EEG Signal Classification Based on WPT-SPCNN and Distributed Virtual Reality
ZHAO Jiechen
WU Hao
Abstract：Current EEG signal classification methods struggle to balance the time-frequency local characteristics and dynamic spatiotemporal representations of signals in feature extraction and pattern discrimination,limiting further improvements in classification performance.This study aims to achieve efficient classification of motor imagery EEG signals and multi-user virtual interaction,proposing a method based on wavelet packet transform-serial-parallel convolutional neural network(WPT-SPCNN)and distributed virtual reality technology.First,multi-channel EEG signals are preprocessed;key frequency bands are decomposed and reconstructed using wavelet packet transform(WPT),and the results are then input into the serial-parallel convolutional neural network(SPCNN)for classification.Finally,distributed virtual reality technology and collision detection technology are integrated to construct an interactive system.Results show that the WPT-SPCNN has an average time consumption of 42.6ms,a compression rate of 82.7％,an average classification accuracy exceeding 84％,and a maximum accuracy of 98％.Additionally,the average synchronization delay of distributed virtual reality scenes is 18.3ms,with an average command execution success rate of 96.4％,providing an effective solution for multi-user EEG-controlled virtual interaction.
Keywords：EEG signal classificationWPTSPCNNVirtual Reality
Publication Date：2025-11-20
Online Publishing Date：2025-12-03(First online date of this platform, not the publication date of the document)
Pages：6( 59-64 )