读书笔记21:MiCT: Mixed 3D/2D Convolutional Tube for Human Action Recognition(CVPR2018)

NoSuchKey