作者: Christoph Sydora , Matthew Guzdial , Anmol Mahajan , Brad Burega , Eugene Chen
DOI:
关键词:
摘要: Procedural content generation via machine learning (PCGML) has recently gained research attention due to its ability generate new game with minimal user input. However, thus far those without expertise have been largely unable use PCGML fit their needs. This paper proposes the of images as input for a process levels. Intuitively, can submit an image, system returning closest valid Super Mario Bros. level. Our results indicate that at least domains like we recreate target level high fidelity.