Using A. I. for the generation of visual references in digital audio games from the perspective of sound design

Abstract

The aim of this article is to examine the application of artificial intelligence in producing visual references during the development of a digital game. These references are a standard inclusion in Game Design Documents or Art Design Documents and are typically displayed as mood boards. In this article, we will discuss sound design and how artificial intelligence, mainly Bing utilizing Dall-E and Midjourney, can produce images that cater to the requirements of a particular game. One of the primary issues we will explore is the efficiency of using artificial intelligence in this creative process in terms of speed, accuracy, efficacy, creativity and how the resulting content compares with human-generated content. One objective is to examine how artificial intelligence can produce visual references that exemplify more detailed sound groups, such as ambiences or interactive sounds linked with specific in-game activities. These visual references will undergo testing within a digital game presently under development, thus achieving validation through real-world application. The game’s genre is an audio game with no standard graphics or interface, tailored to the needs of players with visual impairments. For that reason, generating visual references remains the sole approach to comprehending the actual game environment.