Matt’s research focuses on applying machine learning for automating complex tasks like 3D scene design, playing video games, and image+video editing. Recently he has worked on designing representations for 3D geometry that can be generated by deep networks, and applying generative and adversarial image networks to create new tools for artists. His PhD was under Pat Hanrahan at Stanford University and his undergraduate was at Caltech under Mathieu Desbrun and Peter Schroeder.
His personal website with a complete publication list is here.
His old Stanford website is here.