Zero-filling frequency domain convolutional neural network method suitable for SSVEP classification

Abstract

The invention discloses a zero-filling frequency domain convolutional neural network method suitable for SSVEP classification, is applied to the field of electroencephalogram signal processing, and aims to solve the problems of low classification accuracy and low information transmission rate of SSVEP electroencephalogram signals of SSVEP and SSVEP subclasses in the prior art. The method comprises the following steps: firstly, acquiring SSVEP electroencephalogram signals of nine electrodes of the occipital part of the brain of a human body by using electroencephalogram acquisition equipment; secondly, preprocessing the electroencephalogram signals; performing zero filling on the preprocessed electroencephalogram signals in a time domain, and extracting interested fundamental frequency bands and second harmonic frequency bands in the power spectral density of the SSVEP signals of nine channels to be combined into a feature matrix; and finally, taking the feature matrix as the input of a CNN deep learning model, and identifying different types of SSVEP signals by using nonlinear transformation. By adopting the method provided by the invention, relatively high classification accuracy can be obtained.

Description

technical field [0001] The invention belongs to the field of electroencephalogram signal processing, in particular to an electroencephalogram signal classification technology. Background technique [0002] Brain Computer Interface (BCI) is a communication system. It converts the "ideas" in the brain into instructions, so that the human brain can directly transmit instructions to the designated machine terminal. Express intentions and ideas directly, or manipulate machinery and equipment without words or actions. In the past few decades, among the various modes of BCI, the BCI of Steady-State Visual Evoked Potentials (SSVEP) realized by electroencephalogram (EEG) has a high information transmission rate. (ITR), high signal-to-noise ratio (SNR), less training time and reliability have been widely concerned and researched. It has been widely used in many fields such as rehabilitation for the disabled, entertainment experience, etc., and has made great contributions to improv...

AI Technical Summary

Problems solved by technology

However, these deep learning methods have neither been studied on a large number of stimulus targets nor validated on standard public datasets

And in terms of deep learning, how to improve the classification accuracy of SSVEP's brain-computer interface still needs to be studied.

Images