Using a Multilearner to Fuse Multimodal Features for Human Action Recognition

Figure 4

The projection of the joint points of the human body.