pith. sign in

arxiv: 2604.27033 · v2 · submitted 2026-04-29 · 💻 cs.LG · eess.SP

Cross-Subject Generalization for EEG Decoding: A Survey of Deep Learning Methods

Taida Li , Yujun Yan , Fei Dou , Wenzhan Song , Xiang Zhang This is my paper

Pith reviewed 2026-05-07 13:23 UTC · model grok-4.3

classification 💻 cs.LG eess.SP
keywords cross-subject EEGdomain generalizationdeep learningbrain-computer interfacesEEG decodingsurveyadversarial learning
0
0 comments X

The pith

A survey organizes deep learning methods for EEG decoding into four families to handle differences between training and new subjects.

A machine-rendered reading of the paper's core claim, the machinery that carries it, and where it could break.

The paper reviews deep learning techniques for decoding EEG signals when the test subject differs from those used in training. Inter-subject variability in brain waves creates a domain shift that reduces model accuracy in real applications such as brain-computer interfaces. The survey frames the setting as a multi-source domain problem and requires subject-independent evaluation protocols for fair assessment. It groups existing methods into four families and closes by discussing limits of current approaches, the information carried by subject identity, and the rise of EEG foundation models.

Core claim

This survey formalizes cross-subject EEG decoding as a multi-source domain generalization problem and supplies a taxonomy that divides the literature into feature alignment, adversarial learning, feature disentanglement, and contrastive learning families.

What carries the argument

The four-family taxonomy of methods (feature alignment, adversarial learning, feature disentanglement, contrastive learning) that groups techniques for reducing domain shift caused by inter-subject EEG variability.

If this is right

  • Researchers gain a structured way to compare and choose methods from the four families rather than treating approaches in isolation.
  • Papers must adopt subject-independent protocols to claim valid cross-subject results.
  • Future work should address the theoretical limitations identified for each family.
  • Explicit modeling of subject identity can be tested as an additional signal source.
  • Pre-trained EEG foundation models become a concrete next direction for reducing per-subject data needs.

Where Pith is reading between the lines

These are editorial extensions of the paper, not claims the author makes directly.

  • The taxonomy could serve as a template for surveys on other variable biosignals such as EMG or ECG.
  • Hybrid methods that combine elements from multiple families may be tested for additive gains in generalization.
  • Practical BCI devices could move toward calibration-free use if one family proves consistently superior under real-world conditions.

Load-bearing premise

The taxonomy fully covers all relevant deep learning methods and that subject-independent evaluation protocols constitute the necessary standard for assessing real generalization.

What would settle it

Publication of a cross-subject EEG method that cannot be placed in any of the four families while still improving performance under subject-independent protocols would falsify the taxonomy's completeness.

Figures

Figures reproduced from arXiv: 2604.27033 by Fei Dou, Taida Li, Wenzhan Song, Xiang Zhang, Yujun Yan.

Figure 1
Figure 1. Figure 1: Deep Learning Pipeline for EEG Decoding. view at source ↗
Figure 2
Figure 2. Figure 2: Problem formulation and task taxonomy for cross-subject EEG decoding. Panel (a) formal view at source ↗
Figure 3
Figure 3. Figure 3: Overview of methodology distributions. 3 Methodological Taxonomy To address the challenge of cross-subject generalization in EEG decoding, the research community has developed diverse methodologies, from feature alignment to causal representation learning. This section provides a taxonomy of these approaches, categorized by their underlying strategies for mitigating inter-subject variability view at source ↗
Figure 4
Figure 4. Figure 4: Domain alignment and subject-level contrast. view at source ↗
Figure 5
Figure 5. Figure 5: Domain Adversarial Neural Network. feature representation into distinct, isolated components. For the cross-subject problem, the goal is to learn a representation that is explicitly disentangled into a task-relevant (and subject-invariant) component and an identity (and subject-specific) component. The downstream classifier is then trained using only the task-relevant component, discarding the subject-spec… view at source ↗
read the original abstract

Deep learning for cross-subject EEG decoding is hindered by high inter-subject variability, which introduces a severe domain shift between training and unseen test subjects. This survey presents a comprehensive review of deep learning methodologies specifically engineered to address this cross-subject generalization challenge. To ground this analysis, we formalize the cross-subject setting as a multi-source domain problem and delineate the rigorous, subject-independent evaluation protocols required for valid assessment. Central to this survey is a systematic taxonomy of the current literature into discrete methodological families, including feature alignment, adversarial learning, feature disentanglement, and contrastive learning. We conclude by examining three critical elements for advancing robust, real-world decoding: the theoretical limitations of current methodologies, the structural value of subject identity, and the emergence of EEG foundation models.

Editorial analysis

A structured set of objections, weighed in public.

Desk editor's note, referee report, simulated authors' rebuttal, and a circularity audit. Tearing a paper down is the easy half of reading it; the pith above is the substance, this is the friction.

Referee Report

1 major / 2 minor

Summary. The manuscript surveys deep learning methods for cross-subject EEG decoding, which is challenged by high inter-subject variability and domain shift. It formalizes the problem as a multi-source domain adaptation task, specifies subject-independent evaluation protocols, and organizes the literature via a taxonomy into methodological families including feature alignment, adversarial learning, feature disentanglement, and contrastive learning. The paper concludes by addressing theoretical limitations of current approaches, the structural role of subject identity, and the emergence of EEG foundation models.

Significance. If the taxonomy is comprehensive and the literature coverage balanced, the survey would provide a useful organizing framework for a growing subfield of brain-computer interfaces. The formalization of multi-source protocols and the explicit discussion of open challenges (theoretical bounds, subject identity, foundation models) add value beyond a simple list of papers, potentially guiding standardized evaluations and future research directions.

major comments (1)
  1. The taxonomy (central to the survey) groups methods into four families, but the manuscript does not explicitly address how hybrid approaches (e.g., adversarial training combined with contrastive losses) are classified or whether the categories are intended to be exhaustive and mutually exclusive. This risks arbitrary groupings and should be clarified with decision criteria or examples from the reviewed literature.
minor comments (2)
  1. The abstract states that three critical elements are examined in the conclusion, but the conclusion section would benefit from a more structured subsectioning or bullet-point summary for each element to improve readability.
  2. The problem formalization section would be strengthened by including a small table comparing subject-independent protocols across the cited works, rather than describing them only in prose.

Simulated Author's Rebuttal

1 responses · 0 unresolved

We thank the referee for the positive evaluation of the survey and for highlighting an important point regarding the taxonomy. We address the concern in detail below and will incorporate clarifications into the revised manuscript.

read point-by-point responses

  1. Referee: The taxonomy (central to the survey) groups methods into four families, but the manuscript does not explicitly address how hybrid approaches (e.g., adversarial training combined with contrastive losses) are classified or whether the categories are intended to be exhaustive and mutually exclusive. This risks arbitrary groupings and should be clarified with decision criteria or examples from the reviewed literature.

    Authors: We agree that explicit guidance on classification is necessary. The four families are organized according to the primary mechanism used to mitigate inter-subject domain shift, as described in the introduction to the taxonomy section. Categories are not intended to be mutually exclusive; a method is assigned to the family corresponding to its core technical contribution (e.g., the loss term or architectural component that directly targets subject-invariant representations). Hybrid approaches are classified by their dominant innovation: a method that primarily employs an adversarial discriminator to align distributions while adding a contrastive term for representation quality would be placed under adversarial learning, with the contrastive component noted as a complementary technique. We will add a dedicated paragraph in the taxonomy overview that states these decision criteria explicitly, supplies two or three concrete examples of hybrid papers from the surveyed literature together with their assigned categories, and acknowledges that some works could reasonably be viewed from multiple angles. This addition will make the grouping transparent without altering the overall structure. revision: yes

Circularity Check

0 steps flagged

No significant circularity identified

full rationale

This is a literature survey paper that organizes existing deep learning methods for cross-subject EEG decoding into a taxonomy (feature alignment, adversarial learning, etc.) and frames the problem as multi-source domain adaptation with subject-independent protocols. No mathematical derivations, equations, predictions, or fitted parameters appear in the provided text or abstract. All substantive claims rest on citations to external prior work rather than internal reductions or self-referential definitions. The taxonomy is presented as a descriptive organizational tool, not a derived or uniqueness-forced result. Self-citations, if present, are not load-bearing for any central claim, and the paper explicitly flags open issues without asserting completeness as a testable property. This meets the criteria for a self-contained descriptive review with no circular steps.

Axiom & Free-Parameter Ledger

0 free parameters · 0 axioms · 0 invented entities

The survey's central claims rest on the representativeness of the reviewed literature and the utility of the proposed taxonomy; no free parameters, axioms, or invented entities are introduced beyond standard domain-adaptation concepts.

pith-pipeline@v0.9.0 · 5434 in / 1031 out tokens · 58386 ms · 2026-05-07T13:23:42.351990+00:00 · methodology

discussion (0)

Sign in with ORCID, Apple, or X to comment. Anyone can read and Pith papers without signing in.

Reference graph

Works this paper leans on

101 extracted references · 4 canonical work pages

  1. [1]

    Deep learning for electroencephalogram (eeg) classification tasks: a review,

    A. Craik, Y. He, and J. L. Contreras-Vidal, “Deep learning for electroencephalogram (eeg) classification tasks: a review,”Journal of Neural Engineering, vol. 16, p. 031001, apr 2019

    2019

  2. [2]

    Deep convolutional neural network for the automated detection and diagnosis of seizure using eeg signals,

    U. R. Acharya, S. L. Oh, Y. Hagiwara, J. H. Tan, and H. Adeli, “Deep convolutional neural network for the automated detection and diagnosis of seizure using eeg signals,”Computers in Biology and Medicine, vol. 100, pp. 270–278, 2018

    2018

  3. [3]

    Epileptic seizures detection using deep learning techniques: A review,

    H. u. R. Khan, M. Hussain, I. Ullah, and M. Lee, “Epileptic seizures detection using deep learning techniques: A review,”International Journal of Environmental Research and Public Health, vol. 18, p. 5780, May 2021

    2021

  4. [4]

    Investigating critical frequency bands and channels for eeg-based emotion recognition with deep neural networks,

    W.-L. Zheng and B.-L. Lu, “Investigating critical frequency bands and channels for eeg-based emotion recognition with deep neural networks,”IEEE Transactions on Autonomous Mental Development, vol. 7, no. 3, pp. 162–175, 2015

    2015

  5. [5]

    Deep learning-based eeg emotion recognition: Current trends and future perspectives,

    S. Wagh, G. Gore, S. Ghule, S. Patil, M. Bharti, A. Shukla, and S. Kadam, “Deep learning-based eeg emotion recognition: Current trends and future perspectives,”Frontiers in Psychology, vol. 14, p. 1126994, 2023

    2023

  6. [6]

    Eegnet: a compact convolutional neural network for eeg-based brain–computer interfaces,

    V. J. Lawhern, A. J. Solon, N. R. Waytowich, S. M. Gordon, C. P. Hung, and B. J. Lance, “Eegnet: a compact convolutional neural network for eeg-based brain–computer interfaces,” Journal of Neural Engineering, vol. 15, p. 056013, jul 2018

    2018

  7. [7]

    Inter- and intra-individual variability in alpha peak frequency,

    S. Haegens, H. Cousijn, G. Wallis, N. A. Harrison, and A. C. Nobre, “Inter- and intra-individual variability in alpha peak frequency,”NeuroImage, vol. 92, pp. 46–55, May 2014

    2014

  8. [8]

    Intra- and inter-subject variability in eeg-based sensorimotor brain computer interface: A review,

    S. Saha, K. A. Mamun, K. I. U. Ahmed, S. Mostafa, G. R. Naik, S. Darvishi, and A. H. Khandoker, “Intra- and inter-subject variability in eeg-based sensorimotor brain computer interface: A review,”Frontiers in Computational Neuroscience, vol. 13, p. 87, 2019

    2019

  9. [9]

    Self-supervised contrastive pre-training for time series via time-frequency consistency,

    X. Zhang, Z. Zhao, T. Tsiligkaridis, and M. Zitnik, “Self-supervised contrastive pre-training for time series via time-frequency consistency,” inAdvances in Neural Information Processing Systems (NeurIPS), vol. 35, pp. 3988–4003, 2022

    2022

  10. [10]

    Medformer: A multi-granularity patching transformer for medical time-series classification,

    Y. Wang, N. Huang, T. Li, Y. Yan, and X. Zhang, “Medformer: A multi-granularity patching transformer for medical time-series classification,”Advances in Neural Information Processing Systems, 2024

    2024

  11. [11]

    Medgnn: Towards multi-resolution spatiotemporal graph learning for medical time series classification,

    W. Fan, J. Fei, D. Guo, K. Yi, X. Song, H. Xiang, H. Ye, and M. Li, “Medgnn: Towards multi-resolution spatiotemporal graph learning for medical time series classification,” 2025

    2025

  12. [12]

    Toward cross-subject and cross-session generalization in eeg-based emotion recognition: Systematic review, taxonomy, and methods,

    A. Apicella, P. Arpaia, G. D’Errico, D. Marocco, G. Mastrati, N. Moccaldi, and R. Prevete, “Toward cross-subject and cross-session generalization in eeg-based emotion recognition: Systematic review, taxonomy, and methods,”Neurocomputing, vol. 604, p. 128354, 2024. 16 IOP PublishingJournalvv(yyyy) aaaaaa Authoret al

    2024

  13. [13]

    A systematic review of cross-patient approaches for eeg epileptic seizure prediction,

    S. Shafiezadeh, G. Marco Duma, M. Pozza, and A. Testolin, “A systematic review of cross-patient approaches for eeg epileptic seizure prediction,”Journal of Neural Engineering, vol. 21, no. 6, p. 061004, 2024

    2024

  14. [14]

    Learning invariant representations from eeg via adversarial inference,

    O. ¨Ozdenizci, Y. Wang, T. Koike-Akino, and D. Erdo˘ gmu¸ s, “Learning invariant representations from eeg via adversarial inference,”IEEE Access, vol. 8, pp. 27074–27085, 2020

    2020

  15. [15]

    Cross-patient automatic epileptic seizure detection using patient-adversarial neural networks with spatio-temporal eeg augmentation,

    Z. Zhang, T. Ji, M. Xiao, W. Wang, G. Yu, T. Lin, Y. Jiang, X. Zhou, and Z. Lin, “Cross-patient automatic epileptic seizure detection using patient-adversarial neural networks with spatio-temporal eeg augmentation,”Biomedical Signal Processing and Control, vol. 89, p. 105664, 2024

    2024

  16. [16]

    Contrastive learning of subject-invariant EEG representations for cross-subject emotion recognition,

    X. Shen, X. Liu, X. Hu, D. Zhang, and S. Song, “Contrastive learning of subject-invariant EEG representations for cross-subject emotion recognition,”IEEE Transactions on Affective Computing, vol. 14, no. 3, pp. 2496–2511, 2023

    2023

  17. [17]

    Subject independent EEG-based BCI decoding,

    S. Fazli, C. Grozea, M. Danoczy, B. Blankertz, F. Popescu, and K.-R. M¨ uller, “Subject independent EEG-based BCI decoding,” inAdvances in Neural Information Processing Systems (NIPS), vol. 22, pp. 513–521, 2009

    2009

  18. [18]

    How to evaluate your medical time series classification?,

    Y. Wang, T. Li, Y. Yan, W. Song, and X. Zhang, “How to evaluate your medical time series classification?,” 2024

    2024

  19. [19]

    Data leakage in deep learning studies of translational EEG,

    G. Brookshire, J. Kasper, N. M. Blauch, Y. C. Wu, R. M. Glatt, D. A. Merrill, S. Gerrol, K. J. Yoder, and C. Lucero, “Data leakage in deep learning studies of translational EEG,” Frontiers in Neuroscience, vol. 18, p. 1373515, 2024

    2024

  20. [20]

    Bci2000: a general-purpose brain-computer interface (bci) system,

    G. Schalk, D. J. McFarland, T. Hinterberger, N. Birbaumer, and J. R. Wolpaw, “Bci2000: a general-purpose brain-computer interface (bci) system,”IEEE Transactions on biomedical engineering, vol. 51, no. 6, pp. 1034–1043, 2004

    2004

  21. [21]

    Cross-dataset variability problem in EEG decoding with deep learning,

    L. Xu, M. Xu, Y. Ke, X. An, S. Liu, and D. Ming, “Cross-dataset variability problem in EEG decoding with deep learning,”Frontiers in Human Neuroscience, vol. 14, p. 103, 2020

    2020

  22. [22]

    Bci competition 2008–graz data set b,

    R. Leeb, C. Brunner, G. M¨ uller-Putz, A. Schl¨ ogl, and G. Pfurtscheller, “Bci competition 2008–graz data set b,”Graz University of Technology, Austria, vol. 16, pp. 1–6, 2008

    2008

  23. [23]

    Global adaptive transformer for cross-subject enhanced EEG classification,

    Y. Song, Q. Zheng, Q. Wang, X. Gao, and P.-A. Heng, “Global adaptive transformer for cross-subject enhanced EEG classification,”IEEE Transactions on Neural Systems and Rehabilitation Engineering, vol. 31, pp. 2767–2777, 2023

    2023

  24. [24]

    Eeg datasets for motor imagery brain–computer interface,

    H. Cho, M. Ahn, S. Ahn, M. Kwon, and S. C. Jun, “Eeg datasets for motor imagery brain–computer interface,”GigaScience, vol. 6, no. 7, p. gix034, 2017

    2017

  25. [25]

    Eeg dataset and openbmi toolbox for three bci paradigms: An investigation into bci illiteracy,

    M.-H. Lee, O.-Y. Kwon, Y.-J. Kim, H.-K. Kim, Y.-E. Lee, J. Williamson, S. Fazli, and S.-W. Lee, “Eeg dataset and openbmi toolbox for three bci paradigms: An investigation into bci illiteracy,”GigaScience, vol. 8, no. 5, p. giz002, 2019

    2019

  26. [26]

    Cross-subject motor imagery electroencephalogram decoding with domain generalization,

    Y. Zheng, S. Wu, J. Chen, Q. Yao, and S. Zheng, “Cross-subject motor imagery electroencephalogram decoding with domain generalization,”Bioengineering, vol. 12, no. 5, p. 495, 2025

    2025

  27. [27]

    SPD domain-specific batch normalization to crack interpretable unsupervised domain adaptation in EEG,

    R. J. Kobler, J.-i. Hirayama, Q. Zhao, and M. Kawanabe, “SPD domain-specific batch normalization to crack interpretable unsupervised domain adaptation in EEG,” inAdvances in Neural Information Processing Systems (NeurIPS), vol. 35, pp. 10664–10677, 2022

    2022

  28. [28]

    Subject-invariant deep neural networks based on baseline correction for EEG motor imagery BCI,

    Y. Kwak, K. Kong, W. J. Song, and S.-E. Kim, “Subject-invariant deep neural networks based on baseline correction for EEG motor imagery BCI,”IEEE Journal of Biomedical and Health Informatics, vol. 27, no. 4, pp. 1801–1812, 2023

    2023

  29. [29]

    Subject-independent meta-learning framework towards optimal training of EEG-based classifiers,

    H. W. Ng and C. Guan, “Subject-independent meta-learning framework towards optimal training of EEG-based classifiers,”Neural Networks, vol. 172, p. 106108, 2024

    2024

  30. [30]

    Bci competition 2008–graz data set a,

    C. Brunner, R. Leeb, G. M¨ uller-Putz, A. Schl¨ ogl, and G. Pfurtscheller, “Bci competition 2008–graz data set a,”Institute for knowledge discovery (laboratory of brain-computer interfaces), Graz University of Technology, vol. 16, no. 1-6, p. 34, 2008. 17 IOP PublishingJournalvv(yyyy) aaaaaa Authoret al

    2008

  31. [31]

    Review of the bci competition iv,

    M. Tangermann, K.-R. M¨ uller, A. Aertsen, N. Birbaumer, C. Braun, C. Brunner, R. Leeb, C. Mehring, K. J. Miller, G. R. M¨ uller-Putz,et al., “Review of the bci competition iv,” Frontiers in neuroscience, vol. 6, p. 55, 2012

    2012

  32. [32]

    Cross-subject EEG signal recognition using deep domain adaptation network,

    W. Hang, W. Feng, R. Du, S. Liang, Y. Chen, Q. Wang, and X. Liu, “Cross-subject EEG signal recognition using deep domain adaptation network,”IEEE Access, vol. 7, pp. 128273–128282, 2019

    2019

  33. [33]

    Deep source semi-supervised transfer learning (DS3TL) for cross-subject EEG classification,

    X. Jiang, L. Meng, Z. Wang, and D. Wu, “Deep source semi-supervised transfer learning (DS3TL) for cross-subject EEG classification,”IEEE Transactions on Biomedical Engineering, vol. 71, no. 4, pp. 1308–1318, 2024

    2024

  34. [34]

    Classification of motor imagery electroencephalogram signals by using adaptive cross-subject transfer learning,

    J. Feng, Y. Li, C. Jiang, Y. Liu, M. Li, and Q. Hu, “Classification of motor imagery electroencephalogram signals by using adaptive cross-subject transfer learning,”Frontiers in Human Neuroscience, vol. 16, p. 1068165, 2022

    2022

  35. [35]

    Subject-independent EEG classification based on a hybrid neural network,

    H. Zhang, H. Ji, H. Yu, J. Li, Y. Jin, G. Liu, Y. Bai, and C. Ye, “Subject-independent EEG classification based on a hybrid neural network,”Frontiers in Neuroscience, vol. 17, p. 1124089, 2023

    2023

  36. [36]

    Continuous sensorimotor rhythm based brain computer interface learning in a large population,

    J. R. Stieger, S. A. Engel, and B. He, “Continuous sensorimotor rhythm based brain computer interface learning in a large population,”Scientific Data, vol. 8, no. 1, p. 98, 2021

    2021

  37. [37]

    EEG data of simple and compound limb motor imagery,

    Y. Weibo, “EEG data of simple and compound limb motor imagery,” 2014

    2014

  38. [38]

    Boosting bit rates in noninvasive eeg single-trial classifications by feature combination and multiclass paradigms,

    G. Dornhege, B. Blankertz, G. Curio, and K.-R. Muller, “Boosting bit rates in noninvasive eeg single-trial classifications by feature combination and multiclass paradigms,”IEEE transactions on biomedical engineering, vol. 51, no. 6, pp. 993–1002, 2004

    2004

  39. [39]

    Ovpd: Odor-video elicited physiological signal database for emotion recognition,

    J. Xue, J. Wang, S. Hu, N. Bi, and Z. Lv, “Ovpd: Odor-video elicited physiological signal database for emotion recognition,”IEEE Transactions on Instrumentation and Measurement, vol. 71, pp. 1–12, 2022

    2022

  40. [40]

    Cross-subject emotion EEG signal recognition based on source microstate analysis,

    L. Zhang, D. Xiao, X. Guo, F. Li, W. Liang, and B. Zhou, “Cross-subject emotion EEG signal recognition based on source microstate analysis,”Frontiers in Neuroscience, vol. 17, p. 1288580, 2023

    2023

  41. [41]

    Emotionmeter: A multimodal framework for recognizing human emotions,

    W. Zheng, W. Liu, Y. Lu, B. Lu, and A. Cichocki, “Emotionmeter: A multimodal framework for recognizing human emotions,”IEEE Transactions on Cybernetics, pp. 1–13, 2018

    2018

  42. [42]

    FMLAN: A novel framework for cross-subject and cross-session EEG emotion recognition,

    P. Yu, X. He, H. Li, H. Dou, Y. Tan, H. Wu, and B. Chen, “FMLAN: A novel framework for cross-subject and cross-session EEG emotion recognition,”Biomedical Signal Processing and Control, vol. 100, p. 106912, 2025

    2025

  43. [43]

    Multisource associate domain adaptation for cross-subject and cross-session EEG emotion recognition,

    Q. She, C. Zhang, F. Fang, Y. Ma, and Y. Zhang, “Multisource associate domain adaptation for cross-subject and cross-session EEG emotion recognition,”IEEE Transactions on Instrumentation and Measurement, vol. 72, pp. 1–12, 2023. Art. no. 2515512

    2023

  44. [44]

    Investigating critical frequency bands and channels for EEG-based emotion recognition with deep neural networks,

    W.-L. Zheng and B.-L. Lu, “Investigating critical frequency bands and channels for EEG-based emotion recognition with deep neural networks,”IEEE Transactions on Autonomous Mental Development, vol. 7, no. 3, pp. 162–175, 2015

    2015

  45. [45]

    Differential entropy feature for EEG-based emotion classification,

    R.-N. Duan, J.-Y. Zhu, and B.-L. Lu, “Differential entropy feature for EEG-based emotion classification,” in6th International IEEE/EMBS Conference on Neural Engineering (NER), pp. 81–84, IEEE, 2013

    2013

  46. [46]

    Cross-subject EEG emotion recognition using multi-source domain manifold feature selection,

    Q. She, X. Shi, F. Fang, Y. Ma, and Y. Zhang, “Cross-subject EEG emotion recognition using multi-source domain manifold feature selection,”Computers in Biology and Medicine, vol. 159, p. 106860, 2023

    2023

  47. [47]

    Two-level domain adaptation neural network for EEG-based emotion recognition,

    G. Bao, N. Zhuang, L. Tong, B. Yan, J. Shu, L. Wang, Y. Zeng, and C. Shen, “Two-level domain adaptation neural network for EEG-based emotion recognition,”Frontiers in Human Neuroscience, vol. 14, p. 605246, 2021

    2021

  48. [48]

    Subject-independent EEG-based emotion recognition using adversarial learning,

    S. Hwang, M. Ki, K. Hong, and H. Byun, “Subject-independent EEG-based emotion recognition using adversarial learning,” in2020 8th International Winter Conference on Brain-Computer Interface (BCI), pp. 1–4, IEEE, 2020. 18 IOP PublishingJournalvv(yyyy) aaaaaa Authoret al

    2020

  49. [49]

    Cross-subject EEG-based emotion recognition using deep metric learning and adversarial training,

    H. R. A. Alameer, P. Salehpour, S. H. Aghdasi, and M.-R. Feizi-Derakhshi, “Cross-subject EEG-based emotion recognition using deep metric learning and adversarial training,”IEEE Access, vol. 12, pp. 130241–130252, 2024

    2024

  50. [50]

    A novel baseline removal paradigm for subject-independent features in emotion classification using eeg,

    M. Z. I. Ahmed, N. Sinha, E. Ghaderpour, S. Phadikar, and R. Ghosh, “A novel baseline removal paradigm for subject-independent features in emotion classification using eeg,” Bioengineering, vol. 10, no. 1, p. 54, 2023

    2023

  51. [51]

    Cross-subject EEG emotion recognition combined with connectivity features and meta-transfer learning,

    J. Li, H. Hua, Z. Xu, L. Shu, X. Xu, F. Kuang, and S. Wu, “Cross-subject EEG emotion recognition combined with connectivity features and meta-transfer learning,”Computers in Biology and Medicine, vol. 145, p. 105519, 2022

    2022

  52. [52]

    mixEEG: Enhancing EEG federated learning for cross-subject EEG classification with tailored mixup,

    X.-H. Liu, B.-L. Lu, and W.-L. Zheng, “mixEEG: Enhancing EEG federated learning for cross-subject EEG classification with tailored mixup,” inProceedings of the Annual Meeting of the Cognitive Science Society, vol. 47, 2025

    2025

  53. [53]

    Deap: A database for emotion analysis; using physiological signals,

    S. Koelstra, C. Muhl, M. Soleymani, J.-S. Lee, A. Yazdani, T. Ebrahimi, T. Pun, A. Nijholt, and I. Patras, “Deap: A database for emotion analysis; using physiological signals,”IEEE transactions on affective computing, vol. 3, no. 1, pp. 18–31, 2011

    2011

  54. [54]

    Calibration free meta learning based approach for subject independent eeg emotion recognition,

    S. Bhosale, R. Chakraborty, and S. K. Kopparapu, “Calibration free meta learning based approach for subject independent eeg emotion recognition,”Biomedical Signal Processing and Control, vol. 72, p. 103289, 2022

    2022

  55. [55]

    An eeg dataset for cross-session mental workload estimation: passive bci competition of the neuroergonomics conference 2021,

    M. F. Hinss, L. Darmet, B. Somon, E. Jahanpour, F. Lotte, S. Ladouce, and R. N. Roy, “An eeg dataset for cross-session mental workload estimation: passive bci competition of the neuroergonomics conference 2021,”(No Title), 2021

    2021

  56. [56]

    Brain-computer interfacing using modulations of alpha activity induced by covert shifts of attention,

    M. S. Treder, A. Bahramisharif, N. M. Schmidt, M. A. Van Gerven, and B. Blankertz, “Brain-computer interfacing using modulations of alpha activity induced by covert shifts of attention,”Journal of neuroengineering and rehabilitation, vol. 8, no. 1, p. 24, 2011

    2011

  57. [57]

    Inter-subject transfer learning with an end-to-end deep convolutional neural network for EEG-based BCI,

    F. Fahimi, Z. Zhang, W. B. Goh, T.-S. Lee, K. K. Ang, and C. Guan, “Inter-subject transfer learning with an end-to-end deep convolutional neural network for EEG-based BCI,”Journal of Neural Engineering, vol. 16, no. 2, p. 026007, 2019

    2019

  58. [58]

    Thinking out loud, an open-access eeg-based bci dataset for inner speech recognition,

    N. Nieto, V. Peterson, H. L. Rufiner, J. E. Kamienkowski, and R. Spies, “Thinking out loud, an open-access eeg-based bci dataset for inner speech recognition,”Scientific data, vol. 9, no. 1, p. 52, 2022

    2022

  59. [59]

    Sleep-edf database expanded (version 1.0.0)

    B. Kemp, “Sleep-edf database expanded (version 1.0.0).” PhysioNet, 2013

    2013

  60. [60]

    ManyDG: Many-domain generalization for healthcare applications,

    C. Yang, M. B. Westover, and J. Sun, “ManyDG: Many-domain generalization for healthcare applications,” inThe Eleventh International Conference on Learning Representations, 2023

    2023

  61. [61]

    CHB-MIT Scalp EEG Database (version 1.0.0)

    J. Guttag, “CHB-MIT Scalp EEG Database (version 1.0.0).” PhysioNet, 2010. RRID:SCR 007345

    2010

  62. [62]

    A. H. Shoeb,Application of machine learning to epileptic seizure onset detection and treatment. PhD thesis, Massachusetts Institute of Technology, 2009

    2009

  63. [63]

    Domain adaptation for EEG-based, cross-subject epileptic seizure prediction,

    I. Jemal, L. Abou-Abbas, K. Henni, A. Mitiche, and N. Mezghani, “Domain adaptation for EEG-based, cross-subject epileptic seizure prediction,”Frontiers in Neuroinformatics, vol. 18, p. 1303380, 2024

    2024

  64. [64]

    Cross-patient seizure prediction via continuous domain adaptation and similar sample replay,

    Z. Zhang, A. Liu, Y. Gao, R. Qian, and X. Chen, “Cross-patient seizure prediction via continuous domain adaptation and similar sample replay,”Cognitive Neurodynamics, vol. 19, no. 1, p. 26, 2025

    2025

  65. [65]

    Convolutional transformer networks for epileptic seizure detection,

    N. Ke, T. Lin, Z. Lin, X. Zhou, and T. Ji, “Convolutional transformer networks for epileptic seizure detection,” inProceedings of the 31st ACM International Conference on Information & Knowledge Management, Atlanta, GA, USA, October 17-21, 2022(M. A. Hasan and L. Xiong, eds.), pp. 4109–4113, ACM, 2022

    2022

  66. [66]

    Dataset: Tuh eeg seizure corpus (tusz)

    H. Albaqami, G. M. Hassan, and A. Datta, “Dataset: Tuh eeg seizure corpus (tusz).” Dataset, 2025. 19 IOP PublishingJournalvv(yyyy) aaaaaa Authoret al

    2025

  67. [67]

    Adversarial representation learning for robust patient-independent epileptic seizure detection,

    X. Zhang, L. Yao, M. Dong, Z. Liu, Y. Zhang, and Y. Li, “Adversarial representation learning for robust patient-independent epileptic seizure detection,”IEEE journal of biomedical and health informatics, vol. 24, no. 10, pp. 2852–2859, 2020

    2020

  68. [68]

    DMNet: Self-comparison driven model for subject-independent seizure detection,

    S. Tu, L. Cao, D. Zhang, J. Chen, L. Ma, Y. Zhang, and Y. Yang, “DMNet: Self-comparison driven model for subject-independent seizure detection,” inAdvances in Neural Information Processing Systems (NeurIPS), vol. 37, pp. 28254–28280, 2024

    2024

  69. [69]

    Invariant spatiotemporal representation learning for cross-patient seizure classification,

    Y. Wu, Y. Yang, J. S. Xiao, C. Zhou, H. Sui, and H. Li, “Invariant spatiotemporal representation learning for cross-patient seizure classification,” inNeurIPS 2024 Workshop on NeuroAI, 2024

    2024

  70. [70]

    “siena scalp eeg database (version 1.0.0)

    P. Detti, ““siena scalp eeg database (version 1.0.0)”,” 2020

    2020

  71. [71]

    A dataset of scalp eeg recordings of alzheimer’s disease, frontotemporal dementia and healthy subjects from routine eeg,

    A. Miltiadous, K. D. Tzimourta, T. Afrantou, P. Ioannidis, N. Grigoriadis, D. G. Tsalikakis, P. Angelidis, M. G. Tsipouras, E. Glavas, N. Giannakeas,et al., “A dataset of scalp eeg recordings of alzheimer’s disease, frontotemporal dementia and healthy subjects from routine eeg,”Data, vol. 8, no. 6, p. 95, 2023

    2023

  72. [72]

    Lead: Large foundation model for eeg-based alzheimer’s disease detection

    Y. Wang, N. Huang, N. Mammone, M. Cecchi, and X. Zhang, “Lead: Large foundation model for eeg-based alzheimer’s disease detection,”arXiv preprint arXiv:2502.01678, 2025

    work page arXiv 2025

  73. [73]

    ”40hz auditory entrainment

    M. Lahijanian, H. Aghajan, and Z. Vahabi, “”40hz auditory entrainment”,” 2024

    2024

  74. [74]

    The brainlat project, a multimodal neuroimaging dataset of neurodegeneration from underrepresented backgrounds,

    P. Prado, V. Medel, R. Gonzalez-Gomez, A. Sainz-Ballesteros, V. Vidal, H. Santamar´ ıa-Garc´ ıa, S. Moguilner, J. Mejia, A. Slachevsky, M. I. Behrens,et al., “The brainlat project, a multimodal neuroimaging dataset of neurodegeneration from underrepresented backgrounds,”Scientific Data, vol. 10, no. 1, p. 889, 2023

    2023

  75. [75]

    Eeg p-adic quantum potential accurately identifies depression, schizophrenia and cognitive decline,

    O. Shor, A. Glik, A. Yaniv-Rosenfeld, A. Valevski, A. Weizman, A. Khrennikov, and F. Benninger, “Eeg p-adic quantum potential accurately identifies depression, schizophrenia and cognitive decline,”Plos one, vol. 16, no. 8, p. e0255529, 2021

    2021

  76. [76]

    The two decades brainclinics research archive for insights in neurophysiology (tdbrain) database,

    H. van Dijk, G. van Wingen, D. Denys, S. Olbrich, R. van Ruth, and M. Arns, “The two decades brainclinics research archive for insights in neurophysiology (tdbrain) database,” Scientific Data, vol. 9, p. 333, 2022

    2022

  77. [77]

    Two decades - brainclinics research archive for insights in neurophysiology (td-brain)

    H. van Dijk, G. van Wingen, D. Denys, S. Olbrich, R. van Ruth, and M. Arns, “Two decades - brainclinics research archive for insights in neurophysiology (td-brain).” Dataset (Synapse), 2021

    2021

  78. [78]

    Contrast everything: A hierarchical contrastive framework for medical time-series,

    Y. Wang, Y. Han, H. Wang, and X. Zhang, “Contrast everything: A hierarchical contrastive framework for medical time-series,” inAdvances in Neural Information Processing Systems (NeurIPS), vol. 36, pp. 64601–64624, 2023

    2023

  79. [79]

    The temple university hospital eeg data corpus,

    I. Obeid and J. Picone, “The temple university hospital eeg data corpus,”Frontiers in neuroscience, vol. 10, p. 196, 2016

    2016

  80. [80]

    FAPEX: Fractional amplitude-phase expressor for robust cross-subject seizure prediction,

    R. Zheng, L. Mao, D. Han, T. Luo, Y. Wang, J. Ding, and Y. Yu, “FAPEX: Fractional amplitude-phase expressor for robust cross-subject seizure prediction,” inAdvances in Neural Information Processing Systems (NeurIPS), 2025. Spotlight

    2025

Showing first 80 references.