PROJECT TITLE :

Learning-Based Procedural Content Generation

ABSTRACT:

Procedural content generation (PCG) has recently become one of the most well liked topics in computational intelligence and AI game analysis. Whereas some substantial progress has been created in this space, there are still many challenges starting from content evaluation to personalized content generation. During this paper, we have a tendency to gift a completely unique PCG framework based on Machine Learning, named learning-based procedure content generation (LBPCG), to tackle a range of difficult problems. By exploring and exploiting info gained in game development and public player check, our framework will generate robust content adaptable to finish-user or target players on-line with minimal interruption to their gameplay experience. As the information-driven methodology is stressed in our framework, we tend to develop learning-based mostly enabling techniques to implement the various models needed in our framework. For a symptom of concept, we have developed a prototype primarily based on the classic open supply initial-person shooter game, Quake. Simulation results suggest that our framework is promising in generating quality content.