MAP-Elites Algorithm Enhances FPS Map Generation
A research paper on arXiv investigates the use of MAP-Elites, a quality diversity algorithm, for procedurally generating first-person shooter (FPS) game maps. The study introduces two novel map representations—Point-Line and Spatial-Layout—alongside established All-Black and Grid-Graph methods, aiming to improve map characterization. Metrics are defined to evaluate topological properties (layout-dependent) and emergent properties (gameplay-dependent). The researchers apply MAP-Elites with Sliding Boundaries (MESB) to evolve map populations, finding that the new representations yield higher diversity and quality compared to previous approaches. The paper is categorized under Computer Science > Artificial Intelligence.
Key facts
- MAP-Elites algorithm used for FPS map generation
- Two new map representations: Point-Line and Spatial-Layout
- Existing representations: All-Black and Grid-Graph
- Metrics for topological and emergent properties defined
- MAP-Elites with Sliding Boundaries (MESB) applied
- New representations produce higher diversity and quality
- Paper published on arXiv under AI category
- Study focuses on quality diversity in procedural content generation
Entities
Institutions
- arXiv