Clustering-based Tile Embedding (CTE): A General Representation for Level Design with Skewed Tile Distributions

TL;DR

This paper introduces Clustering-based Tile Embedding (CTE), a novel representation for game level design that effectively handles skewed tile distributions, improving upon previous tile embedding methods in generating levels for various games.

Contribution

The paper proposes CTE, a clustering-based tile embedding that discretizes continuous representations, enabling better level generation with skewed tile distributions in procedural game design.

Findings

Outperforms original tile embeddings in level generation quality.

Effective for both seen and unseen games with skewed tile distributions.

Introduces a two-step level generation leveraging CTE.

Abstract

There has been significant research interest in Procedural Level Generation via Machine Learning (PLGML), applying ML techniques to automated level generation. One recent trend is in the direction of learning representations for level design via embeddings, such as tile embeddings. Tile Embeddings are continuous vector representations of game levels unifying their visual, contextual and behavioural information. However, the original tile embedding struggled to generate levels with skewed tile distributions. For instance, Super Mario Bros. (SMB) wherein a majority of tiles represent the background. To remedy this, we present a modified tile embedding representation referred to as Clustering-based Tile Embedding (CTE). Further, we employ clustering to discretize the continuous CTE representation and present a novel two-step level generation to leverage both these representations. We evaluate the performance of our approach in generating levels for seen and unseen games with skewed tile distributions and outperform the original tile embeddings.