P300 event-related potential detection using one-dimensional convolutional capsule networks

作者：

Highlights：

• A new method to P300 ERP detection called 1D-CapsNet.

• The improvement of the convolution dimension allows better detection of P300 ERP.

• Two classifiers based on 1D-CapsNet are proposed to better transform BCI.

• The performance of 1D-CapsNet is better than other state-of-the-art algorithms.

摘要

•A new method to P300 ERP detection called 1D-CapsNet.•The improvement of the convolution dimension allows better detection of P300 ERP.•Two classifiers based on 1D-CapsNet are proposed to better transform BCI.•The performance of 1D-CapsNet is better than other state-of-the-art algorithms.

论文关键词：Brain-computer interface,EEG classification,Features extraction,Ergonomics,Capsule network

论文评审过程：Received 30 October 2020, Revised 9 January 2021, Accepted 6 February 2021, Available online 12 February 2021, Version of Record 6 March 2021.

论文官网地址：https://doi.org/10.1016/j.eswa.2021.114701