Using a Multilearner to Fuse Multimodal Features for Human Action Recognition

Figure 6

Sample images from the Cornell Activity Dataset 60.