EEG temporal-spatial transformer for person identification.

An increasing number of studies have been devoted to electroencephalogram (EEG) identity recognition since EEG signals are not easily stolen. Most of the existing studies on EEG person identification have only addressed brain signals in a single state, depending upon specific and repetitive sensory stimuli. However, in reality, human states are diverse and rapidly changing, which limits their practicality in realistic settings. Among many potential solutions, transformer is widely used and achieves an excellent performance in natural language processing, which demonstrates the outstanding ability of the attention mechanism to model temporal signals. In this paper, we propose a transformer-based approach for the EEG person identification task that extracts features in the temporal and spatial domains using a self-attention mechanism. We conduct an extensive study to evaluate the generalization ability of the proposed method among different states. Our method is compared with the most advanced EEG biometrics techniques and the results show that our method reaches state-of-the-art results. Notably, we do not need to extract any features manually.