Hierarchically Learned View-Invariant Representations for Cross-View Action Recognition | IEEE Journals & Magazine | IEEE Xplore