EMPT: a sparsity Transformer for EEG-based motor imagery recognition

Ming Liu, Yanbing Liu, Weiyou Shi, Yitai Lou, Yuan Sun, Meng Qi, Dezheng Wang, Fangzhou Xu, Yang Zhang, Lei Zhang, Jiancai LengFrontiers in Neuroscience 2024 · DOI

Abstract

The MoE layer and ProbSparse Self-attention inside the EMPT are subjected to visualisation experiments. The experiments prove that sparsity can be introduced to the Transformer neural network by introducing MoE and kullback-leibler divergence attention pooling mechanism, thereby enhancing its applicability on EEG datasets. A novel deep learning approach is presented for decoding EEG data based on MI.
BibTeX
@inproceedings{liu2024empt,
  author = {Ming Liu and Yanbing Liu and Weiyou Shi and Yitai Lou and Yuan Sun and Meng Qi and Dezheng Wang and Fangzhou Xu and Yang Zhang and Lei Zhang and Jiancai Leng},
  title = {{EMPT: a sparsity Transformer for EEG-based motor imagery recognition}},
  booktitle = {Frontiers in Neuroscience},
  year = {2024},
  doi = {10.3389/fnins.2024.1366294},
}

Loading related content...

Figure