Pith. sign in

REVIEW 5 major objections 6 minor 74 references

A channel-swap plus adversarial training recipe lets ECoG models generalize to unseen subjects in Parkinson's prediction, outperforming baselines across every setting tested.

Reviewed by Pith at T0; open to challenge. T0 means a machine referee read the full paper against a public rubric. the ladder, T0–T4 →

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

T0 review · deepseek-v4-flash

2026-08-03 01:04 UTC pith:GL7YKNMA

load-bearing objection Plausible method and a potentially useful ECoG benchmark, but the benchmark's time-locked labels with no sham control undercut the central generalization claim. the 5 major comments →

arxiv 2602.10528 v2 pith:GL7YKNMA submitted 2026-02-11 cs.LG cs.AI

A swap-adversarial framework for improving domain generalization in electrocorticography-based Parkinson's disease classification

Seongwon Jin , Hanseul Choi , Sunggu Yang , Sungho Park , Jibum Kim This is my paper
classification cs.LG cs.AI
keywords ECoGEEGParkinson's diseasedomain generalizationadversarial learningdata augmentationchannel swappingbenchmark dataset
verification ladder T0 review T1 audit T2 compute T3 formal T4 reserved

The pith

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

The paper claims that a two-step recipe — swapping functionally matched channels between same-class subjects, then adversarially training a subject discriminator — makes deep models generalize to unseen subjects in ECoG-based Parkinson's disease classification. The authors argue that the swap weakens subject-specific structural cues before adversarial training, so the adversarial step removes remaining subject bias without destroying task-relevant signal. They report consistent gains over EEGNet and DMMR baselines across cross-subject, cross-modality (wireless-wired), and cross-dataset (EEG benchmark) settings, with the largest gains in highly variable environments. They also introduce MOCOP, an open ECoG benchmark from six 6-OHDA rats, intended as the first reproducible benchmark for ECoG-based PD prediction.

Core claim

On its own terms, the paper establishes that inter-subject channel swapping (ISBCS), applied at the data level before domain-adversarial learning, reduces the F-statistic of between-subject vs within-subject feature variance from 24.48 to 0.99, and that the full framework outperforms baselines in every setting tested. The central claim is that the two components are complementary: ISBCS weakens subject-specific structure that adversarial learning can then remove without harming task-relevant features, yielding subject-invariant representations that generalize to unseen subjects and unseen recording environments.

What carries the argument

Inter-Subject Balanced Channel Swap (ISBCS) is the data-level engine: with probability p per channel, it swaps the signal of a channel between two same-class samples from different subjects, guided by a brain map to pair functionally corresponding channels. This is coupled with domain-adversarial learning through a Gradient Reversal Layer (GRL) and a domain classifier predicting subject identity from features, plus a mutual-information penalty on the predicted subject distribution. The backbone is EEGNet, a compact CNN for brain signals.

Load-bearing premise

The two classes are defined by recording time — before versus one week after stimulation — with no sham or counterbalanced control, so any slow drift in electrodes, behavior, or environment is perfectly confounded with the label; if the model exploits that drift, the central generalization claim collapses.

What would settle it

Train the framework on MOCOP with class labels shuffled within each subject's time sequence, or collect a sham-stimulated cohort where the same electrode implant and handling occur without active stimulation; if accuracy remains near 0.99, the model is reading time-linked artifacts rather than PD-related neural state.

Watch this falsifier — get emailed when new claim-graph text bears on it.

If this is right

  • If correct, cross-subject ECoG classification becomes practical with small cohorts: channel swapping requires no new data and the gains are largest exactly where subject variability is high.
  • The framework transfers across modalities — training on wireless recordings and testing on wired (or vice versa) works, and it also transfers across EEG datasets from different institutions.
  • The MOCOP dataset, once public, would give the ECoG community a standard benchmark with reproducible subject splits and a trained baseline, enabling fair comparisons that the paper argues were previously missing.
  • The reduction in F-statistic from 24.48 to 0.99 provides a measurable target for other augmentation methods: subject separability in feature space can be driven to near zero without necessarily destroying task information.

Where Pith is reading between the lines

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

  • The claim that corresponding channels preserve task-relevant signal relies on an anatomical correspondence that the paper does not verify; a testable extension is to swap random channels and compare accuracy to see how much the brain-map guidance matters.
  • The benchmark's labels are time-confounded (pre-stimulation versus after one week of stimulation), so part of the reported accuracy may reflect slow electrode drift or behavioral state rather than PD-specific neural change; a sham-stimulated control group would separate these.
  • The F-statistic reduction shows ISBCS makes subjects indistinguishable in feature space, but it does not show that the remaining features carry the class information; a class-conditional or mutual-information measure would sharpen the claim.
  • Because the framework generalizes from ECoG to EEG, it may also generalize to other electrophysiological modalities such as local field potentials or single-unit recordings, which could be tested with existing public datasets.

Editorial analysis

A structured set of objections, weighed in public.

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

Referee Report

5 major / 6 minor

Summary. The paper proposes a Swap-Adversarial Framework (SAF) for domain generalization in ECoG/EEG classification, combining preprocessing (band-pass, notch, ASR), an Inter-Subject Balanced Channel Swap (ISBCS) augmentation that swaps corresponding channels between subjects of the same class, and domain-adversarial learning with a gradient reversal layer and an entropy-based mutual-information penalty. The authors introduce MOCOP, an ECoG benchmark built from 6-OHDA-lesioned rat recordings, and report cross-subject, cross-modality (wireless/wired), ablation, and cross-dataset EEG experiments claiming consistent improvements over EEGNet and DMMR baselines, with the largest gains in the most variable environments. The central claims are that SAF learns task-relevant, subject-invariant features and that MOCOP is the first reproducible benchmark for ECoG-based PD prediction.

Significance. If the benchmark labels were valid and the reported gains were robust, the framework would be a useful contribution: it is a simple, modular recipe (preprocessing + data augmentation + adversarial training) that could transfer across brain-signal modalities, and MOCOP would fill a real gap as a public ECoG benchmark for PD prediction. The paper has concrete strengths: the ablation study supports complementarity of ISBCS and DAL, the method is evaluated under several shift scenarios, and the authors commit to releasing data and code. However, the label construction in MOCOP, the per-target hyperparameter selection, and the circular F-statistic evidence all undermine the central generalization and benchmark claims. As presented, the evidence does not establish that the model learns PD-related neural state rather than time-correlated nuisance variation.

major comments (5)
  1. [Section 4.1.2] The dataset labels are perfectly confounded with recording time. Class 0 is defined as the state without electrical stimulation and class 1 as the state after one week of continuous stimulation, with no sham control, unstimulated control group, or counterbalanced design described. Because all class-0 recordings precede all class-1 recordings within each animal, any slow drift in electrode impedance, recording equipment, behavior, or environment can produce the global spectral changes reported in Figure 7. The high accuracies in Experiments 1 and 2 (e.g., 98.9% wireless cross-subject) may therefore reflect classification of temporal drift rather than PD-related neural state. This is a load-bearing validity issue for MOCOP as a 'reproducible benchmark' and for the central generalization claim. The manuscript must show control analyses (e.g., shuffled-time labels, unstimulated animals) or s
  2. [Section 5.3 / Table 3] Hyperparameters lambda_MI and lambda_GRL are grid-searched and reported separately for each target rat and each transfer direction (e.g., Wireless Rat1: 0.001/0.001; Wireless Rat3: 3.334/1.25). If the validation split used to select these values includes any target-domain data, the 'unseen domain' evaluation is transductive and the comparison is optimistic. The paper must state exactly how validation folds were constructed relative to the source/target split. In addition, the EEG cross-dataset results in Figure 17 are reported as point estimates with no error bars or significance tests, so the claim that the method 'consistently outperforms all baselines across all settings' is not statistically supported.
  3. [Section 5.6] The F-statistic drop from 24.48 to 0.99 after ISBCS is a direct consequence of the augmentation construction: swapping channels between subjects mixes subject identities in the swapped channels by design, so between-subject variance in the feature space decreases mechanically. This is not independent evidence that task-relevant, subject-invariant features are learned. The passage should be reframed as a sanity check of the augmentation's intended effect, not as 'statistical evidence' for the method's generalization benefit. The circularity weakens the central claim that ISBCS improves generalization.
  4. [Section 5.5, Experiments 1-2] Each ECoG setting contains only three source subjects; the cross-subject evaluation is at most a 2-vs-1 split repeated three times. The reported +/- values (e.g., 0.989 +/- 0.010 for wireless accuracy) are standard deviations across three fold results and are not a stable estimate of model performance. No significance tests are provided. With n=3, one easy or hard target subject can drive the average, so the claimed 41% improvement over EEGNet in wireless accuracy should be interpreted with this limitation explicitly acknowledged.
  5. [Contribution 1 / Section 4.1] The 'first reproducible benchmark' claim is not currently verifiable. MOCOP is a re-annotation and re-partitioning of recordings from Shin et al. (2025), and the manuscript provides no data DOI, accession number, or code link ('will be made publicly available upon publication'). Reproducibility cannot be assessed. Together with the label confound identified in Section 4.1.2, the benchmark contribution as stated is not supportable.
minor comments (6)
  1. [Abstract and Title] The abstract's first sentence swaps EEG and ECoG definitions: 'Electroencephalography (ECoG)' and 'electrocorticography (EEG)'. The title also says 'Electroencephalography-Based' while the paper is about ECoG. Please correct throughout.
  2. [Eq. (3)] L_MI is defined as the entropy of the domain classifier's predicted subject distribution H(h_psi(z)). This is not the mutual information between z and s; it is a proxy. Please rename it as a conditional-entropy proxy or derive the actual mutual information.
  3. [Section 5.1 and Figures] The baseline name is inconsistent: Section 5.1 lists 'DDMR' while Figures 13-17 and the Results sections use 'DMMR'. Choose one spelling.
  4. [Section 4.2.1 vs Section 5.3] Section 4.2.1 says ASR was not applied to the EEG datasets, but Section 5.3 says clean_asr() was used for both ECoG and EEG data. Clarify which preprocessing the EEG models actually received.
  5. [Section 3.3.1] ISBCS assumes channel indices correspond functionally across subjects, described as 'guided by a brain map,' but no anatomical registration or channel-mapping procedure is described. Please make the correspondence explicit or discuss its validity.
  6. [Sections 4.1.3 and 4.2.2] The silhouette score histograms and t-SNE plots are used to support claims of inter-subject separability, but no confidence intervals or permutation tests are provided. Please temper the wording or add quantitative uncertainty.

Circularity Check

1 steps flagged

ISBCS validation via F-statistic is circular; central accuracy claims remain empirical

specific steps
  1. self definitional [Section 5.6, Ablation Study (Figure 16)]
    "To quantitatively assess the degree of inter-subject separability, we computed the F-statistic, which measures the ratio of between-subject variance to within-subject variance. For the original data, the F-statistic was 24.48, indicating significant subject-specific differences. After applying ISBCS augmentation, the F-statistic dramatically decreased to 0.99, which indicates negligible between-subject differences. This reduction provides statistical evidence that ISBCS effectively reduces inter-subject variability."

    ISBCS constructs augmented samples by swapping channel data between subjects of the same class while retaining the host subject's label. After swapping, the subject label no longer identifies the provenance of all channels in the sample, so a between-subject variance computed over these labels is definitionally deflated. The decrease from 24.48 to 0.99 is a direct consequence of how the augmentation scrambles subject identity, not an independent measurement of learned domain invariance. Presenting this constructed decrease as 'statistical evidence' of ISBCS effectiveness is therefore circular as a validation step, although it is not the main accuracy result.

full rationale

The central contribution of SAF is an empirical training recipe: ISBCS augmentation plus domain-adversarial learning with a GRL, evaluated on held-out subjects and cross-modality transfers. The reported accuracies (e.g., 98.9% wireless cross-subject, 91.2% wired-to-wireless) are obtained on unseen test subjects/domains and are not derived from the fitted hyperparameters or from the construction of ISBCS itself, so the core generalization claim is not circular in the derivation chain. The one defensible circular step is the F-statistic evidence in the ablation study: after ISBCS swaps channels across subjects, computing an F-statistic on the original subject labels measures a mixture artifact rather than independent evidence of subject-invariance. This step is supporting validation only, not the main result. The dataset labeling issue (class 0 = pre-stimulation, class 1 = after one week of stimulation, with no sham control) is a serious validity confound, but it is not a circularity of the derivation. The self-citations to Shin et al. for data acquisition and behavioral labels function as provenance citations for a published dataset, not as an imported uniqueness theorem or unverified ansatz. Eq. (3) calls the domain classifier's output entropy 'mutual information' without derivation; that is an overclaim/correctness concern rather than a circular reasoning step. Overall, the paper's primary empirical comparisons are self-contained; the circularity is limited to a construct-dependent validation metric.

Axiom & Free-Parameter Ledger

3 free parameters · 6 axioms · 0 invented entities

The method is empirical ML on a small animal dataset. No new physical entities are postulated. The key untested premises are the time-based label validity and the functional equivalence of channels across rats. The hyperparameters λ_MI and λ_GRL are fitted per setting, adding degrees of freedom that could absorb some of the reported gains.

free parameters (3)
  • lambda_MI (per setting) = Table 3 values, e.g., 0.001, 3.334, 7.916, 0.417
    Grid-searched per experimental setting on validation data; controls entropy regularization strength.
  • lambda_GRL (per setting) = Table 3 values, e.g., 0.001, 1.25, 8.889, 4.445
    Grid-searched per experimental setting on validation data; controls gradient reversal strength.
  • ISBCS swap probability p = not stated in final experiments
    Bernoulli probability for channel swapping; sensitivity is shown in Figure 4, but the value used in the main results is not reported.
axioms (6)
  • domain assumption Pre-stimulation recordings correspond to Parkinsonian state and post-stimulation recordings to improved state, with no sham/control condition.
    Section 4.1.2 labels classes by time relative to stimulation; this is the main confound and is not validated against a control group.
  • domain assumption 6-OHDA-lesioned rats are a valid model of human Parkinson's disease for ECoG-based prediction.
    Inherited from prior studies (Shin et al., 2025; Wang et al., 2024a); no independent validation in this paper.
  • domain assumption Artifact Subspace Reconstruction removes noise while preserving task-relevant neural signals.
    Applied with default settings; no quantitative check that the removed components are purely artifactual.
  • domain assumption Domain-adversarial learning can suppress subject-specific features without removing task-relevant information.
    The premise of DANN (Ganin et al., 2016); the paper relies on it but the trade-off is known to be sensitive to hyperparameters.
  • domain assumption Channel indices correspond functionally across rats, so same-index channel swapping is meaningful.
    ISBCS swaps 'corresponding channels' guided by a brain map (Section 3.3.1), but no anatomical registration or channel-matching protocol is described.
  • domain assumption Swapping channels between two samples of the same class preserves the class label.
    Necessary for ISBCS; assumed without empirical check on a held-out set.

pith-pipeline@v1.3.0-alltime-deepseek · 21517 in / 13223 out tokens · 133929 ms · 2026-08-03T01:04:11.344793+00:00 · methodology

0 comments
read the original abstract

We propose a novel swap-adversarial framework that mitigates high inter-subject variability and the high-dimensional low-sample-size problem in electrocorticography (ECoG) data. It achieves robust domain generalization across ECoG and electroencephalography (EEG)-based brain-computer interface datasets. Our framework integrates (1) robust preprocessing, (2) inter-subject balanced channel swap (ISBCS) for cross-subject augmentation, and (3) domain-adversarial learning (DAL) to suppress subject-specific bias. The ISBCS method is a bio-inspired channel swapping strategy that exchanges only functionally corresponding channels across subjects, guided by a brain map, to mitigate inter-subject distribution differences. The DAL strategy encourages the model to learn task-relevant shared features. We validate the effectiveness of this framework through extensive experiments under cross-subject, cross-session, and cross-dataset settings. Our framework consistently outperforms all baselines across all settings, showing the most significant improvements in highly variable environments. It also achieves superior cross-dataset performance between public EEG benchmarks, demonstrating strong generalization capability not only for ECoG but also for EEG data. In addition, we introduce a new ECoG dataset, the first reproducible benchmark, which is constructed from long-term ECoG recordings of 6-hydroxydopamine-induced rat models and annotated with neural responses measured before and after electrical stimulation.

Figures

Figures reproduced from arXiv: 2602.10528 by Hanseul Choi, Jibum Kim, Seongwon Jin, Sunggu Yang, Sungho Park.

Figure 1
Figure 1. Figure 1: The overall architecture of Swap-Adversarial Framework. (a) The overall pipeline [PITH_FULL_IMAGE:figures/full_fig_p010_1.png] view at source ↗
Figure 2
Figure 2. Figure 2: Visualization of the preprocessing pipeline. A representative 15-second segment [PITH_FULL_IMAGE:figures/full_fig_p012_2.png] view at source ↗
Figure 3
Figure 3. Figure 3: Inter-Subject Balanced Channel Swap (ISBCS): A counterfactual data augmen [PITH_FULL_IMAGE:figures/full_fig_p013_3.png] view at source ↗
Figure 4
Figure 4. Figure 4: Sensitivity analysis of the ISBCS channel-swap probability [PITH_FULL_IMAGE:figures/full_fig_p014_4.png] view at source ↗
Figure 5
Figure 5. Figure 5: t-SNE visualization of the ECoG data after power spectral density (PSD) trans [PITH_FULL_IMAGE:figures/full_fig_p016_5.png] view at source ↗
Figure 6
Figure 6. Figure 6: Silhouette score histograms based on t-SNE visualization of the ECoG data. [PITH_FULL_IMAGE:figures/full_fig_p017_6.png] view at source ↗
Figure 7
Figure 7. Figure 7: Power Spectral Density Analysis Across Different Datasets and Classes: Com [PITH_FULL_IMAGE:figures/full_fig_p018_7.png] view at source ↗
Figure 8
Figure 8. Figure 8: Inter-Subject Variability Analysis Using Coefficient of Variation (CV): Compari [PITH_FULL_IMAGE:figures/full_fig_p019_8.png] view at source ↗
Figure 9
Figure 9. Figure 9: t-SNE visualization of the UI and UNM dataset after power spectral density [PITH_FULL_IMAGE:figures/full_fig_p020_9.png] view at source ↗
Figure 10
Figure 10. Figure 10: Silhouette score histograms based on t-SNE visualization of the EEG data. [PITH_FULL_IMAGE:figures/full_fig_p021_10.png] view at source ↗
Figure 11
Figure 11. Figure 11: Power Spectral Density Analysis Across Different Datasets and Classes: Com [PITH_FULL_IMAGE:figures/full_fig_p022_11.png] view at source ↗
Figure 12
Figure 12. Figure 12: Inter-Subject Variability Analysis Using Coefficient of Variation (CV): Com [PITH_FULL_IMAGE:figures/full_fig_p023_12.png] view at source ↗
Figure 13
Figure 13. Figure 13: Results of Experiments 1: (Left) Wireless (train) [PITH_FULL_IMAGE:figures/full_fig_p027_13.png] view at source ↗
Figure 14
Figure 14. Figure 14: Results of Experiments 2: (Left) Wireless (train) [PITH_FULL_IMAGE:figures/full_fig_p027_14.png] view at source ↗
Figure 15
Figure 15. Figure 15: Results of Experiments 3: (a) Wireless (train) [PITH_FULL_IMAGE:figures/full_fig_p029_15.png] view at source ↗
Figure 16
Figure 16. Figure 16: t-SNE visualization of the original and ISBCS-augmented data features. Each [PITH_FULL_IMAGE:figures/full_fig_p030_16.png] view at source ↗
Figure 17
Figure 17. Figure 17: Results of Experiments 4: (Left) UI (train) [PITH_FULL_IMAGE:figures/full_fig_p030_17.png] view at source ↗

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

74 extracted references · 34 canonical work pages · 4 internal anchors

  1. [1]

    Parkinson's disease diagnosis using deep learning: A bibliometric analysis and literature review

    Abumalloh, R.A., Nilashi, M., Samad, S., Ahmadi, H., Alghamdi, A., et al., 2024. Parkinson's disease diagnosis using deep learning: A bibliometric analysis and literature review. Ageing Res Rev. 96, 102285. https://doi.org/10.1016/j.arr.2024.102285

    arXiv 2024

  2. [2]

    Agrawal, S., Sahu, S.P., 2025. Quantitative EEG-Based Regional Spectral Analysis for Parkinson's Disease Detection, In: 2025 IEEE International Conference on Interdisciplinary Approaches in Technology and Management for Social Innovation, Vol. 3. pp. 1-6. https://doi.org/10.1109/IATMSI64286.2025.10985366

    arXiv 2025

  3. [3]

    Predictive accuracy of CNN for cortical oscillatory activity in an acute rat model of parkinsonism

    Ali, A.A.N., Alam, M., Klein, S.C., Behmann, N., Krauss, J.K., et al., 2022. Predictive accuracy of CNN for cortical oscillatory activity in an acute rat model of parkinsonism. Neural Networks. 146, 334–340. https://doi.org/10.1016/j.neunet.2021.11.025

    work page doi:10.1016/j.neunet.2021.11.025 2022

  4. [4]

    Invariant risk minimization

    Arjovsky, M., Bottou, L., Gulrajani, I., Lopez-Paz, D., 2020. Invariant risk minimization. arXiv preprint. https://doi.org/10.48550/arXiv.1907.02893

    Pith/arXiv arXiv doi:10.48550/arxiv.1907.02893 2020

  5. [5]

    Deep brain stimulation of the subthalamic nucleus for the treatment of Parkinson's disease

    Benabid, A.L., Chabardes, S., Mitrofanis, J., Pollak, P., 2009. Deep brain stimulation of the subthalamic nucleus for the treatment of Parkinson's disease. Lancet Neurol. 8 (1), 67-81. https://doi.org/10.1016/S1474-4422(08)70291-6

    doi:10.1016/s1474-4422(08)70291-6 2009

  6. [6]

    A riemannian modification of artifact subspace reconstruction for EEG artifact handling

    Blum, S., Jacobsen, N.S.J., Bleichner, M.G., Debener, S., 2019. A riemannian modification of artifact subspace reconstruction for EEG artifact handling. Front Hum Neurosci. 13, 141. https://doi.org/10.3389/fnhum.2019.00141

    arXiv 2019

  7. [7]

    Prediction of mild Parkinsonism revealed by neural oscillatory changes and machine learning

    Bore, J.C., Campbell, B.A., Cho, H., Gopalakrishnan, R., Machado, A.G., et al., 2020. Prediction of mild Parkinsonism revealed by neural oscillatory changes and machine learning. J Neurophysiol. 124 (6), 1698-1705. https://doi.org/10.1152/jn.00534.2020

    arXiv 2020

  8. [8]

    Deep Learning-Driven EEG Analysis for Personalized Deep Brain Stimulation Programming in Parkinson's Disease

    Calvo Peiro, N., Haugland, M.R., Kutuzova, A., Graef, C., Bocum, A., et al., 2025. Deep Learning-Driven EEG Analysis for Personalized Deep Brain Stimulation Programming in Parkinson's Disease. medRxiv preprint. https://doi.org/10.1101/2025.02.11.25321886

    work page doi:10.1101/2025.02.11.25321886 2025

  9. [9]

    A method for optimizing the artifact subspace reconstruction performance in low-density EEG

    Cataldo, A., Criscuolo, S., De Benedetto, E., Masciullo, A., Pesola, M., Schiavoni, R., Invitto, S., 2022. A method for optimizing the artifact subspace reconstruction performance in low-density EEG. IEEE Sens J. 22 (21), 21257–21265. https://doi.org/10.1109/JSEN.2022.3208768

    arXiv 2022

  10. [10]

    Diminished EEG habituation to novel events effectively classifies Parkinson's patients

    Cavanagh, J.F., Kumar, P., Mueller, A.A., Richardson, S.P., Mueen, A., 2018. Diminished EEG habituation to novel events effectively classifies Parkinson's patients. Clin Neurophysiol. 129 (2), 409–418. https://doi.org/10.1016/j.clinph.2017.11.023

    work page doi:10.1016/j.clinph.2017.11.023 2018

  11. [11]

    Evaluation of artifact subspace reconstruction for automatic EEG artifact removal

    Chang, C.Y., Hsu, S.H., Pion-Tonachini, L., Jung, T.P., 2018. Evaluation of artifact subspace reconstruction for automatic EEG artifact removal. In: 2018 40th annual international conference of the IEEE engineering in medicine and biology society (EMBC). pp. 1242-1245

    2018

  12. [12]

    Transformer-based neural speech decoding from surface and depth electrode signals

    Chen, J., Chen, X., Wang, R., Le, C., Khalilian-Gourtani, A., Jensen, E., Dugan, P., Doyle, W., Devinsky, O., Friedman, D., Flinker, A., Wang, Y., 2025. Transformer-based neural speech decoding from surface and depth electrode signals. J Neural Eng 22. https://doi.org/10.1088/1741-2552/adab21

    work page doi:10.1088/1741-2552/adab21 2025

  13. [13]

    Kernel two-sample tests for manifold data

    Cheng, X., Xie, Y., 2024. Kernel two-sample tests for manifold data. 30 (4), 2572-2597. https://doi.org/10.3150/23-BEJ1685

    work page doi:10.3150/23-bej1685 2024

  14. [14]

    Cluster embedding joint-probability-discrepancy transfer for cross-subject seizure detection

    Cui, X., Cao, J., Lai, X., Jiang, T., Gao, F., 2022. Cluster embedding joint-probability-discrepancy transfer for cross-subject seizure detection. IEEE Trans Neural Syst Rehabil Eng. 31, 593–605. https://doi.org/10.1109/TNSRE.2022.3229066

    arXiv 2022

  15. [15]

    Towards best practice of interpreting deep learning models for EEG-based brain computer interfaces

    Cui, J., Yuan, L., Wang, Z., Li, R., Jiang, T., 2023. Towards best practice of interpreting deep learning models for EEG-based brain computer interfaces. Front Comput Neurosci. 17, 1232925. https://doi.org/10.3389/fncom.2023.1232925

    arXiv 2023

  16. [16]

    MADG: margin-based adversarial learning for domain generalization

    Dayal, A., KB, V., Cenkeramaddi, L.R., Mohan, C., Kumar, A., et al., 2023. MADG: margin-based adversarial learning for domain generalization. Adv Neural Inf Process Syst. 36, 58938-58952

    2023

  17. [17]

    Disentangling document topic and author gender in multiple languages: Lessons for adversarial debiasing

    Dayanik, E., Padó, S., 2021. Disentangling document topic and author gender in multiple languages: Lessons for adversarial debiasing. In: Proceedings of the Eleventh Workshop on Computational Approaches to Subjectivity, Sentiment and Social Media Analysis. pp. 50-61

    2021

  18. [18]

    Global, regional, and national burden of Parkinson's disease, 1990–2016: a systematic analysis for the Global Burden of Disease Study 2016

    Dorsey, E.R., Elbaz, A., Nichols, E., Abd-Allah, F., Abdelalim, A., et al., 2018. Global, regional, and national burden of Parkinson's disease, 1990–2016: a systematic analysis for the Global Burden of Disease Study 2016. Lancet Neurol. 17 (11), 939–953. https://doi.org/10.1016/S1474-4422(18)30295-3

    work page doi:10.1016/s1474-4422(18)30295-3 2018

  19. [19]

    Variational mode decomposition

    Dragomiretskiy, K., Zosso, D., 2014. Variational mode decomposition. IEEE Trans Signal Process. 62 (3), 531–544. https://doi.org/10.1109/TSP.2013.2288675

    arXiv 2014

  20. [20]

    Domain-adversarial training of neural networks

    Ganin, Y., Ustinova, E., Ajakan, H., Germain, P., Larochelle, H., et al., 2016. Domain-adversarial training of neural networks. J Mach Learn Res. 17 (59), 1-35

    2016

  21. [21]

    Elucidating relations between fMRI, ECoG, and EEG through a common natural stimulus

    Haufe, S., Deguzman, P., Henin, S., Arcaro, M., Honey, C.J., et al., 2018. Elucidating relations between fMRI, ECoG, and EEG through a common natural stimulus. NeuroImage. 179, 79-81. https://doi.org/10.1016/j.neuroimage.2018.06.016

    work page doi:10.1016/j.neuroimage.2018.06.016 2018

  22. [22]

    The empirical mode decomposition and the Hilbert spectrum for nonlinear and non-stationary time series analysis, In: Proceedings of the Royal Society of London

    Huang, N.E., Shen, Z., Long, S.R., Wu, M.C., Shih, H.H., et al., 1998. The empirical mode decomposition and the Hilbert spectrum for nonlinear and non-stationary time series analysis, In: Proceedings of the Royal Society of London. Series A: mathematical, physical and engineering sciences, vol. 454, no 1971. pp. 903-995. https://doi.org/10.1098/rspa.1998.0193

    arXiv 1998

  23. [23]

    Discrepancy between inter- and intra-subject variability in EEG-based motor imagery brain-computer interface: Evidence from multiple perspectives

    Huang, G., Zhao, Z., Zhang, S., Hu, Z., Fan, J., et al., 2023. Discrepancy between inter- and intra-subject variability in EEG-based motor imagery brain-computer interface: Evidence from multiple perspectives. Front Neurosci. 17, 1122661. https://doi.org/10.3389/fnins.2023.1122661

    arXiv 2023

  24. [24]

    M., Hersche, M., Wang, X., Kobayashi, N., Cavigelli, L., et al., 2020

    Ingolfsson, T. M., Hersche, M., Wang, X., Kobayashi, N., Cavigelli, L., et al., 2020. EEG-TCNet: An accurate temporal convolutional network for embedded motor-imagery brain–machine interfaces. In: 2020 IEEE International Conference on Systems, Man, and Cybernetics (SMC). pp. 2958-2965

    2020

  25. [25]

    Do, Song, I.U., Chung, Y.A., Jeong, J., 2016

    Jeong, D.H., Kim, Y. Do, Song, I.U., Chung, Y.A., Jeong, J., 2016. Wavelet energy and wavelet coherence as eeg biomarkers for the diagnosis of Parkinson's disease-related dementia and Alzheimer's disease. Entropy. 18 (1). https://doi.org/10.3390/e18010008

    work page doi:10.3390/e18010008 2016

  26. [26]

    Bi-Band ECoGNet for ECoG Decoding on Classification Task

    Ji, C., 2024. Bi-Band ECoGNet for ECoG Decoding on Classification Task. arXiv preprint. https://doi.org/10.48550/arXiv.2412.00378

    work page internal anchor Pith review Pith/arXiv arXiv doi:10.48550/arxiv.2412.00378 2024

  27. [27]

    Parkinson's disease

    Kalia, L.V., Lang, A.E., 2015. Parkinson's disease. Lancet. 386 (9996), 896-912. https://doi.org/10.1016/S0140-6736(14)61393-3

    work page doi:10.1016/s0140-6736(14)61393-3 2015

  28. [28]

    Unsupervised domain adaptation in brain lesion segmentation with adversarial networks

    Kamnitsas, K., Baumgartner, C., Ledig, C., Newcombe, V.F.J., Simpson, J.P., et al., 2016. Unsupervised domain adaptation in brain lesion segmentation with adversarial networks. In: International conference on information processing in medical imaging. Springer, pp. 597-609. https://doi.org/10.1007/978-3-319-59050-9\_47

    doi:10.1007/978-3-319-59050-9 2016

  29. [29]

    Data augmentation with suboptimal warping for time-series classification

    Kamycki, K., Kapuscinski, T., Oszust, M., 2020. Data augmentation with suboptimal warping for time-series classification. Sensors. 20 (1), 98. https://doi.org/10.3390/s20010098

    work page doi:10.3390/s20010098 2020

  30. [30]

    Electrocorticogram (ECOG) is highly informative in primate visual cortex

    Kanth, S.T., Ray, S., 2020. Electrocorticogram (ECOG) is highly informative in primate visual cortex. J Neurosci. 40 (12), 2430–2444. https://doi.org/10.1523/JNEUROSCI.1368-19.2020

    work page doi:10.1523/jneurosci.1368-19.2020 2020

  31. [31]

    Distinguishing Parkinson's disease with GLCM features from the hankelization of EEG signals

    Karakaş, M.F., Latifoğlu, F., 2023. Distinguishing Parkinson's disease with GLCM features from the hankelization of EEG signals. Diagnostics. 13 (10), 1769. https://doi.org/10.3390/diagnostics13101769

    work page doi:10.3390/diagnostics13101769 2023

  32. [32]

    The influence of filters on EEG-ERP testing: Analysis of motor cortex in healthy subjects

    Karpiel, I., Kurasz, Z., Kurasz, R., Duch, K., 2021. The influence of filters on EEG-ERP testing: Analysis of motor cortex in healthy subjects. Sensors. 21 (22), 7711. https://doi.org/10.3390/s21227711

    work page doi:10.3390/s21227711 2021

  33. [33]

    Learning not to learn: Training deep neural networks with biased data

    Kim, B., Kim, H., Kim, K., Kim, S., Kim, J., 2019. Learning not to learn: Training deep neural networks with biased data. In: Proceedings of the IEEE/CVF conference on computer vision and pattern recognition. pp. 9012-9020

    2019

  34. [34]

    Explainable AI-Driven Neural Activity Analysis in Parkinsonian Rats under Electrical Stimulation

    Kim, J., Choi, H., Kim, G., Yang, S., Baeg, E., et al., 2025. Explainable AI-Driven Neural Activity Analysis in Parkinsonian Rats under Electrical Stimulation. arXiv preprint. https://doi.org/10.48550/arXiv.2502.12471

    work page internal anchor Pith review Pith/arXiv arXiv doi:10.48550/arxiv.2502.12471 2025

  35. [35]

    Efficient artifact removal from low-density wearable EEG using artifacts subspace reconstruction

    Kumaravel, V.P., Kartsch, V., Benatti, S., Vallortigara, G., Farella, E., et al., 2021. Efficient artifact removal from low-density wearable EEG using artifacts subspace reconstruction. In: 2021 43rd annual international conference of the IEEE engineering in medicine & biology society (EMBC). pp. 333–336. https://doi.org/10.1109/EMBC46164.2021.9629771

    arXiv 2021

  36. [36]

    Unsupervised representation learning visualization for brain activity dynamics in REM-sleep

    Kyllönen, M., 2024. Unsupervised representation learning visualization for brain activity dynamics in REM-sleep

    2024

  37. [37]

    High-frequency neural activity and human cognition: Past, present and possible future of intracranial EEG research

    Lachaux, J.P., Axmacher, N., Mormann, F., Halgren, E., Crone, N.E., 2012. High-frequency neural activity and human cognition: Past, present and possible future of intracranial EEG research. Prog Neurobiol. 98 (3), 279-301. https://doi.org/10.1016/j.pneurobio.2012.06.008

    doi:10.1016/j.pneurobio.2012.06.008 2012

  38. [38]

    Self-supervised learning for seizure classification using ECoG spectrograms

    Lam, V., Oliugbo, C., Parida, A., Linguraru, M.G., Anwar, S.M., 2024. Self-supervised learning for seizure classification using ECoG spectrograms. In: Medical Imaging 2024: Computer-Aided Diagnosis, Vol. 12927. pp. 597-602. https://doi.org/10.1117/12.3007431

    work page doi:10.1117/12.3007431 2024

  39. [39]

    EEGNet: A compact convolutional neural network for EEG-based brain-computer interfaces

    Lawhern, V.J., Solon, A.J., Waytowich, N.R., Gordon, S.M., Hung, C.P., et al., 2018. EEGNet: A compact convolutional neural network for EEG-based brain-computer interfaces. J Neural Eng. 15 (5), 056013. https://doi.org/10.1088/1741-2552/aace8c

    doi:10.1088/1741-2552/aace8c 2018

  40. [40]

    Domain generalization with adversarial feature learning

    Li, H., Pan, S.J., Wang, S., Kot, A.C., 2018. Domain generalization with adversarial feature learning. In: Proceedings of the IEEE conference on computer vision and pattern recognition. pp. 5400-5409

    2018

  41. [41]

    Adaptive deep brain stimulation in advanced Parkinson disease

    Little, S., Pogosyan, A., Neal, S., Zavala, B., Zrinzo, L., et al., 2013. Adaptive deep brain stimulation in advanced Parkinson disease. Ann Neurol. 74 (3), 449–457. https://doi.org/10.1002/ana.23951

    doi:10.1002/ana.23951 2013

  42. [42]

    Reducing the subject variability of eeg signals with adversarial domain generalization, In: International Conference on Neural Information Processing

    Ma, B.Q., Li, H., Zheng, W.L., Lu, B.L., 2019. Reducing the subject variability of eeg signals with adversarial domain generalization, In: International Conference on Neural Information Processing. Springer, pp. 30–42. https://doi.org/10.1007/978-3-030-36708-4\_3

    work page doi:10.1007/978-3-030-36708-4 2019

  43. [43]

    Independent component analysis of electroencephalographic data

    Makeig, S., Jung, T.P., Bell, A.J., Sejnowski, T.J., 1995. Independent component analysis of electroencephalographic data. Adv Neural Inf Process Syst. 8

    1995

  44. [44]

    Conserved structures of neural activity in sensorimotor cortex of freely moving rats allow cross-subject decoding

    Melbaum, S., Russo, E., Eriksson, D., Schneider, A., Durstewitz, D., et al., 2022. Conserved structures of neural activity in sensorimotor cortex of freely moving rats allow cross-subject decoding. Nat Commun. 13 (1), 7420. https://doi.org/10.1038/s41467-022-35115-6

    work page doi:10.1038/s41467-022-35115-6 2022

  45. [45]

    RISE-iEEG: Robust to Inter-Subject Electrodes Implantation Variability iEEG Classifier

    Memar, M.O., Ziaei, N., Nazari, B., Yousefi, A., 2025. RISE-iEEG: Robust to Inter-Subject Electrodes Implantation Variability iEEG Classifier. arXiv preprint. https://doi.org/10.48550/arXiv.2408.14477

    work page doi:10.48550/arxiv.2408.14477 2025

  46. [46]

    Cortical activity during motor execution, motor imagery, and imagery-based online feedback

    Miller, K.J., Schalk, G., Fetz, E.E., Den Nijs, M., Ojemann, J.G., Rao, R.P.N., 2010. Cortical activity during motor execution, motor imagery, and imagery-based online feedback. Proc Natl Acad Sci. 107 (9), 4430–4435. https://doi.org/10.1073/pnas.0913697107

    work page doi:10.1073/pnas.0913697107 2010

  47. [47]

    Domain generalization via invariant feature representation

    Muandet, K., Balduzzi, D., Schölkopf, B., 2013. Domain generalization via invariant feature representation. In: International conference on machine learning, PMLR. pp. 10-18

    2013

  48. [48]

    Real-time neuroimaging and cognitive monitoring using wearable dry EEG

    Mullen, T.R., Kothe, C.A.E., Chi, Y.M., Ojeda, A., Kerth, T., et al., 2015. Real-time neuroimaging and cognitive monitoring using wearable dry EEG. IEEE Trans Biomed Eng. 62 (11), 2553–2567. https://doi.org/10.1109/TBME.2015.2481482

    arXiv 2015

  49. [49]

    Parkinson's disease detection from resting state EEG using multi-head graph structure learning with gradient weighted graph attention explanations

    Neves, C., Zeng, Y., Xiao, Y., 2024. Parkinson's disease detection from resting state EEG using multi-head graph structure learning with gradient weighted graph attention explanations. In: International Workshop on Machine Learning in Clinical Neuroimaging. Springer, pp. 3-12. https://doi.org/10.1007/978-3-031-78761-4\_1

    work page doi:10.1007/978-3-031-78761-4 2024

  50. [50]

    FASTER: fully automated statistical thresholding for EEG artifact rejection

    Nolan, H., Whelan, R., Reilly, R.B., 2010. FASTER: fully automated statistical thresholding for EEG artifact rejection. J Neurosci Methods. 192 (1), 152–162. https://doi.org/10.1016/j.jneumeth.2010.07.015

    doi:10.1016/j.jneumeth.2010.07.015 2010

  51. [51]

    A novel automated Parkinson's disease identification approach using deep learning and EEG

    Obayya, M., Saeed, M.K., Maashi, M., Alotaibi, S.S., Salama, A.S., et al., 2023. A novel automated Parkinson's disease identification approach using deep learning and EEG. PeerJ Comput Sci. 9, e1663. https://doi.org/10.7717/peerj-cs.1663

    work page doi:10.7717/peerj-cs.1663 2023

  52. [52]

    Past, present, and future of Parkinson's disease: A special essay on the 200th Anniversary of the Shaking Palsy

    Obeso, J.A., Stamelou, M., Goetz, C.G., Poewe, W., Lang, A.E., et al., 2017. Past, present, and future of Parkinson's disease: A special essay on the 200th Anniversary of the Shaking Palsy. Mov Disord. 32 (9), 1264-1310. https://doi.org/10.1002/mds.27115

    work page doi:10.1002/mds.27115 2017

  53. [53]

    A deep learning approach for Parkinson's disease diagnosis from EEG signals

    Oh, S.L., Hagiwara, Y., Raghavendra, U., Yuvaraj, R., Arunkumar, N., et al., 2020. A deep learning approach for Parkinson's disease diagnosis from EEG signals. Neural Comput Appl. 32 (15), 10927–10933. https://doi.org/10.1007/s00521-018-3689-5

    work page doi:10.1007/s00521-018-3689-5 2020

  54. [54]

    AMICA: An adaptive mixture of independent component analyzers with shared components

    Palmer, J.A., Kreutz-Delgado, K., Makeig, S., 2012. AMICA: An adaptive mixture of independent component analyzers with shared components. Swartz Center for Computational Neuroscience. University of California San Diego, Tech Rep

    2012

  55. [55]

    On the high dimensional power of a linear-time two sample test under mean-shift alternatives

    Reddi, S.J., Ramdas, A., Póczos, B., Singh, A., Wasserman, L., 2015. On the high dimensional power of a linear-time two sample test under mean-shift alternatives. In: Artificial Intelligence and Statistics, PMLR. pp. 772-780

    2015

  56. [56]

    CADDA: Class-wise Automatic Differentiable Data Augmentation for EEG Signals

    Rommel, C., Moreau, T., Paillard, J., Gramfort, A., 2022. CADDA: Class-wise automatic differentiable data augmentation for EEG signals. arXiv preprint. https://doi.org/10.48550/arXiv.2106.13695

    work page internal anchor Pith review Pith/arXiv arXiv doi:10.48550/arxiv.2106.13695 2022

  57. [57]

    Distributionally robust neural networks for group shifts: On the importance of regularization for worst-case generalization

    Sagawa, S., Koh, P.W., Hashimoto, T.B., Liang, P., 2019. Distributionally robust neural networks for group shifts: On the importance of regularization for worst-case generalization. arXiv preprint. https://doi.org/10.48550/arXiv.1911.08731

    Pith/arXiv arXiv doi:10.48550/arxiv.1911.08731 2019

  58. [58]

    Single-trial EEG analysis using an adaptive autoregressive model

    Schlögl, A., Pfurtscheller, G., Schack, B., 1996. Single-trial EEG analysis using an adaptive autoregressive model. In: Proceedings of Fourth International Symposium on Central Nervous Monitoring

    1996

  59. [59]

    Dynamical system based compact deep hybrid network for classification of Parkinson disease related EEG signals

    Shah, S.A.A., Zhang, L., Bais, A., 2020. Dynamical system based compact deep hybrid network for classification of Parkinson disease related EEG signals. Neural Networks. 130, 75–84. https://doi.org/10.1016/j.neunet.2020.06.018

    work page doi:10.1016/j.neunet.2020.06.018 2020

  60. [60]

    Generalizing across domains via cross-gradient training

    Shankar, S., Piratla, V., Chakrabarti, S., Chaudhuri, S., Jyothi, P., et al, 2018. Generalizing across domains via cross-gradient training. arXiv preprint. https://doi.org/10.48550/arXiv.1804.10745

    Pith/arXiv arXiv doi:10.48550/arxiv.1804.10745 2018

  61. [61]

    A Wireless Cortical Surface Implant for Diagnosing and Alleviating Parkinson's Disease Symptoms in Freely Moving Animals

    Shin, H., Kim, K., Lee, J., Nam, J., Baeg, E., et al., 2025. A Wireless Cortical Surface Implant for Diagnosing and Alleviating Parkinson's Disease Symptoms in Freely Moving Animals. Adv Healthc Mater. 14 (17). https://doi.org/10.1002/adhm.202405179

    work page doi:10.1002/adhm.202405179 2025

  62. [62]

    Frontal theta and beta oscillations during lower-limb movement in Parkinson's disease

    Singh, A., Cole, R.C., Espinoza, A.I., Brown, D., Cavanagh, J.F., et al., 2020. Frontal theta and beta oscillations during lower-limb movement in Parkinson's disease. Clin Neurophysiol. 131 (3), 694–702. https://doi.org/10.1016/j.clinph.2019.12.399

    work page doi:10.1016/j.clinph.2019.12.399 2020

  63. [63]

    Dusk2Dawn: an EEGLAB plugin for automatic cleaning of whole-night sleep electroencephalogram using Artifact Subspace Reconstruction

    Somervail, R., Cataldi, J., Stephan, A.M., Siclari, F., Iannetti, G.D., 2023. Dusk2Dawn: an EEGLAB plugin for automatic cleaning of whole-night sleep electroencephalogram using Artifact Subspace Reconstruction. Sleep. 46 (12), zsad208. https://doi.org/10.1093/sleep/zsad208

    work page doi:10.1093/sleep/zsad208 2023

  64. [64]

    EEG conformer: Convolutional transformer for EEG decoding and visualization

    Song, Y., Zheng, Q., Liu, B., Gao, X., 2023. EEG conformer: Convolutional transformer for EEG decoding and visualization. IEEE Trans Neural Syst Rehabil Eng 31, 710–719. https://doi.org/10.1109/TNSRE.2022.3230250

    arXiv 2023

  65. [65]

    EEG and MEG coherence: Measures of functional connectivity at distinct spatial scales of neocortical dynamics

    Srinivasan, R., Winter, W.R., Ding, J., Nunez, P.L., 2007. EEG and MEG coherence: Measures of functional connectivity at distinct spatial scales of neocortical dynamics. J Neurosci Methods. 166 (1), 41–52. https://doi.org/10.1016/j.jneumeth.2007.06.026

    doi:10.1016/j.jneumeth.2007.06.026 2007

  66. [66]

    Generalizable electroencephalographic classification of Parkinson's disease using deep learning

    Sugden, R.J., Diamandis, P., 2023. Generalizable electroencephalographic classification of Parkinson's disease using deep learning. Inform Med Unlocked. 42, 101352. https://doi.org/10.1016/j.imu.2023.101352

    arXiv 2023

  67. [67]

    Local domain generalization with low-rank constraint for EEG-based emotion recognition

    Tao, J., Dan, Y., Zhou, D., 2024. Local domain generalization with low-rank constraint for EEG-based emotion recognition. Front Neurosci. 17, 1213099. https://doi.org/10.3389/fnins.2023.1213099

    arXiv 2024

  68. [68]

    Intracranial EEG fluctuates over months after implanting electrodes in human brain

    Ung, H., Baldassano, S.N., Bink, H., Krieger, A.M., Williams, S., et al., 2017. Intracranial EEG fluctuates over months after implanting electrodes in human brain. J Neural Eng. 14 (5), 056011. https://doi.org/10.1088/1741-2552/aa7f40

    work page doi:10.1088/1741-2552/aa7f40 2017

  69. [69]

    A longitudinal electrophysiological and behavior dataset for PD rat in response to deep brain stimulation

    Wang, X., Chen, M., Shen, Y., Li, Y., Li, S., et al., 2024. A longitudinal electrophysiological and behavior dataset for PD rat in response to deep brain stimulation. Sci Data. 11 (1), 500. https://doi.org/10.12751/g-node.lzvqb5

    work page doi:10.12751/g-node.lzvqb5 2024

  70. [70]

    DMMR: Cross-subject domain generalization for EEG-based emotion recognition via denoising mixed mutual reconstruction

    Wang, Y., Zhang, B., Tang, Y., 2024. DMMR: Cross-subject domain generalization for EEG-based emotion recognition via denoising mixed mutual reconstruction. In: Proceedings of the AAAI conference on artificial intelligence, Vol. 38, No. 1. pp. 628-636. https://doi.org/10.1609/aaai.v38i1.27819

    work page doi:10.1609/aaai.v38i1.27819 2024

  71. [71]

    Bilateral deep brain stimulation vs best medical therapy for patients with advanced parkinson disease: A randomized controlled trial

    Weaver, F.M., Follett, K., Stern, M., Hur, K., Harris, C., et al., 2009. Bilateral deep brain stimulation vs best medical therapy for patients with advanced parkinson disease: A randomized controlled trial. JAMA. 301 (1), 63–73. https://doi.org/10.1001/jama.2008.929

    work page doi:10.1001/jama.2008.929 2009

  72. [72]

    Adversarial Training based Domain Adaptation for Cross-Subject Emotion Recognition

    Woo, S., Zubair, M., Lim, S., Kim, D., 2024. Adversarial Training based Domain Adaptation for Cross-Subject Emotion Recognition. In: The Third Workshop on New Frontiers in Adversarial Machine Learning

    2024

  73. [73]

    Towards Stable and Comprehensive Domain Alignment: Max-Margin Domain-Adversarial Training

    Yang, J., Zou, H., Zhou, Y., Xie, L., 2020. Towards stable and comprehensive domain alignment: Max-margin domain-adversarial training. arXiv preprint. https://doi.org/10.48550/arXiv.2003.13249

    work page internal anchor Pith review Pith/arXiv arXiv doi:10.48550/arxiv.2003.13249 2020

  74. [74]

    Impact of dataset size and long-term ECoG-based BCI usage on deep learning decoders performance

    Śliwowski, M., Martin, M., Souloumiac, A., Blanchart, P., Aksenova, T., 2023. Impact of dataset size and long-term ECoG-based BCI usage on deep learning decoders performance. Front Hum Neurosci. 17, 1111645. https://doi.org/10.3389/fnhum.2023.1111645

    arXiv 2023