We create a moving 3d image from pictures using a neural network. This new machine-learning algorithm uses mannequin videos to create pictures for the network to train on and create a depth map to rotate pictures in 3d.