Machine Learning-Based Hand Gesture Recognition Using Wearable Sensor Data

Abstract

This paper presents a novel approach to hand gesture recognition based on machine learning techniques applied to data collected from wearable sensors. The proposed system aims to accurately identify intuitive hand gestures to facilitate the development of contactless user interfaces. To achieve robust and reliable recognition performance, various feature extraction methods were systematically employed, and an ensemble learning strategy was developed using Support Vector Machines (SVM), Random Forests (RF), Gradient Boosting (GB), and a Voting Classifier (VC). The gesture dataset utilized in this study consists of over 4,000 samples collected from multiple individuals, capturing diverse gesture classes and conditions. Rigorous experimental evaluation demonstrates that the proposed ensemble approach attains superior accuracy and robustness compared to traditional single-model methods. This research contributes valuable methodological insights and provides an accessible, scalable framework that enhances the practical applicability of sensor-based gesture recognition systems in diverse real-world applications.