Multi-channels CNN temporal features for depth-based action recognition.

Jacek Trelinski, Bogdan Kwolek

Browse the full ICMV paper archive.