English (United Kingdom)

https://curated-unify.zendy.io/wp-json/zendy-region/v1/featured_content/oa?rat=en

https://curated-unify.zendy.io/wp-json/zendy-region/v1/highlighted_journal/

Global: Zendy Plus annual

Presents the access of premium content as premium feature

Premium Content

Presents the keyphrase highlighting as premium feature

Keyphrase Highlighting

Presents the summarisation as premium feature

Summarisation

Insights

Presents the pdf analysis as premium feature

PDF Analysis

Presents the zaia usage as premium feature

ZAIA

Global: Zendy Plus monthly

Global: Zendy Tools annual

Global: Zendy Tools monthly

Global: Zendy Free Plan

Motor imagery EEG classification plays a crucial role in non-invasiveBrain-Computer Interface (BCI) research. However, the classification isaffected by the non-stationarity and individual variations of EEG signals.Simply pooling EEG data with different statistical distributions to train aclassification model can severely degrade the generalization performance. Toaddress this issue, the existing methods primarily focus on domain adaptation,which requires access to the target data during training. This is unrealisticin many EEG application scenarios. In this paper, we propose a novelmulti-source domain generalization framework called EEG-DG, which leveragesmultiple source domains with different statistical distributions to buildgeneralizable models on unseen target EEG data. We optimize both the marginaland conditional distributions to ensure the stability of the joint distributionacross source domains and extend it to a multi-source domain generalizationframework to achieve domain-invariant feature representation, therebyalleviating calibration efforts. Systematic experiments on a simulative datasetand BCI competition datasets IV-2a and IV-2b demonstrate the superiority of ourproposed EEG-DG over state-of-the-art methods. Specifically, EEG-DG achieves anaverage classification accuracy/kappa value of 81.79%/0.7572 and 87.12%/0.7424on datasets IV-2a and IV-2b, respectively, which even outperforms some domainadaptation methods. Our code is available athttps://github.com/XC-ZhongHIT/EEG-DG for free download and evaluation.

EEG-DG: A Multi-Source Domain Generalization Framework for Motor Imagery EEG Classification