EMG SIGNALS FOR FINGER MOVEMENT CLASSIFICATION BASED ON SHORT-TERM FOURIER TRANSFORM AND DEEP LEARNING

• DOI: 10.14311/ctj.2021.1.02
• Publication Date: 2022-12-19T12:49:45Z
• Authors: Ivana Kralikova; Branko Babusiak; Lubomir Kralik
• Keywords: Spectrogram; Interface (matter)

An interface based on electromyographic (EMG) signals is considered one of the central fields in human-machine (HCI) research with broad practical use. This paper presents recognition 13 individual finger movements time-frequency representation EMG via spectrograms. A deep learning algorithm, namely a convolutional neural network (CNN), used to extract features and classify them. Two approaches data representations are investigated: different window segmentation lengths reduction measured channels. The overall highest accuracy classification reaches 95.5% for segment length 300 ms. average attains more than 90% by reducing channels from four three.