Abstract: We propose a simple, yet effective approach for spatiotemporal feature learning using deep 3-dimensional convolutional networks (3D ConvNets) trained on a large scale supervised video ...
Some results have been hidden because they may be inaccessible to you
Show inaccessible results