Another research path we are exploring is the generation of a 3D model of a coffee cup based on an AI generated cup. We fed more than 1000 pictures of coffee cups to a StyleGAN deep learning network. The result is an algorithm that can generate different coffee cups based on a set of parameters. The next step is automatically transform the 2D image into a 3D model. The video shows different features the network uses to generate a cup. For more information about StyleGAN and deep learning, check this post. It’s a nice explanation of the process and very understandable for those who are less familiar with artificial intelligence.