Deep Feature Mining via Attention-based BiLSTM-GCN for Human Motor Imagery Recognition
This work addresses the need for practical EEG-based brain-computer interfaces by enhancing recognition performance, though it appears incremental as it combines existing deep learning components.
The paper tackled the problem of improving accuracy and response time in EEG-based motor imagery recognition for brain-computer interfaces, achieving 98.81% and 94.64% accuracy with a 0.4-second detection framework that outperformed state-of-the-art methods.
Recognition accuracy and response time are both critically essential ahead of building practical electroencephalography (EEG) based brain-computer interface (BCI). Recent approaches, however, have either compromised in the classification accuracy or responding time. This paper presents a novel deep learning approach designed towards remarkably accurate and responsive motor imagery (MI) recognition based on scalp EEG. Bidirectional Long Short-term Memory (BiLSTM) with the Attention mechanism manages to derive relevant features from raw EEG signals. The connected graph convolutional neural network (GCN) promotes the decoding performance by cooperating with the topological structure of features, which are estimated from the overall data. The 0.4-second detection framework has shown effective and efficient prediction based on individual and group-wise training, with 98.81% and 94.64% accuracy, respectively, which outperformed all the state-of-the-art studies. The introduced deep feature mining approach can precisely recognize human motion intents from raw EEG signals, which paves the road to translate the EEG based MI recognition to practical BCI systems.