Official implementation of CVPR2022 paper "Capturing Humans in Motion: Temporal-Attentive 3D Human Pose and Shape Estimation from Monocular Video" - View it on GitHub