Skeleton-Based Action Recognition Based on Distance Vector and Multihigh View Adaptive Networks

Figure 8

Comparative experiment of the four models on the NTU-RGB + D datasets.