Discovering discriminative action parts from mid-level video representations.

Michalis Raptis, Iasonas Kokkinos, Stefano Soatto

Browse the full CVPR paper archive.