Abstract:Background and objective: The identification of motor intention by analyzing electroencephalogram (EEG) signals has become an important issue in brain-computer interface (BCI) applications. Since the individual differences of EEG are large, it's hard to design a universal BCI system, which means the labeled data from current user have to be collected for calibration. We propose a multi-source fusion transfer learning algorithm for this calibration, which only needs a small amount of new data from the current user. Methods: The core idea of proposed algorithm is a two-step calibration based on transfer learning, which works on the levels of both subjects and features. The spatial covariance matrices represented in the Riemannian space are used to characterize the multi-channel EEG signals. Source subjects whose features are similar to the current user are selected as the transfer source via a Riemannian geometry alignment algorithm. Their features obtained from the Riemannian tangent space, as well as the ones of the current user, are fused and calibrated by a balanced distribution adaptation algorithm. Finally, a classification model could be trained for the current user with only a few new training data. Results: Validated on both the online ERD-BCI database and the BCI Competition IV 2a database for motor imagery tasks, experimental results show that our method outperforms the state-of-the-art methods. When the new training data from current user is insufficient, the classification accuracy and kappa values compared with the outstanding counterparts have advantages of 3.30 % and 0.05 respectively on the online ERD-BCI database, 3.95 % and 0.03 respectively on the BCI Competition IV 2a database. Even when the new training data from current user is sufficient, the classification accuracy and kappa values of our proposed method also outperforms the counterparts with highest performance by 2.37 % and 0.05 respectively on the online ERD-BCI database, 2.25 % and 0.03 respectively on the BCI Competition IV 2a database. Conclusions: The algorithm provides a new idea of solution for calibrating EEG features. It is expected to contribute to the practical applications of BCI.

Source Data Selection for Brain-Computer Interfaces based on Simple Features

Optimization of transfer learning based on source sample selection in Euclidean space for P300-based brain-computer interfaces

Calibrating EEG features in motor imagery classification tasks with a small amount of current data using multisource fusion transfer learning

An inter-subject model to reduce the calibration time for motion imagination-based brain-computer interface

A novel dual-step transfer framework based on domain selection and feature alignment for motor imagery decoding

A cross-dataset adaptive domain selection transfer learning framework for motor imagery-based brain-computer interfaces

Minimizing subject-dependent calibration for BCI with Riemannian transfer learning

Transfer Learning Based on Regularized Common Spatial Patterns Using Cosine Similarities of Spatial Filters for Motor-Imagery BCI

Multi-layer transfer learning algorithm based on improved common spatial pattern for brain-computer interfaces

Transfer learning with data alignment and optimal transport for EEG based motor imagery classification

Transfer Learning Enhanced Common Spatial Pattern Filtering for Brain Computer Interfaces (BCIs): Overview and a New Approach

Domain Adaptive Algorithm Based on Multi-Manifold Embedded Distributed Alignment for Brain-Computer Interfaces

Unsupervised Multi-Source Variational Domain Adaptation for Inter-Subject SSVEP-based BCIs

Automatic feature selection for sensorimotor rhythms brain-computer interface fusing expert and data-driven knowledge

Suppression of negative transfer in motor imagery brain–computer interface based on mutual information and Pearson correlation coefficient

Transferring BCI models from calibration to control: Observing shifts in EEG features

SCDAN: Learning Common Feature Representation of Brain Activation for Intersubject Motor Imagery EEG Decoding

Cross-Subject Transfer Method Based on Domain Generalization for Facilitating Calibration of SSVEP-Based BCIs

Mutual information-based feature selection for low-cost BCIs based on motor imagery

Parameter transfer and Riemannian space coordinate alignment for EEG intention recognition

Riemannian Geometric Instance Filtering for Transfer Learning in Brain-Computer Interfaces.