Machine learning-assisted wearable sensing for high-sensitivity gesture recognition

The emergence of the Internet of Things (IoT) and the ubiquity of 5 G technologies have increased the demand for self-powered, flexible, and high-precision sensors, especially in the context of human-computer interaction. These sensors should be able to accurately capture gesture information and provide a favorable experience for users. Existing solutions often fall short in terms of flexibility, energy efficiency, and gesture recognition accuracy, highlighting the need for sensor innovation. In this paper, we address this need by presenting a novel flexible wearable sensor array based on friction electric technology. To address this challenge, we propose a novel sensor array that utilizes the principles of triboelectric and electrostatic effects to detect and capture various hand gestures. By integrating data from three sensors, our system achieves a high level of accuracy in gesture recognition. The array can recognize nine different gesture actions, making it highly versatile in various applications. The contribution of this research lies in the integration of triboelectric technology with advanced machine learning techniques, specifically using Linear Discriminant Analysis (LDA). This integration enables the sensor array to achieve superior accuracy of more than 95% in recognizing different gestures. Furthermore, in terms of gesture recognition, wearable technology, and human-computer interaction applications, this research brings a strong boost to the progress in these fields.