Creating two-dimensional tile maps by hand is a long and tedious process. Tile maps contain an inherent structure and repetition that we should be able to exploit. We provide a way of generating these maps automatically using machine learning techniques. With these set of tools level designers will be able to create worlds that are much larger in scope.
keywords: Procedural Modeling, Markov Chains, C#, IC, Semester Project, Machine Learning, Masters, Tile Maps
master’s semester project student: Pedro Amorim
thesis supervisor: Immanuel Koh
thesis professor: Jeffrey Huang
Image credits: MxD