A Shallow Spiking Neural Network for P300 Classification

P300 is widely used in brain-computer interface (BCI) systems due to its high bit rates. Fast and accurate classification of P300 is the key to ensure the performance of P300-based brain-computer interface (BCI). In this paper, a shallow spiking neural network (SNN) model based on ANN-to-SNN method is proposed for P300 detection. Firstly, a shallow convolutional neural network (CNN) is pre-trained, and then its weight is transferred to the equivalent SNN for final recognition. The effect of the model is verified based on the public dataset provided by ITBA University. We compare the accuracy of CNN and SNN on the test dataset. The results show that equivalent SNN shows the comparable performance as the CNN, which proves the effectiveness.