SokobanPCG

This project focused on the generation of Sokoban game levels using a combination of Reinforcement Learning (RL) and Wave Function Collapse (WFC) techniques. The goal was to create dynamic and challenging game levels that provided an engaging player experience.

Key Responsibilities and Accomplishments:

• Utilized the Wave Function Collapse (WFC) algorithm to generate layouts for Sokoban levels, which comprised walls and ground elements. This ensured the creation of diverse and unique level designs.
• Employed Reinforcement Learning (RL) techniques to analyze the level layouts, optimizing the placement of player characters, boxes, and goals to generate coherent and enjoyable gameplay experiences.
• Leveraged C# in the Unity Game Engine to implement and fine-tune the procedural content generation process, enhancing the level creation and player engagement.

This project exemplified the power of combining advanced algorithms and game development tools to create procedurally generated content that challenged and entertained players. The synergy of RL, WFC, C#, and Unity Game Engine facilitated the dynamic and engaging generation of Sokoban game levels.