LASR: Learning Articulated Shape Reconstruction from a Monocular Video