Competitive Physical Action Recognition via Inertial Sensing and Graph Convolutional Transformer

Key Points

• This research aims to explore the use of inertial sensing and graph convolutional transformers for recognizing physical actions.
• Utilized inertial sensing data for action recognition
• Incorporated graph convolutional transformers in the analysis
• Evaluated the model's performance in competitive physical actions
• Successfully identified various physical actions with high accuracy
• Showed the effectiveness of graph convolutional transformers in processing inertial data