Classification of hand motions in EEG signals using recurrent neural networks

This paper describes hand motion detection and the method for classification of 32-component EEG signals. This method is based on using recurrent convolution neural network as multi-class classifier. In this paper, we propose and empirically evaluate several architectures of recurrent convolutional neural network, and show advantages of using recurrent convolutional neural network for investigating problem. The results prove that this type of classifier can effectively distinguish characteristic features in the initial EEG signals and provide correct values of neural network outputs. Using recurrent convolution layer instead of the standard convolution layer can significantly improve the quality of classification. Adding recurrent connections for convolutional layer neurons increases the depth of the network, maintaining a constant number of parameters by weight sharing between layers.

[1]  S. Adamovich,et al.  Analysis of a commercial EEG device for the control of a robot arm , 2010, Proceedings of the 2010 IEEE 36th Annual Northeast Bioengineering Conference (NEBEC).

[2]  R. Johansson,et al.  Coordinated isometric muscle commands adequately and erroneously programmed for the weight during lifting task with precision grip , 2004, Experimental Brain Research.

[3]  Charu C. Aggarwal,et al.  Neural Networks and Deep Learning , 2018, Springer International Publishing.

[4]  P J Webros.  BACKPROPAGATION THROUGH TIME: WHAT IT DOES AND HOW TO DO IT , 1990 .

[5]  Geoffrey E. Hinton,et al.  ImageNet classification with deep convolutional neural networks , 2012, Commun. ACM.

[6]  Guigang Zhang,et al.  Deep Learning , 2016, Int. J. Semantic Comput..

[7]  Matthew D. Zeiler.  ADADELTA: An Adaptive Learning Rate Method , 2012, ArXiv.

[8]  R. Johansson,et al.  Programmed and triggered actions to rapid load changes during precision grip , 2004, Experimental Brain Research.

[9]  Xiaolin Hu,et al.  Recurrent convolutional neural network for object recognition , 2015, 2015 IEEE Conference on Computer Vision and Pattern Recognition (CVPR).

[10]  Yoshua Bengio,et al.  Gradient-based learning applied to document recognition , 1998, Proc. IEEE.

[11]  Ewa Jarocka,et al.  Multi-channel EEG recordings during 3,936 grasp and lift trials with varying weight and friction , 2014, Scientific Data.

[12]  H. Jasper,et al.  The ten-twenty electrode system of the International Federation. The International Federation of Clinical Neurophysiology. , 1999, Electroencephalography and clinical neurophysiology. Supplement.

[13]  Yann LeCun,et al.  Convolutional networks and applications in vision , 2010, Proceedings of 2010 IEEE International Symposium on Circuits and Systems.

[14]  R. Johansson,et al.  Factors influencing the force control during precision grip , 2004, Experimental Brain Research.

[15]  Kunihiko Fukushima,et al.  Neocognitron: A self-organizing neural network model for a mechanism of pattern recognition unaffected by shift in position , 1980, Biological Cybernetics.