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

DOI Open Access

Description

Recently, research on generative adversarial networks (GANs) in deep learning has advanced rapidly. For instance, in the field of image recognition, using a GAN, the number of training data was increased. GAN have also been used to create new similar images using training images, which can help designers such as car designer, character designer, game designer and more. In this study, we consider a method to automatically generate a game stage based on a dot-picture using GAN. However, the images generated by GAN using the original game stages are blurred. To solve this problem, we perform preprocessing to the original images. Specifically, we painted out colors for each type of object, and deleted backgrounds and unnecessary objects. In our experiments, we automatically generated game stages using DC-GAN for two 2D action games with different scroll directions (vertical or horizontal). As a result, the proposed method can generate game stages clearly in some game systems.

Journal

Details 詳細情報について

Report a problem

Back to top