Journal
IEEE TRANSACTIONS ON COGNITIVE AND DEVELOPMENTAL SYSTEMS
Volume 14, Issue 1, Pages 4-19Publisher
IEEE-INST ELECTRICAL ELECTRONICS ENGINEERS INC
DOI: 10.1109/TCDS.2020.3007453
Keywords
Electroencephalography; Task analysis; Calibration; Brain modeling; Machine learning; Visualization; Probability distribution; Adversarial attacks; affective brain-computer interface (BCI); brain-computer interfaces; domain adaptation; electroencephalogram (EEG); transfer learning (TL)
Funding
- Hubei Province Funds for Distinguished Young Scholors [2020CFA050]
- Hubei Technology Innovation Platform [2019AEA171]
- National Natural Science Foundation of China [61873321, U1913207, 61673266, 61976135]
- International Science and Technology Cooperation Program of China [2017YFE0128300]
- National Key Research and Development Program of China [2017YFB1002501]
- SJTU Transmed Awards Research [WF540162605]
- Fundamental Research Funds for the Central Universities
- 111 Project
Ask authors/readers for more resources
This article reviews journal publications on transfer learning approaches in EEG-based BCIs since 2016. TL methods applied to different paradigms and applications, such as motor imagery and event-related potentials, are reviewed in the context of cross-subject/session and cross-device/task settings. Observations and conclusions provide insights for future research directions.
A brain-computer interface (BCI) enables a user to communicate with a computer directly using brain signals. The most common noninvasive BCI modality, electroencephalogram (EEG), is sensitive to noise/artifact and suffers between-subject/within-subject nonstationarity. Therefore, it is difficult to build a generic pattern recognition model in an EEG-based BCI system that is optimal for different subjects, during different sessions, for different devices and tasks. Usually, a calibration session is needed to collect some training data for a new subject, which is time consuming and user unfriendly. Transfer learning (TL), which utilizes data or knowledge from similar or relevant subjects/sessions/devices/tasks to facilitate learning for a new subject/session/device/task, is frequently used to reduce the amount of calibration effort. This article reviews journal publications on TL approaches in EEG-based BCIs in the last few years, i.e., since 2016. Six paradigms and applications-motor imagery, event-related potentials, steady-state visual evoked potentials, affective BCIs, regression problems, and adversarial attacks-are considered. For each paradigm/application, we group the TL approaches into cross-subject/session, cross-device, and cross-task settings and review them separately. Observations and conclusions are made at the end of the article, which may point to future research directions.
Authors
I am an author on this paper
Click your name to claim this paper and add it to your profile.
Reviews
Recommended
No Data Available