Domain-Generalized EEG Classification With Category-Oriented Feature Decorrelation and Cross-View Consistency Learning
Generalizing the electroencephalogram (EEG) decoding methods to unseen subjects is an important research direction for realizing practical application of brain-computer interfaces (BCIs). Since distribution shifts across subjects, the performance of most current deep neural networks for decoding EEG...
| Main Authors: | , , , , , , , |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
IEEE
2023-01-01
|
| Series: | IEEE Transactions on Neural Systems and Rehabilitation Engineering |
| Subjects: | |
| Online Access: | https://ieeexplore.ieee.org/document/10198467/ |
| _version_ | 1828717079627300864 |
|---|---|
| author | Shuang Liang Changsheng Xuan Wenlong Hang Baiying Lei Jun Wang Jing Qin Kup-Sze Choi Yu Zhang |
| author_facet | Shuang Liang Changsheng Xuan Wenlong Hang Baiying Lei Jun Wang Jing Qin Kup-Sze Choi Yu Zhang |
| author_sort | Shuang Liang |
| collection | DOAJ |
| description | Generalizing the electroencephalogram (EEG) decoding methods to unseen subjects is an important research direction for realizing practical application of brain-computer interfaces (BCIs). Since distribution shifts across subjects, the performance of most current deep neural networks for decoding EEG signals degrades when dealing with unseen subjects. Domain generalization (DG) aims to tackle this issue by learning invariant representations across subjects. To this end, we propose a novel domain-generalized EEG classification framework, named FDCL, to generalize EEG decoding through category-relevant and -irrelevant Feature Decorrelation and Cross-view invariant feature Learning. Specifically, we first devise data augmented regularization through mixing the segments of same-category features from multiple subjects, which increases the diversity of EEG data by spanning the space of subjects. Furthermore, we introduce feature decorrelation regularization to learn the weights of the augmented EEG trials to remove the dependencies between their features, so that the true mapping relationship between relevant features and corresponding labels can be better established. To further distill subject-invariant EEG feature representations, cross-view consistency learning regularization is introduced to encourage consistent predictions of category-relevant features induced from different augmented EEG views. We seamlessly integrate three complementary regularizations into a unified DG framework to jointly improve the generalizability and robustness of the model on unseen subjects. Experimental results on motor imagery (MI) based EEG datasets validate that the proposed FDCL outperforms the available state-of-the-art methods. |
| first_indexed | 2024-03-12T14:21:43Z |
| format | Article |
| id | doaj.art-b582883763964360a1668c63200f171f |
| institution | Directory Open Access Journal |
| issn | 1558-0210 |
| language | English |
| last_indexed | 2024-03-12T14:21:43Z |
| publishDate | 2023-01-01 |
| publisher | IEEE |
| record_format | Article |
| series | IEEE Transactions on Neural Systems and Rehabilitation Engineering |
| spelling | doaj.art-b582883763964360a1668c63200f171f2023-08-18T23:00:09ZengIEEEIEEE Transactions on Neural Systems and Rehabilitation Engineering1558-02102023-01-01313285329610.1109/TNSRE.2023.330096110198467Domain-Generalized EEG Classification With Category-Oriented Feature Decorrelation and Cross-View Consistency LearningShuang Liang0https://orcid.org/0000-0003-0305-7558Changsheng Xuan1Wenlong Hang2https://orcid.org/0000-0001-8029-3613Baiying Lei3https://orcid.org/0000-0002-3087-2550Jun Wang4https://orcid.org/0000-0001-9548-0411Jing Qin5https://orcid.org/0000-0002-2961-0860Kup-Sze Choi6https://orcid.org/0000-0003-0836-7088Yu Zhang7https://orcid.org/0000-0003-4087-6544School of Geographic and Biologic Information, Nanjing University of Posts and Telecommunications, Nanjing, ChinaCollege of Computer and Information Engineering and the College of Artificial Intelligence, Nanjing Tech University, Nanjing, ChinaCollege of Computer and Information Engineering and the College of Artificial Intelligence, Nanjing Tech University, Nanjing, ChinaHealth Science Center, School of Biomedical Engineering, Shenzhen University, Shenzhen, ChinaSchool of Communication and Information Engineering, Shanghai University, Shanghai, ChinaCentre for Smart Health, The Hong Kong Polytechnic University, Kowloon, Hong KongCentre for Smart Health, The Hong Kong Polytechnic University, Kowloon, Hong KongDepartment of Bioengineering and the Department of Electrical and Computer Engineering, Lehigh University, Bethlehem, PA, USAGeneralizing the electroencephalogram (EEG) decoding methods to unseen subjects is an important research direction for realizing practical application of brain-computer interfaces (BCIs). Since distribution shifts across subjects, the performance of most current deep neural networks for decoding EEG signals degrades when dealing with unseen subjects. Domain generalization (DG) aims to tackle this issue by learning invariant representations across subjects. To this end, we propose a novel domain-generalized EEG classification framework, named FDCL, to generalize EEG decoding through category-relevant and -irrelevant Feature Decorrelation and Cross-view invariant feature Learning. Specifically, we first devise data augmented regularization through mixing the segments of same-category features from multiple subjects, which increases the diversity of EEG data by spanning the space of subjects. Furthermore, we introduce feature decorrelation regularization to learn the weights of the augmented EEG trials to remove the dependencies between their features, so that the true mapping relationship between relevant features and corresponding labels can be better established. To further distill subject-invariant EEG feature representations, cross-view consistency learning regularization is introduced to encourage consistent predictions of category-relevant features induced from different augmented EEG views. We seamlessly integrate three complementary regularizations into a unified DG framework to jointly improve the generalizability and robustness of the model on unseen subjects. Experimental results on motor imagery (MI) based EEG datasets validate that the proposed FDCL outperforms the available state-of-the-art methods.https://ieeexplore.ieee.org/document/10198467/Electroencephalographdomain generalizationdata augmentationmotor imagery |
| spellingShingle | Shuang Liang Changsheng Xuan Wenlong Hang Baiying Lei Jun Wang Jing Qin Kup-Sze Choi Yu Zhang Domain-Generalized EEG Classification With Category-Oriented Feature Decorrelation and Cross-View Consistency Learning IEEE Transactions on Neural Systems and Rehabilitation Engineering Electroencephalograph domain generalization data augmentation motor imagery |
| title | Domain-Generalized EEG Classification With Category-Oriented Feature Decorrelation and Cross-View Consistency Learning |
| title_full | Domain-Generalized EEG Classification With Category-Oriented Feature Decorrelation and Cross-View Consistency Learning |
| title_fullStr | Domain-Generalized EEG Classification With Category-Oriented Feature Decorrelation and Cross-View Consistency Learning |
| title_full_unstemmed | Domain-Generalized EEG Classification With Category-Oriented Feature Decorrelation and Cross-View Consistency Learning |
| title_short | Domain-Generalized EEG Classification With Category-Oriented Feature Decorrelation and Cross-View Consistency Learning |
| title_sort | domain generalized eeg classification with category oriented feature decorrelation and cross view consistency learning |
| topic | Electroencephalograph domain generalization data augmentation motor imagery |
| url | https://ieeexplore.ieee.org/document/10198467/ |
| work_keys_str_mv | AT shuangliang domaingeneralizedeegclassificationwithcategoryorientedfeaturedecorrelationandcrossviewconsistencylearning AT changshengxuan domaingeneralizedeegclassificationwithcategoryorientedfeaturedecorrelationandcrossviewconsistencylearning AT wenlonghang domaingeneralizedeegclassificationwithcategoryorientedfeaturedecorrelationandcrossviewconsistencylearning AT baiyinglei domaingeneralizedeegclassificationwithcategoryorientedfeaturedecorrelationandcrossviewconsistencylearning AT junwang domaingeneralizedeegclassificationwithcategoryorientedfeaturedecorrelationandcrossviewconsistencylearning AT jingqin domaingeneralizedeegclassificationwithcategoryorientedfeaturedecorrelationandcrossviewconsistencylearning AT kupszechoi domaingeneralizedeegclassificationwithcategoryorientedfeaturedecorrelationandcrossviewconsistencylearning AT yuzhang domaingeneralizedeegclassificationwithcategoryorientedfeaturedecorrelationandcrossviewconsistencylearning |