3D Contextual Transformer & Double Inception Network for Human Action Recognition | IEEE Conference Publication | IEEE Xplore