A master’s thesis by researcher Duha Mahlal Ajimi was defended at the Department of Computer Engineering, College of Engineering, University of Basrah, under the supervision of Dr. Hanadi Abbas Jaber. The thesis is titled... Electromyographic Hand Gesture Recognition Using Deep Learning Techniques

Hand gesture recognition using surface electromyography (sEMG) has become a promising research area because of its usage in intelligent prosthetic hand control and human-machine interaction (HMI). Nevertheless, there are several difficulties connected with the recognition of gestures using such signals due to the non-stationarity of EMG. In this thesis, we propose a simple framework based on Convolutional Neural Networks (CNN) for HD-sEMG hand gesture recognition. Three public datasets from the CapgMyo database (DB-a, DB-b, and DB-c) were used to validate the proposed framework. Two time-frequency analysis methods, namely, the Short-Time Fourier Transform (STFT) and the Continuous Wavelet Transform (CWT) are used for discriminant features extraction from raw HD-sEMG signals before classification. Time-frequency representations generated by these methods are used as input data to our CNN-based model for hand gesture recognition. The results of the proposed approach are analyzed with several classification criteria. The obtained results prove the efficiency of time-frequency analysis and deep learning combination for hand gesture recognition using HD-sEMG