3D Human Shape and Pose from a Single Low-Resolution Image with Self-Supervised Learning - 42Papers