EMG-Based Hand Gesture Recognition: Comparative Evaluation of Machine Learning Models

Abstract

Hand gesture recognition is becoming increasingly crucial for enhancing interactions in a variety of fields, such as human-computer interfaces, rehabilitation, and prosthetics. By using Electromyography sensors to track forearm muscle activity during hand gestures, this study tackles the need for precise gesture recognition. The main objective is to use a variety of machine learning algorithms to analyze these EMG signals and categorize various hand gestures. The study intends to efficiently classify gestures based on the recorded muscle activity by utilizing deep learning techniques in conjunction with conventional algorithms like Random Forest and k-Nearest Neighbors. This research contributes to the development of sophisticated human-computer interaction systems, more efficient rehabilitation tools, and advanced prosthetic devices by increasing the accuracy of gesture classification. The goal of this research is to progress the field of gesture recognition by advocating for safer and easier-to-use technologies that enhance the user experience across a range of applications.