CA3D: Convolutional-Attentional 3D Nets for Efficient Video Activity Recognition on the Edge.

Gabriele Lagani, Fabrizio Falchi, Claudio Gennaro, Giuseppe Amato

Browse the full ECCV paper archive.