Learning a Mid-Level Representation for Multiview Action Recognition

Figure 3

Confusion matrix of our action recognition method on the MuHAVi-MAS dataset.