported. No sensitivity analysis is provided. This is load-bearing because the paper's value proposition is a tuning-free deterministic classifier. Moreover, RPME itself is shown to change results significantly (Figure 10: SMD = -0.297, p < 0.001 for P300; Figure 6: SMD = -0.11, p < 0.01 for MI), so at least one introduced choice materially affects performance. The paper should either provide a sensitivity analysis over the means-field grid, the number of means, the ADCSP thresholds, and zth, or substantially soften the 'virtually hyperparameter-free' framing. The sentence 'Further research should establish a convenient choice of these hyperparameters to be used universally' also concedes that the choice is not yet settled.
  2. [Section 3.1, Table 4, Figures 2-3] The statement that MF 'significantly outperforms' MDM and MDMF 'for all databases taken individually' cannot be correct for databases with very few subjects. For example, Zhou2016 has only 4 subjects (Table 1), so the exact paired permutation test's minimum one-sided p-value is 1/16 = 0.0625, which exceeds 0.05. The authors should clarify what 'individually' means here: if they mean the effect direction is positive in every database, that wording should be used; if they mean per-database statistical significance, the claim is logically impossible for the small-n databases in this study.
minor comments (5)
  1. [Abstract and Section 5] The abstract says 'our code will be released as open source,' but Section 5 says 'Our code is published at ...' — please align these statements.
  2. [Figure 4 caption] The caption contains a typo: 'ACSTP' should be 'ADCSP'.
  3. [Table 2] The entry '1/8' for the number of sessions in Brain Invaders 2013a is unclear; please clarify whether this means one session for some subjects and eight for others, or something else.
  4. [Section 2.3] The statistical analysis would benefit from a statement about whether any correction for multiple comparisons was applied across the many pipeline pairs tested; as written, the reader cannot tell whether the reported p-values are corrected.
  5. [Section 4, Discussion of RPME] The sentence 'This suggests that the P300 data included in MOABB needs to be curated' is presented without evidence; consider rephrasing to avoid implying a data-quality judgment solely from improved outlier rejection.

Circularity Check

0 steps flagged · score 0.0 of 10

No circularity: the classifier's gains are established by empirical benchmarking on public MOABB datasets against external and prior baselines, with no fitted parameter renamed as a prediction.

full rationale

The paper's central claim is an empirical performance comparison, not a derivation. The MF classifier feeds squared Riemannian distances to a field of power means into a hyperparameter-free LDA, and the claimed advantages over MDM, MDMF, and TS+LR are obtained by cross-validated evaluation on 20 public MOABB databases with identical folds. No input quantity is defined in terms of the output metric: the power means are computed by published fixed-point algorithms, the LDA weights are learned on training folds, and the AUC-ROC scores are measured on held-out data. The MDMF baseline is the authors' prior method, but it is a real algorithmic alternative implemented in pyRiemann, and the comparison against it is data-driven rather than forced by construction. The citations to the authors' previous power-mean work [19] and MDMF [20] are citations to published mathematical algorithms and do not reduce the present result to those papers. The use of six Brain Invaders P300 databases from the authors' group is a data-independence concern, not a circularity concern, because the data are public and the benchmark also includes external databases. Section 4 asserts without sensitivity analysis that the newly introduced hyperparameters are 'not critical and reasonable a priori choices give results that are statistically indistinguishable.' This is an unsupported robustness claim, and the paper itself reports that RPME significantly changes results (e.g., SMD = -0.297, p < 0.001 for P300), which could weaken the 'virtually hyperparameter-free' framing. However, an unsupported assertion is not a circular derivation: the paper does not define the hyperparameters in terms of the outcome, nor does it fit them to the test databases and then call the outcome a prediction. No quoted step exhibits an equation or fitted parameter that is equivalent to the reported result by construction, so the honest finding is no significant circularity.

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

The central result is empirical, so the main burden is carried by the correctness of the software pipeline and the representativeness of the chosen mean field. The paper contributes no new mathematical axioms; it imports power means, affine-invariant geometry, and MOABB/pyRiemann as given. The only ad hoc elements are the uniform hyperparameter values and the assumption that they are non-critical.

free parameters (4)
  • Power means field definition = h in {-1,-0.75,-0.5,-0.25,-0.1,0,0.1,0.25,0.5,0.75,1}
    Chosen by the authors as a representative sample of the [-1,1] interval; claimed to be non-critical but no sensitivity analysis is given.
  • ADCSP dimension thresholds = reduce to 28 if d>=28; reduce to 10 if d>=10
    Hand-chosen thresholds for the two-stage CSP spatial filter used on motor-imagery data.
  • RPME outlier parameters = I=4 iterations, zth=2.5
    Hand-chosen values for the robust power mean estimation; again no sensitivity analysis.
  • Number of spatial filters = 4 filters per class
    Fixed number of CSP/Xdawn filters retained; standard but still a choice.
assumptions (5)
  • standard math Power means of SPD matrices are well-defined and computable by the fixed-point algorithm of [19] for h in [-1,1], with the geometric mean as the limit h->0.
    Relied on in Section 2.1.2 to define the mean field over 11 values of h.
  • standard math The affine-invariant Fisher-Rao metric is the appropriate distance on the SPD manifold and the geometric mean minimizes the sum of squared distances.
    Used by MDM, MDMF, and MF for all distances; taken from refs [6,8].
  • domain assumption MOABB v1.1.1 and pyRiemann v0.7 implement the preprocessing, cross-validation, and classifiers identically for all pipelines without systematic bias.
    The entire benchmark depends on these open-source tools; the authors note they fixed a pyRiemann initialization bug.
  • ad hoc to paper The global outlier threshold zth=2.5 and the power-mean interval [-1,1] are valid uniformly across all 20 databases.
    Chosen a priori with no sensitivity analysis, yet applied to every dataset.
  • domain assumption LDA on the squared-distance feature vectors is well-posed without shrinkage or regularization.
    The paper states LDA requires no hyperparameter tuning, implicitly assuming the 22-dimensional within-class covariance is non-singular on all datasets.

how reviews work

0 comments
Cite this review

Pith. "Pith review of The Riemannian Means Field Classifier for EEG-Based BCI Data." pith.science (2026). https://pith.science/paper/QE2AR7RO

@misc{pith2026250417352,
  author       = {Pith},
  title        = {Pith review of: The Riemannian Means Field Classifier for EEG-Based BCI Data},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/QE2AR7RO}},
  note         = {Machine review of arXiv:2504.17352}
}
read the original abstract

A substantial amount of research has demonstrated the robustness and accuracy of the Riemannian minimum distance to mean (MDM) classifier for all kinds of EEG-based brain--computer interfaces (BCIs). This classifier is simple, fully deterministic, robust to noise, computationally efficient, and prone to transfer learning. Its training is very simple, requiring just the computation of a geometric mean of a symmetric positive-definite (SPD) matrix per class. We propose an improvement of the MDM involving a number of power means of SPD matrices instead of the sole geometric mean. By the analysis of 20 public databases, 10 for the motor-imagery BCI paradigm and 10 for the P300 BCI paradigm, comprising 587 individuals in total, we show that the proposed classifier clearly outperforms the MDM, approaching the state-of-the art in terms of performance while retaining the simplicity and the deterministic behavior. In order to promote reproducible research, our code will be released as open source.

Figures

Figures reproduced from arXiv: 2504.17352 by the authors.

Figure 1
Figure 1. ). Provided that this classifier does not require hyperparameter tuning, the MF approach also remains hyperparameter-free, except for the initial selection of the means field. Nevertheless, as long as the means field offers a representative set of power means sampled along the interval [−1, 1], this choice has minimal impact on overall performance [PITH_FULL_IMAGE:figures/full_fig_p003_1.png] view at source ↗
Figure 2
Figure 2. MI data: standardized mean differences (grey diamonds) with 95% confidence interval (grey horizontal lines) and p-values per database. The comparison concerns the MF with ADCSP and the MDM with ADCSP pipelines. A positive SMD value indicates that the algorithm on the right is better and the opposite if the value is negative. The closer the SMD is to zero, the lower the effect size. The p-values for the statistical t… view at source ↗
Figure 3
Figure 3. MI data: comparison of the MF and MDMF pipeline, both with ADCSP. See [PITH_FULL_IMAGE:figures/full_fig_p008_3.png] view at source ↗
Figures from the paper (7 more)
Figure 4
Figure 4. Figure 4: MI data: comparison of the MF with ADCSP and TS + LR with ACSTP pipeline. See [PITH_FULL_IMAGE:figures/full_fig_p009_4.png]
Figure 5
Figure 5. Figure 5: MI data: comparison of the MF pipeline, with ADCSP and the standard CSP. See [PITH_FULL_IMAGE:figures/full_fig_p009_5.png]
Figure 6
Figure 6. Figure 6: shows that MF with ADCSP and RPME is slightly better than MF with ADCSP only (SMD = −0.11, p < 0.01); thus, the robust power mean estimation does help in improving the robustness of the pipeline. However, this comes at the price of an increase in execution time ( [PIT…
Figure 7
Figure 7. Figure 7: P300 data: comparison of the proposed MF to the MDM, both with Xdawn. See [PITH_FULL_IMAGE:figures/full_fig_p011_7.png]
Figure 8
Figure 8. Figure 8: P300 data: comparison of the proposed MF to the MDMF pipelines, both with Xdawn. See [PITH_FULL_IMAGE:figures/full_fig_p012_8.png]
Figure 9
Figure 9. Figure 9: P300 data: comparison of the proposed MF to the golden-standard TS + LR pipelines, both with Xdawn. See [PITH_FULL_IMAGE:figures/full_fig_p012_9.png]
Figure 10
Figure 10. Figure 10: P300 data: comparison of the proposed MF pipeline with Xdawn, with and without RPME. See [PITH_FULL_IMAGE:figures/full_fig_p013_10.png]

Discussion (0). Continue with ORCID to comment.

Reference graph

Works this paper leans on

50 extracted references · 32 canonical work pages

  1. [1]

    The largest EEG- based BCI reproducibility study for open science: The MOABB benchmark

    Chevallier, S.; Carrara, I.; Aristimunha, B.; Guetschel, P.; Sedlar, S.; Lopes, B.; Velut, S.; Khazem, S.; Moreau, T. The largest EEG- based BCI reproducibility study for open science: The MOABB benchmark. arXiv 2024, arXiv:2404.15319. https://doi.org/10.48550/arXiv.2404.15319

    arXiv doi:10.48550/arxiv.2404.15319 2024

  2. [2]

    A review of classification algorithms for EEG-based brain–computer interfaces: A 10 year update

    Lotte, F.; Bougrain, L.; Cichocki, A.; Clerc, M.; Congedo, M.; Rakotomamonjy, A.; Yger, F. A review of classification algorithms for EEG-based brain–computer interfaces: A 10 year update. J. Neural Eng. 2018, 15, 031005. https://doi.org/10.1088/1741- 2552/aab2f2

    doi:10.1088/1741- 2018

  3. [3]

    Lezioni Di Calcolo Differenziale Assoluto; Nicola Zanichelli Editore: Bologna, Italy, 1925

    Levi-Civita, T. Lezioni Di Calcolo Differenziale Assoluto; Nicola Zanichelli Editore: Bologna, Italy, 1925

    work page 1925

  4. [4]

    Positive Definite Matrices; Princeton University Press: Princeton, NJ, USA, 2015

    Bhatia, R. Positive Definite Matrices; Princeton University Press: Princeton, NJ, USA, 2015

    work page 2015

  5. [5]

    Multiclass brain-computer interface classification by Riemannian geometry

    Barachant, A.; Bonnet, S.; Congedo, M.; Jutten, C. Multiclass brain-computer interface classification by Riemannian geometry. IEEE Trans. Bio-Med. Eng. 2012, 59, 920–928. https://doi.org/10.1109/TBME.2011.2172210

    work page arXiv 2012

  6. [6]

    EEG Source Analysis

    Congedo, M. EEG Source Analysis. Habilitation Thesis, Université de Grenoble, Saint-Martin-d'Hères, France, 2013. Available online: https://tel.archives-ouvertes.fr/tel-00880483 (accessed on 1 February 2025)

    work page 2013

  7. [7]

    Advanced Methods for BCI-EEG Processing for Improved Classification Performance and Reproducibility

    Carrara, I. Advanced Methods for BCI-EEG Processing for Improved Classification Performance and Reproducibility. Phdthesis, Université Côte d’Azur, 2024. Available online: https://inria.hal.science/tel-04797267 (accessed on 12 February 2025)

    work page 2024

  8. [8]

    Riemannian geometry for EEG-based brain-computer interfaces; a primer and a review

    Congedo, M.; Barachant, A.; Bhatia, R. Riemannian geometry for EEG-based brain-computer interfaces; a primer and a review. Brain-Comput. Interfaces 2017, 4, 155–174. https://doi.org/10.1080/2326263X.2017.1297192

    work page arXiv 2017

Show all 50 references
  1. [9]

    Riemannian Approaches in Brain-Computer Interfaces: A Review

    Yger, F.; Berar, M.; Lotte, F. Riemannian Approaches in Brain-Computer Interfaces: A Review. IEEE Trans. Neural Syst. Rehabil. Eng. 2017, 25, 1753–1762. https://doi.org/10.1109/TNSRE.2016.2627016

    2017

  2. [10]

    Single-trial classification of multi-user P300-based brain-computer interface using rie- mannian geometry

    Korczowski, L.; Congedo, M.; Jutten, C. Single-trial classification of multi-user P300-based brain-computer interface using rie- mannian geometry. In Proceedings of the 2015 37th Annual International Conference of the IEEE Engineering in Medicine and Biology Society (EMBC), Mi...

    2015

  3. [11]

    Mayaud, L.; Cabanilles, S.; Van Langhenhove, A.; Congedo, M.; Barachant, A.; Pouplin, S.; Filipe, S.; Pétégnief, L.; Rochecouste, O.; Azabou, E.; et al. Brain-computer interface for the communication of acute patients: A feasibility study and a randomized controlled trial comp...

    2016

  4. [12]

    Brain Invaders Cali- bration-less P300-Based BCI Using Dry EEG Electrodes Dataset (bi2014a); GIPSA -lab, Research Report, juillet; 2019

    Korczowski, L.; Ostaschenko, E.; Andreev, A.; Cattan, G.; Rodrigues, P.L.C.; Gautheret, V.; Congedo, M. Brain Invaders Cali- bration-less P300-Based BCI Using Dry EEG Electrodes Dataset (bi2014a); GIPSA -lab, Research Report, juillet; 2019. Available online: https://hal.scienc...

    2019

  5. [13]

    Recreational Applications of OpenViBE: Brain Invaders and Use -the-Force; chap

    Andreev, A.; Barachant, A.; Lotte, F.; Congedo, M. Recreational Applications of OpenViBE: Brain Invaders and Use -the-Force; chap. 14; John Wiley, Sons: Hoboken, NJ, USA, 2016. Available online: https://hal.archives-ouvertes.fr/hal-01366873/document (ac- cessed on 1 February 2025)

    2016

  6. [14]

    A Differential Geometric Approach to the Geometric Mean of Symmetric Positive-Definite Matrices

    Moakher, M. A Differential Geometric Approach to the Geometric Mean of Symmetric Positive-Definite Matrices. SIAM J. Ma- trix Anal. Appl. 2005, 26, 735–747. https://doi.org/10.1137/S0895479803436937

    2005 doi

  7. [15]

    Matrix power means and the Karcher mean

    Lim, Y.; Pálfia, M. Matrix power means and the Karcher mean. J. Funct. Anal. 2012, 262, 1498 –1514. https://doi.org/10.1016/j.jfa.2011.11.012

    2012 doi

  8. [16]

    Operator means of probability measures and generalized Karcher equations

    Pálfia, M. Operator means of probability measures and generalized Karcher equations. arXiv 2016, arXiv:1601.06777. https://doi.org/10.48550/arXiv.1601.06777

    2016 doi arXiv

  9. [17]

    Weighted means and Karcher equations of positive operators

    Lawson, J.; Lim, Y. Weighted means and Karcher equations of positive operators. Proc. Natl. Acad. Sci. USA 2014, 110, 15626– 15632. https://doi.org/10.1073/pnas.1313640110

    2014 doi

  10. [18]

    Karcher means and Karcher equations of positive definite operators

    Lawson, J.; Lim, Y. Karcher means and Karcher equations of positive definite operators. Trans. Amer. Math. Soc. Ser. B 2014, 1, 1–22. https://doi.org/10.1090/S2330-0000-2014-00003-4

    2014 doi

  11. [19]

    Fixed Point Algorithms for Estimating Power Means of Positive Definite Matrices

    Congedo, M.; Barachant, A.; Bhatia, R. Fixed Point Algorithms for Estimating Power Means of Positive Definite Matrices. IEEE Trans. Signal Process. 2017, 65, 2211–2220. https://doi.org/10.1109/TSP.2017.2649483

    2017

  12. [20]

    The Riemannian Minimum Distance to Means Field Classifier

    Congedo, M.; Rodrigues, P.L.C.; Jutten, C. The Riemannian Minimum Distance to Means Field Classifier. In Proceedings of the BCI 2019 —8th International Brain -Computer Interface Conference, Graz, Austria, 16–20 September 2019. https://doi.org/10.3217/978-3-85125-682-6-02 16 of 17

    2019 doi

  13. [21]

    Mother of All BCI Benchmarks, version v1.0.0; 23 October 2023; Zenodo: Genève, Switzerland

    Bruno, A.; Igor, C.; Pierre, G.; Sara, S.; Pedro, R.; Jan, S.; Divyesh, N.; Erik, B.; Barthelemy, Q.; Tibor, S.R.; et al. Mother of All BCI Benchmarks, version v1.0.0; 23 October 2023; Zenodo: Genève, Switzerland. https://doi.org/10.5281/zenodo.10034224

    2023 doi

  14. [22]

    MOABB: Trustworthy algorithm benchmarking for BCIs

    Jayaram, V.; Barachant, A. MOABB: Trustworthy algorithm benchmarking for BCIs. J. Neural Eng. 2018, 15, 066011. https://doi.org/10.1088/1741-2552/aadea0

    2018 doi

  15. [23]

    -R.; Gramfort, A.; Chevallier, S.; Rodrigues, P.L.C.; Olivetti, E.; Goncharenko, V.; Berg, G.W.v.; Reguig, G.; et al

    Barachant, A.; Barthélemy, Q.; King, J. -R.; Gramfort, A.; Chevallier, S.; Rodrigues, P.L.C.; Olivetti, E.; Goncharenko, V.; Berg, G.W.v.; Reguig, G.; et al. pyRiemann, version v0.8; Zenodo: Genève, Switzerland, 2024. https://doi.org/10.5281/zenodo.593816

    2024 doi

  16. [24]

    Optimizing Spatial filters for Robust EEG Single-Trial Analysis

    Blankertz, B.; Tomioka, R.; Lemm, S.; Kawanabe, M.; Muller, K. Optimizing Spatial filters for Robust EEG Single-Trial Analysis. IEEE Signal Process. Mag. 2008, 25, 41–56. https://ieeexplore.ieee.org/document/4408441

    2008

  17. [25]

    Xdawn algorithm to enhance evoked potentials: Application to brain-computer interface

    Rivet, B.; Souloumiac, A.; Attina, V.; Gibert, G. Xdawn algorithm to enhance evoked potentials: Application to brain-computer interface. IEEE Trans. Biomed. Eng. 2009, 56, 2035–2043. https://ieeexplore.ieee.org/document/4760273

    2009

  18. [26]

    Joint Approximate Diagonalization of Positive Definite Hermitian Matrices

    Pham, T. Joint Approximate Diagonalization of Positive Definite Hermitian Matrices. SIAM J. Matrix Anal. Appl. 2001, 22, 1136–

    2001

  19. [27]

    The Riemannian Potato Field: A Tool for Online Signal Quality Index of EEG

    Barthélemy, Q.; Mayaud, L.; Ojeda, D.; Congedo, M. The Riemannian Potato Field: A Tool for Online Signal Quality Index of EEG. IEEE Trans. Neural Syst. Rehabil. Eng. 2019, 27, 244. https://doi.org/10.1109/TNSRE.2019.2893113

    2019

  20. [28]

    Review of the BCI Competition IV

    Tangermann, M.; Müller, K.R.; Aertsen, A.; Birbaumer, N.; Braun, C.; Brunner, C.; Leeb, R.; Mehring, C.; Miller, K.J.; Müller - Putz, G.R.; et al. Review of the BCI Competition IV. Front. Neurosci. 2012, 6, 55. https://doi.org/10.3389/fnins.2012.00055

    2012 arXiv

  21. [29]

    Brain -computer communication: Motivation, aim, and impact of exploring a virtual apartment

    Leeb, R.; Lee, F.; Keinrath, C.; Scherer, R.; Bischof, H.; Pfurtscheller, G. Brain -computer communication: Motivation, aim, and impact of exploring a virtual apartment. IEEE Trans. Neural. Syst. Rehabil. Eng . 2007, 15, 473 –482. https://doi.org/10.1109/TNSRE.2007.906956

    2007

  22. [30]

    EEG datasets for motor imagery brain-computer interface

    Cho, H.; Ahn, M.; Ahn, S.; Kwon, M.; Jun, S.C. EEG datasets for motor imagery brain-computer interface. Gigascience 2017, 6, gix034. https://doi.org/10.1093/gigascience/gix034

    2017 doi

  23. [31]

    Beamforming in noninvasive brain-computer interfaces

    Grosse-Wentrup, M.; Liefhold, C.; Gramann, K.; Buss, M. Beamforming in noninvasive brain-computer interfaces. IEEE Trans. Biomed. Eng. 2009, 56, 1209–1219. https://doi.org/10.1109/TBME.2008.2009768

    2009

  24. [32]

    EEG dataset and OpenBMI toolbox for three BCI paradigms: An investigation into BCI illiteracy

    Lee, M.H.; Kwon, O.Y.; Kim, Y.J.; Kim, H.K.; Lee, Y.E.; Williamson, J.; Fazli, S.; Lee, S.W. EEG dataset and OpenBMI toolbox for three BCI paradigms: An investigation into BCI illiteracy. Gigascience 2019, 8, giz002. https://doi.org/10.1093/gigascience/giz002

    2019 doi

  25. [33]

    BCI2000: A general-purpose brain-computer interface (BCI) system

    Schalk, G.; McFarland, D.J.; Hinterberger, T.; Birbaumer, N.; Wolpaw, J.R. BCI2000: A general-purpose brain-computer interface (BCI) system. IEEE Trans. Biomed. Eng. 2004, 51, 1034–1043. https://doi.org/10.1109/TBME.2004.827072

    2004

  26. [34]

    Deep learning with convolutional neural networks for EEG decoding and visualization

    Schirrmeister, R.T.; Springenberg, J.T.; Fiederer, L.D.J.; Glasstetter, M.; Eggensperger, K.; Tangermann, M.; Hutter, F.; Burgard, W.; Ball, T. Deep learning with convolutional neural networks for EEG decoding and visualization. Hum. Brain Mapp. 2017, 38, 5391–5420. https://do...

    2017 doi

  27. [35]

    Open Access Dataset for EEG+NIRS Single-Trial Classification

    Shin, J.; von Lühmann, A.; Blankertz, B.; Kim, D.W.; Jeong, J.; Hwang, H.J.; Müller, K.R. Open Access Dataset for EEG+NIRS Single-Trial Classification. IEEE Trans. Neural . Syst. Rehabil. Eng . 2017, 25, 1735 –1745. https://doi.org/10.1109/TNSRE.2016.2628057

    2017

  28. [36]

    Evaluation of EEG oscillatory patterns and cognitive process during simple and compound limb motor imagery

    Yi, W.; Qiu, S.; Wang, K.; Qi, H.; Zhang, L.; Zhou, P.; He, F.; Ming, D. Evaluation of EEG oscillatory patterns and cognitive process during simple and compound limb motor imagery. PLoS ONE 2014, 9, e114853. https://doi.org/10.1371/jour- nal.pone.0114853

    2014 doi

  29. [37]

    A Fully Automated Trial Selection Method for Optimization of Motor Imagery Based Brain-Computer Interface

    Zhou, B.; Wu, X.; Lv, Z.; Zhang, L.; Guo, X. A Fully Automated Trial Selection Method for Optimization of Motor Imagery Based Brain-Computer Interface. PLoS ONE 2016, 11, e0162657. https://doi.org/10.1371/journal.pone.0162657

    2016 doi

  30. [38]

    Attention and P300- based BCI performance in people with amyotrophic lateral sclerosis

    Riccio, A.; Simione, L.; Schettini, F.; Pizzimenti, A.; Inghilleri, M.; Belardinelli, M.O.; Mattia, D.; Cincotti, F. Attention and P300- based BCI performance in people with amyotrophic lateral sclerosis. Front. Hum. Neurosci. 2013, 7, 732. https://doi.org/10.3389/fnhum.2013.00732

    2013

  31. [39]

    Influence of P300 latency jitter on event related potential- based brain-computer interface performance

    Aricò, P.; Aloise, F.; Schettini, F.; Salinari, S.; Mattia, D.; Cincotti, F. Influence of P300 latency jitter on event related potential- based brain-computer interface performance. J. Neural Eng. 2014, 11, 035008. https://doi.org/10.1088/1741-2560/11/3/035008

    2014 doi

  32. [40]

    How many people are able to control a P300-based brain-computer interface (BCI)? Neurosci

    Guger, C.; Daban, S.; Sellers, E.; Holzner, C.; Krausz, G.; Carabalona, R.; Gramatica, F.; Edlinger, G. How many people are able to control a P300-based brain-computer interface (BCI)? Neurosci. Lett . 2009, 462, 94 –98. https://doi.org/10.1016/j.neulet.2009.06.045

    2009 doi

  33. [41]

    Building Brain Invaders: EEG Data of An Experimental Validation

    Van Veen, G.F.P.; Barachant, A.; Andreev, A.; Cattan, G.; Rodrigues, P.L.C.; Congedo, M. Building Brain Invaders: EEG Data of An Experimental Validation. GIPSA-Lab, Research Report 1, mai 2019. Available online: https://hal.archives-ouvertes.fr/hal- 02126068 (accessed on 1 Feb...

    2019

  34. [42]

    Brain Invaders Adaptive Versus Non-Adap- tive P300 Brain-Computer Interface Dataset

    Vaineau, E.; Barachant, A.; Andreev, A.; Rodrigues, P.C.; Cattan, G.; Congedo, M. Brain Invaders Adaptive Versus Non-Adap- tive P300 Brain-Computer Interface Dataset. GIPSA-LAB, Research Report 1, avril 2019. Available online: https://hal.archives- ouvertes.fr/hal-02103098 (ac...

    2019

  35. [43]

    L.C.; Gautheret, V.; Congedo, M

    Korczowski, L.; Ostaschenko, E.; Andreev, A.; Cattan, G.; Rodrigues, P. L.C.; Gautheret, V.; Congedo, M. Brain Invaders Solo Versus Collaboration: Multi-User P300-Based Brain-Computer Interface Dataset (bi2014b); GIPSA-lab, Research Report, juillet

  36. [44]

    Brain In- vaders Calibrationless P300-Based BCI with Modulation of Flash Duration Dataset (bi2015a); GIPSA-lab, Research Report, juillet; 2019

    Korczowski, L.; Cederhout, M.; Andreev, A.; Cattan, G.; Rodrigues, P.L.C.; Gautheret, V.; Congedo, M. Brain In- vaders Calibrationless P300-Based BCI with Modulation of Flash Duration Dataset (bi2015a); GIPSA-lab, Research Report, juillet; 2019. Available online: https://hal.s...

    2019

  37. [45]

    Brain Invaders Coopera- tive Versus Competitive: Multi-User P300-based Brain-Computer Interface Dataset (bi2015b); GIPSA-lab, Research Report, juil- let; 2019

    Korczowski, L.; Cederhout, M.; Andreev, A.; Cattan, G.; Rodrigues, P.L.; Gautheret , V.; Congedo, M. Brain Invaders Coopera- tive Versus Competitive: Multi-User P300-based Brain-Computer Interface Dataset (bi2015b); GIPSA-lab, Research Report, juil- let; 2019. Available online...

    2019

  38. [46]

    Randomization Tests, 3rd ed.; Marcel Dekker: New York, NY, USA, 1995

    Edgington, E.S. Randomization Tests, 3rd ed.; Marcel Dekker: New York, NY, USA, 1995

    1995

  39. [47]

    Multivariate Permutation Tests: With Applications in Biostatistics

    Huson, L.W. Multivariate Permutation Tests: With Applications in Biostatistics. J. R. Stat. Soc. Ser. D Stat. 2003, 52, 247 . https://doi.org/10.1111/1467-9884.t01-6-00356

    2003 doi

  40. [48]

    Approximate Joint Diagonalization and Geometric Mean of Symmetric Positive Definite Matrices

    Congedo, M.; Afsari, B.; Barach ant, A.; Moakher, M. Approximate Joint Diagonalization and Geometric Mean of Symmetric Positive Definite Matrices. PLoS ONE 2015, 10, e0121423. https://doi.org/10.1371/journal.pone.0121423

    2015 doi

  41. [1152]

    https://doi.org/10.1137/S089547980035689X

    doi

  42. [2019]

    Available online: https://hal.science/hal-02173958/ (accessed on 1 February 2025)

    2025

Pith tools

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