Pattern recognition of surface electromyography based on multi-scale convolutional neural network with attention mechanism
by Beibei Wang; Hui Zheng; Jing Jie; Miao Zhang; Yintao Ke; Yang Liu
International Journal of Wireless and Mobile Computing (IJWMC), Vol. 23, No. 3/4, 2022

Abstract: Natural control methods based on Surface Electromyography (sEMG) pattern recognition have been widely applied in the field of hand prostheses. However, the control robustness and accuracy are difficult to meet many real-life applications. This paper proposes a Multi-Scale Convolutional Neural Network (MSCNN) model based on the attention mechanism, which can automatically learn gesture features through convolution. The model generates features through convolution kernels of different sizes to achieve the fusion of features of different degrees firstly. After that, the attention mechanism is used to calculate the weights of different scales, and then the fused comprehensive features are obtained. The proposed model has been verified on the SIA_delsys_16_movement and NinaPro data sets. The experimental results showed that the proposed model has better classification accuracy, and the attention mechanism can validly improve the classification performance of the convolutional neural network.

Online publication date: Mon, 12-Dec-2022

The full text of this article is only available to individual subscribers or to users at subscribing institutions.

 
Existing subscribers:
Go to Inderscience Online Journals to access the Full Text of this article.

Pay per view:
If you are not a subscriber and you just want to read the full contents of this article, buy online access here.

Complimentary Subscribers, Editors or Members of the Editorial Board of the International Journal of Wireless and Mobile Computing (IJWMC):
Login with your Inderscience username and password:

Username:        Password:          Forgotten your password?

Want to subscribe?
A subscription gives you complete access to all articles in the current issue, as well as to all articles in the previous three years (where applicable). See our Orders page to subscribe.

If you still need assistance, please email [email protected]