An Effective Video Transformer With Synchronized Spatiotemporal and Spatial Self-Attention for Action Recognition | IEEE Journals & Magazine | IEEE Xplore