SSVEP Analysis Toolbox

This repository provides a python package for SSVEP datasets and recognition algorithms. The goal of this toolbox is to make researchers be familier with SSVEP signals and related recognition algorithms quickly, and focus on improving algorithms with shortest preparation time.

Most conventional recognition algorithms are implemented using both eigen decomposition and least-square unified framework. The least-square unified framework demonstrates various design strategies applied in the correlatiion analysis (CA)-based SSVEP spatial filtering algorithms and their relationships.

The lease-square unified framework has been published in IEEE TNSRE: https://ieeexplore.ieee.org/document/10587150/. If you use this toolbox, the lease-square unified framework, or the new spatial filtering methods (ms-MsetCCA-R-1, ms-MsetCCA-R-2, ms-MsetCCA-R-3, ms-(e)TRCA-R-1, ms-(e)TRCA-R-2), please cite this paper.

@article{LSFrameWork,
  title = {A least-square unified framework for spatial filtering in {SSVEP}-based {BCIs}},
  volume = {32},
  url = {https://ieeexplore.ieee.org/document/10587150/},
  doi = {10.1109/TNSRE.2024.3424410},
  journal = {IEEE Trans. Neural Syst. Rehabil. Eng.},
  author = {Wang, Ze and Shen, Lu and Yang, Yi and Ma, Yueqi and Wong, Chi Man and Liu, Zige and Lin, Cuiyun and Hon, Chi Tin and Qian, Tao and Wan, Feng},
  year = {2024},
  pages = {2470--2481},
}

Features

  • Mutiple implementations of various algorithms: + Eigen decomposition + Least-square unified framework

  • Unify formats of SSVEP datasets

  • Provide a standard processing procedure for faire performance comparisons

  • Python implementations of SSVEP recognition algorithms

Datasets and Algorithms