Abstract:This survey explores Procedural Content Generation via Machine Learning (PCGML), defined as the generation of game content using machine learning models trained on existing content. As the importance of PCG for game development increases, researchers explore new avenues for generating high-quality content with or without human involvement; this paper addresses the relatively new paradigm of using machine learning (in contrast with search-based, solver-based, and constructive methods). We focus on what is most often considered functional game content such as platformer levels, game maps, interactive fiction stories, and cards in collectible card games, as opposed to cosmetic content such as sprites and sound effects. In addition to using PCG for autonomous generation, co-creativity, mixed-initiative design, and compression, PCGML is suited for repair, critique, and content analysis because of its focus on modeling existing content. We discuss various data sources and representations that affect the resulting generated content. Multiple PCGML methods are covered, including neural networks, long short-term memory (LSTM) networks, autoencoders, and deep convolutional networks; Markov models, $n$-grams, and multi-dimensional Markov chains; clustering; and matrix factorization. Finally, we discuss open problems in the application of PCGML, including learning from small datasets, lack of training data, multi-layered learning, style-transfer, parameter tuning, and PCG as a game mechanic.

Toward Co-creative Dungeon Generation via Transfer Learning

Procedural Content Generation via Machine Learning (PCGML)

Player-Oriented Procedural Generation: Producing Desired Game Content by Natural Language

Procedural Content Generation via Generative Artificial Intelligence

PCGRL: Procedural Content Generation via Reinforcement Learning

Learning to Generate Levels From Nothing

PCGPT: Procedural Content Generation via Transformers

Joint Level Generation and Translation Using Gameplay Videos

Procedural Content Generation via Knowledge Transformation (PCG-KT)

A Controllable Co-Creative Agent for Game System Design

Co-generation of game levels and game-playing agents

An Interactive Evolution Strategy Based Deep Convolutional Generative Adversarial Network for 2D Video Game Level Procedural Content Generation

Learning Constructive Primitives for Online Level Generation and Real-time Content Adaptation in Super Mario Bros

PCGRL+: Scaling, Control and Generalization in Reinforcement Learning Level Generators

Game Level Blending using a Learned Level Representation

Adversarial Reinforcement Learning for Procedural Content Generation

Toward Game Level Generation from Gameplay Videos

Two-step Constructive Approaches for Dungeon Generation

Integrating Automated Play in Level Co-Creation

Mastering the Dungeon: Grounded Language Learning by Mechanical Turker Descent

Deep Learning for Procedural Content Generation