EEG-TCNet: An Accurate Temporal Convolutional Network for Embedded Motor-Imagery Brain–Machine Interfaces

@article{Ingolfsson2020EEGTCNetAA,
  title={EEG-TCNet: An Accurate Temporal Convolutional Network for Embedded Motor-Imagery Brain–Machine Interfaces},
  author={Thorir Mar Ingolfsson and Michael Hersche and Xiaying Wang and Nobuaki Kobayashi and Lukas Cavigelli and Luca Benini},
  journal={2020 IEEE International Conference on Systems, Man, and Cybernetics (SMC)},
  year={2020},
  pages={2958-2965}
}
In recent years, deep learning (DL) has contributed significantly to the improvement of motor-imagery brain–machine interfaces (MI-BMIs) based on electroencephalography (EEG). While achieving high classification accuracy, DL models have also grown in size, requiring a vast amount of memory and computational resources. This poses a major challenge to an embedded BMI solution that guarantees user privacy, reduced latency, and low power consumption by processing the data locally. In this paper, we… 
View on IEEE
research-collection.ethz.ch
30 Citations
Highly Influential Citations
10
Background Citations
19
Methods Citations
13
Results Citations
4

Figures and Tables from this paper

30 Citations

Physics-Informed Attention Temporal Convolutional Network for EEG-Based Motor Imagery Classification

The proposed ATCNet model outperforms the current state-of-the-art techniques in the BCI Competition IV-2a dataset with an accuracy of 85.38% and 70.97% for the subject-dependent and subject-independent modes, respectively.

MI-BMInet: An Efficient Convolutional Neural Network for Motor Imagery Brain--Machine Interfaces with EEG Channel Selection

An efficient convolutional neural network (CNN) for EEG-based MI classi-cation that achieves comparable accuracy while being orders of magnitude less resource-demanding and significantly more energy-efficient than state-of-the-art (SoA) models for a long-lifetime battery operation is proposed.

EEG-ITNet: An Explainable Inception Temporal Convolutional Network for motor imagery classification

An end-to-end deep learning architecture called EEG-ITNet is proposed and a more comprehensible method to visualise the network learned patterns is proposed to comprehensively explain and support the validity of network illustration from a neuroscientific perspective.

Electroencephalography-based motor imagery classification using temporal convolutional network fusion

A Multi-Branch Convolutional Neural Network with Squeeze-and-Excitation Attention Blocks for EEG-Based Motor Imagery Signals Classification

This study introduces a new model for decoding MI known as a Multi-Branch EEGNet with squeeze-and-excitation blocks (MBEEGSE), a multi-branch CNN model with attention blocks is employed to adaptively change channel-wise feature responses.

Rethinking CNN Architecture for Enhancing Decoding Performance of Motor Imagery-Based EEG Signals

A novel model, called M–ShallowConvNet, is proposed, which solves the existing problems of the conventional model and demonstrates that performance improvement can be achieved with only a few small modifications that resolve the problems ofThe conventional model.

REEG-BTCNet: A Novel Framework for EEG-based Motor Imagery Classification

  • Jiyao LiuHuifu Li
  • Computer Science
    2022 IEEE International Conference on Bioinformatics and Biomedicine (BIBM)
  • 2022
The proposed REEG-BTCNet is a novel framework that achieves outstanding accuracy with stronger robustness for EEG-based motor imagery classification and outperforms the current state-of-the-art techniques in the BCI Competition IV-2a dataset.

A Multibranch of Convolutional Neural Network Models for Electroencephalogram-Based Motor Imagery Classification

The proposed multibranch CNN models outperformed other state-of-the-art EEG motor imagery classification methods and improved the accuracy of a single-scale network by 6.84%.

Sub-100uW Multispectral Riemannian Classification for EEG-based Brain--Machine Interfaces

This work proposes a multispectral Riemannian classifier that sets the new state-of-the-art in terms of accuracy-energy trade-off for near-sensor classification in embedded MI-BMIs.

MAtt: A Manifold Attention Network for EEG Decoding

The proposed manifold attention network (mAtt), a novel geometric deep learning (GDL)-based model, featuring a manifold attention mechanism that characterizes spatiotemporal representations of EEG data fully on a Riemannian symmetric positive definite (SPD) manifold suggests its superiority over other leading DL methods for general EEG decoding.

References

SHOWING 1-10 OF 32 REFERENCES

An Accurate EEGNet-based Motor-Imagery Brain–Computer Interface for Low-Power Edge Computing

This paper presents an accurate and robust embedded motor-imagery brain–computer interface (MI-BCI), based on EEGNet, that matches the requirements of memory footprint and computational resources of low-power microcontroller units (MCUs), such as the ARM Cortex-M family.

A Parallel Multiscale Filter Bank Convolutional Neural Networks for Motor Imagery EEG Classification

A layered end-to-end network structure, in which a feature-extraction network is used to extract temporal and spatial features, and the results suggest that the proposed network has superior performance, robustness, and transfer learning ability.

Densely Feature Fusion Based on Convolutional Neural Networks for Motor Imagery EEG Classification

A densely feature fusion convolutional neural networks (DFFN) framework is proposed, which considers the correlation between adjacent layers and cross layer features, which reduces the information loss in the process of Convolutional operation and considers the local and global characteristics of the network.

Deep Learning Solutions for Motor Imagery Classification: A Comparison Study

  • N. LuTao YinXue Jing
  • Computer Science
    2020 8th International Winter Conference on Brain-Computer Interface (BCI)
  • 2020
Experiments have shown that TCN based method has obtaining the best performance and the paralleled combination of CNN and RNN has obtained the second best performance, which inspired us to explore the spatial-temporal feature learning deep network solutions for further improvement.

EEGNet: a compact convolutional neural network for EEG-based brain–computer interfaces

This work introduces EEGNet, a compact convolutional neural network for EEG-based BCIs, and introduces the use of depthwise and separable convolutions to construct an EEG-specific model which encapsulates well-known EEG feature extraction concepts for BCI.

On the improvement of classifying EEG recordings using neural networks

A significant improvement of classification accuracy is presented on classifying electroencephalography recordings using deep learning on the Brain-Computer Interfaces Competition IV dataset, and the results of various state of the art neural network structures are compared.

Applying Transfer Learning To Deep Learned Models For EEG Analysis

This project proposes a method that can produce reliable electroencephalography (EEG) signal classification, based on modest amounts of training data through the use of transfer learning, and explores transferability of knowledge between trained models over different experiments, called inter-experimental transfer learning.

A Novel MI-EEG Imaging With the Location Information of Electrodes

The experimental results show that the effectiveness of proposed MI-EEG imaging method, and it is well-matched with mVGG, has lower class skew and error costs.

Classification of Hand Motions within EEG Signals for Non-Invasive BCI-Based Robot Hand Control

This study analyzed the decoding of five different hand executions and imageries from EEG signals, for a robot hand control, using the common spatial patterns and the regularized linear discriminant analysis for the data analysis.

A review of classification algorithms for EEG-based brain–computer interfaces: a 10 year update

This paper provides a comprehensive overview of the modern classification algorithms used in EEG-based BCIs, presents the principles of these methods and guidelines on when and how to use them, and identifies a number of challenges to further advance EEG classification in BCI.