Inertial Sensing and Graph Convolutional Transformer for Wearable Real-Time Motion Recognition

Key Points

• The aim is to improve real-time motion recognition through advanced inertial sensing and graph convolutional networks.
• Incorporated inertial sensors in wearable devices for data collection.
• Utilized graph convolutional transformers for processing and analysis of motion data.
• Evaluated the accuracy and efficiency of the proposed methods in real-time settings.
• Achieved over 90% accuracy in recognizing various motion patterns across participants.
• Demonstrated significantly reduced processing time compared to traditional methods.
• Highlighted practical applications in sports and health monitoring.

Abstract