Fast realistic multi-action recognition using mined dense spatio-temporal features.

Andrew Gilbert, John Illingworth, Richard Bowden

Browse the full ICCV paper archive.