Sparseness Meets Deepness: 3D Human Pose Estimation from Monocular Video | Read Paper on Bytez