Current action recognition systems require large amounts of training dat...
It was recently shown that the structure of convolutional neural network...
We propose novel Stacked Spatio-Temporal Graph Convolutional Networks
(S...
Many imaging tasks require global information about all pixels in an ima...
share