Alignment-Based Adversarial Training (ABAT) for Improving the Robustness and Accuracy of EEG-Based BCIs

• Published in: IEEE transactions on neural systems and rehabilitation engineering Vol. 32; pp. 1703 - 1714
• Main Authors: Chen, Xiaoqing, Wang, Ziwei, Wu, Dongrui
• Format: Journal Article
• Language: English
• Published: United States IEEE 2024; The Institute of Electrical and Electronics Engineers, Inc. (IEEE)
• Subjects: Accuracy; adversarial attack; adversarial training; Algorithms; Alignment; Artificial neural networks; Brain modeling; brain-computer interface; Brain-Computer Interfaces; Classification; Classifiers; Computer vision; Covariance matrices; data alignment; Data models; EEG; Electroencephalogram; Electroencephalography; Electroencephalography - methods; Evoked Potentials - physiology; Human-computer interface; Humans; Image classification; Imagination - physiology; Interfaces; Machine Learning; Mental task performance; Model accuracy; Neural networks; Neural Networks, Computer; Perturbation methods; Robustness; Training

Machine learning has achieved great success in electroencephalogram (EEG) based brain-computer interfaces (BCIs). Most existing BCI studies focused on improving the decoding accuracy, with only a few considering the adversarial security. Although many adversarial defense approaches have been propose...