A Hierarchical Static-Dynamic Encoder-Decoder Structure for 3D Human Motion Prediction with Residual CNNs

Figure 8

Visualization of predictive performance on (a) G3D and (b) FNTU.

(a)

(b)