A Study on Stage Generation for 2D Action Game based on DC-GAN

Abstract

In recent years, Researchers has studied generative adversarial networks (GAN) in an image generation. We have verified a method to automatically generate a dot picture game stage using GAN. Specifically, we eliminated the blur of the boundary line in the image generated using GAN by coloring the objects in the stage in advance. In this research, we verify to automatically generate stages of 2D action games with different systems in the same way.