Does Meta-Learning Improve EEG Motor Imagery Classification?

Xiaoli Wu; Rosa H. M. Chan

doi:10.1109/EMBC48229.2022.9871035

Does Meta-Learning Improve EEG Motor Imagery Classification?

^*Corresponding author for this work

Department of Electrical Engineering

Research output: Chapters, Conference Papers, Creative and Literary Works › RGC 32 - Refereed conference paper (with host publication) › peer-review

13 Citations (Scopus)

Abstract

Deep learning has been applied to enhance the performance of EEG-based brain-computer interface applications. However, the cross-subject variations in EEG signals cause domain shifts and negatively affect the model performance and generalization. Meta-learning algorithms have shown fast new domain adaption in various fields, which may help solve the domain shift problems in EEG. Reptile, with satisfactory performance and low computational costs, stands out from other existing meta-learning algorithms. We integrated Reptile with a deep neural network as Reptile-EEG for the EEG motor imagery tasks, and compared Reptile-EEG with other state-of-the-art models in three motor imagery BCI benchmark datasets. Results show that Reptile-EEGdoes not outperform simple training of deep neural networks in motor imagery BCI tasks.

Original language	English
Title of host publication	2022 44th Annual International Conference of the IEEE Engineering in Medicine and Biology Society (EMBC)
Place of Publication	978-1-7281-2783-5
Publisher	IEEE
Pages	4048-4051
ISBN (Electronic)	9781728127828
DOIs	https://doi.org/10.1109/EMBC48229.2022.9871035
Publication status	Published - 2022
Event	44th Annual International Conference of the IEEE Engineering in Medicine and Biology Society (EMBC 2022): “Biomedical Engineering transforming the provision of healthcare: promoting wellness through personalized & predictable provision at the point of care” - Scottish Event Campus (SEC) Centre, Glasgow, United Kingdom Duration: 11 Jul 2022 → 15 Jul 2022 https://embc.embs.org/2022/

Publication series

Name	Proceedings of the Annual International Conference of the IEEE Engineering in Medicine and Biology Society, EMBS
ISSN (Print)	1557-170X
ISSN (Electronic)	2694-0604

Conference

Conference	44th Annual International Conference of the IEEE Engineering in Medicine and Biology Society (EMBC 2022)
Place	United Kingdom
City	Glasgow
Period	11/07/22 → 15/07/22
Internet address	https://embc.embs.org/2022/

Funding

This work was supported in part by a grant from the Research Grants Council of the Hong Kong Special Administrative Region, China, under Project CityU 11215618, and by a grant from Guangdong-Hong KongMacao Greater Bay Area Center for Brain Science and Brain-Inspired Intelligence, Guangzhou 510515, P. R. China, by Key-Area Research and Development Program of Guangdong Province, 2018B030340001.

RGC Funding Information

RGC-funded

Access Output

10.1109/EMBC48229.2022.9871035

Other Access Options

Check CityUHK Library

Permanent Link

Right click to copy link

Cite this

Wu, X., & Chan, R. H. M. (2022). Does Meta-Learning Improve EEG Motor Imagery Classification? In 2022 44th Annual International Conference of the IEEE Engineering in Medicine and Biology Society (EMBC) (pp. 4048-4051). (Proceedings of the Annual International Conference of the IEEE Engineering in Medicine and Biology Society, EMBS). IEEE. https://doi.org/10.1109/EMBC48229.2022.9871035

@inproceedings{08661b4b53ea452a8f979b908cb9caae,

title = "Does Meta-Learning Improve EEG Motor Imagery Classification?",

abstract = "Deep learning has been applied to enhance the performance of EEG-based brain-computer interface applications. However, the cross-subject variations in EEG signals cause domain shifts and negatively affect the model performance and generalization. Meta-learning algorithms have shown fast new domain adaption in various fields, which may help solve the domain shift problems in EEG. Reptile, with satisfactory performance and low computational costs, stands out from other existing meta-learning algorithms. We integrated Reptile with a deep neural network as Reptile-EEG for the EEG motor imagery tasks, and compared Reptile-EEG with other state-of-the-art models in three motor imagery BCI benchmark datasets. Results show that Reptile-EEGdoes not outperform simple training of deep neural networks in motor imagery BCI tasks.",

author = "Xiaoli Wu and Chan, {Rosa H. M.}",

year = "2022",

doi = "10.1109/EMBC48229.2022.9871035",

language = "English",

series = "Proceedings of the Annual International Conference of the IEEE Engineering in Medicine and Biology Society, EMBS",

publisher = "IEEE",

pages = "4048--4051",

booktitle = "2022 44th Annual International Conference of the IEEE Engineering in Medicine and Biology Society (EMBC)",

address = "United States",

note = "44th Annual International Conference of the IEEE Engineering in Medicine and Biology Society (EMBC 2022) : “Biomedical Engineering transforming the provision of healthcare: promoting wellness through personalized & predictable provision at the point of care” ; Conference date: 11-07-2022 Through 15-07-2022",

url = "https://embc.embs.org/2022/",

}

Wu, X & Chan, RHM 2022, Does Meta-Learning Improve EEG Motor Imagery Classification? in 2022 44th Annual International Conference of the IEEE Engineering in Medicine and Biology Society (EMBC). Proceedings of the Annual International Conference of the IEEE Engineering in Medicine and Biology Society, EMBS, IEEE, 978-1-7281-2783-5, pp. 4048-4051, 44th Annual International Conference of the IEEE Engineering in Medicine and Biology Society (EMBC 2022), Glasgow, United Kingdom, 11/07/22. https://doi.org/10.1109/EMBC48229.2022.9871035

Does Meta-Learning Improve EEG Motor Imagery Classification? / Wu, Xiaoli; Chan, Rosa H. M.
2022 44th Annual International Conference of the IEEE Engineering in Medicine and Biology Society (EMBC). 978-1-7281-2783-5: IEEE, 2022. p. 4048-4051 (Proceedings of the Annual International Conference of the IEEE Engineering in Medicine and Biology Society, EMBS).

Research output: Chapters, Conference Papers, Creative and Literary Works › RGC 32 - Refereed conference paper (with host publication) › peer-review