Navigating Attention-Centric: A Machine Learning Approach to EMG-Based Hand Gesture Recognition for Interactive RC Car

This research investigates the application of electromyogram (EMG) signals for real-time human-computer interaction through the control of an RC vehicle using hand gestures. By employing a dual-channel EMG sensor, EMG signals associated with four distinct hand gestures were collected from ten participants. These signals were processed to extract both time-domain and frequency-domain features. A classification model was developed using a dataset of signals produced by four different gestures performed with two hands. This model, trained using a Boosted Trees learning algorithm, achieved an accuracy of 84.1% in gesture recognition. The findings highlight the viability of EMG-based interfaces for interactive device control and potential advancements in assistive technology.