Aggregating the temporal coherent descriptors in videos using multiple learning kernel for action recognition | Publicación