Recognising action as clouds of space-time interest points.

Matteo Bregonzio, Shaogang Gong, Tao Xiang

Browse the full CVPR paper archive.