Research on Discriminative Skeleton-Based Action Recognition in Spatiotemporal Fusion and Human-Robot Interaction

Figure 2

Local posture motion-based channel attention module (LMP-CAM).