Theses and Dissertations Collection

Asset generation in video games and simulations is an expensive process that takes considerable resources to do well. Procedural Content Generation (PCG) is a technique that can be used to generate this content for video games and simulations. Although, there are certain flaws related to content quality and diversity that need to be addressed in order to utilize this technique generally. The Innovation Engine is a relatively new theoretical framework that is incrementally attempting to encapsulate human creativity by computational means. Optimizing techniques such as Genetic Algorithms (GAs) are inadequate when trying to accomplish goals related to creativity. A new software framework presented in this work, the Diversity Engine, acts as a practical application of the ideas found in Innovation Engines and shows procedurally generated flowers can be created, with mixed results, using a new Quality Diversity (QD) algorithm Centroidal Voronoi Tessalation MAP-Elites (CVT-MAP-Elites).