Welcome to our chapter on human pose estimation. In this chapter, we will be building a neural network that will predict 3D human poses using 2D images. We will do this with the help of transfer learning by using the VGG16 model architecture and modifying it accordingly for our current problem. By the end of this chapter, you will have a deep learning (DL) model that does a really good job of predicting human poses.
Visual effects (VFX) in movies are expensive. They involve using a lot of expensive sensors that will be placed on the body of the actor when shooting. The information from these sensors will then be used to build visual effects, all of which ends up being super expensive. We have been asked (in this hypothetical use case) by a major movie studio whether we can help their graphics department build cheaper and better visual...
