Deep Convolutional Poses for Human Interaction Recognition in Monocular Videos | Read Paper on Bytez