研究目的
To reduce the computational complexity and improve the recognition accuracy for human activity recognition based on Kinect sensor.
研究成果
The extraction of features is discriminative, and the method further refines the core findings, significance of the research, and suggests future studies.
研究不足
The technical and application constraints of the experiments, as well as potential areas for optimization, are not explicitly mentioned.