In recent years, deep generative models have shown great performance on several tasks in the 2D domain. In my PhD, I am working on generative models in the 3D domain. My aims are to generate full 3D objects and to reconstruct 3D objects given some input condition like an image of the object.