Action snippets: How many frames does human action recognition require?

Konrad Schindler, Luc Van Gool

Browse the full CVPR paper archive.