Training code for "Coarse-to-Fine Volumetric Prediction for Single-Image 3D Human Pose", CVPR 2017 - View it on GitHub