Abstract
Electromyographic (EMG) signals provide information about muscle activity. In hand movements, each gesture's execution involves the activation of different combinations of the forearm muscles, which generate distinct electrical patterns. Furthermore, the analysis of muscle activation patterns represented by EMG signals allows recognizing these gestures. We aimed to develop an automatic hand gesture recognition system based on supervised Machine Learning (ML) techniques. We trained eight computational models to recognize six hand gestures and generalize between different subjects using raw data recordings of EMG signals from 36 subjects. We found that the random forest model and fully connected artificial neural network showed the best performances, indicated by 96.25% and 96.09% accuracy, respectively. These results improve on computational time and resources by avoiding data preprocessing operations and model generalization capabilities by including data from a larger number of subjects. In addition to the application in the health sector, in the context of Smart Cities, digital inclusion should be aimed at individuals with physical disabilities, with which, this model could contribute to the development of identification and interaction devices that can emulate the movement of hands.