Skip to search boxSkip to navigationSkip to main content

Enhancing game development process using AI: a comparative analysis of image generative AI

  • University of Staffordshire
Research Output: Contribution to conference Paper Peer-review

Abstract

The gaming industry continually faces the challenge of high costs and extensive time requirements for producing 2D assets. Gaming studies hire and pay high costs to artists to create images for the game which is also time consuming. With the pervasive integration of artificial intelligence across various sectors, the application of AI in game development presents a promising solution to these challenges. This study focuses on the comparison and outcomes of two advanced generative AI tools, DALL-E and Midjourney, to reduce the time and cost associated with the creation of dynamic 2D game assets. The comparison process for both the AI tools involves creation of a Unity based game that integrates DALL-E and Midjourney to generate dynamic game assets which were used in the gameplay to get the user feedback inside the game. The findings reveal that both DALL-E and Midjourney significantly reduce the time and financial barriers traditionally associated with game asset creation, democratizing game development and allowing smaller studios to produce competitive content. DALL-E was found to be highly effective in rapidly transforming textual descriptions into high-quality visual assets, making it superior in terms of speed and cost-efficiency compared to Midjourney. Player feedback indicated a higher level of satisfaction and engagement with DALL-E-generated assets, particularly due to its superior performance in handling prompts and generating contextually relevant images.

Publication Information

Output type

Research Output: Contribution to conference Paper Peer-review

Original language

English

Pages from-to (Number of pages)

Pages 946-951 (6 pages)

Publication milestones

  • Published - 24/12/2024

Publication status

Published - 24/12/2024

External Publication IDs

  • Scopus: 85216126714