Procedural Content Generation via Machine Learning (PCGML)

Summerville, Adam; Snodgrass, Sam; Guzdial, Matthew; Holmgård, Christoffer; Hoover, Amy K.; Isaksen, Aaron; Nealen, Andy; Togelius, Julian

doi:10.48550/arxiv.1702.00539

Cited by 17 publications

(33 citation statements)

References 0 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Procedural Content Generation via Machine Learning (PCGML) can be used for both level generation and rule generation, both of which we attempt in this work [24]. Initially PCGML research focused on level generation rather than rule generation or autonomous game design.…”

Section: Procedural Content Generation Via Machine Learningmentioning

confidence: 99%

“…traditional PCG, and can create high quality content [24]. However, there has been relatively little work on using PCGML for generating entire games.…”

mentioning

confidence: 99%

“…The task of autonomous game generation requires both autonomous level design, and autonomous rule design at minimum. Machine learning has been used for PCG but generally focuses on level generation [24]. The problem of generating levels through PCGML is often easier as the representation of the data is simpler, and there is a much larger dataset.…”

mentioning

confidence: 99%

See 2 more Smart Citations

Adversarial Random Forest Classifier for Automated Game Design

Maurer

Guzdial

2021

The 16th International Conference on the Foundations of Digital Games (FDG) 2021

Self Cite

View full text Add to dashboard Cite

Autonomous game design, generating games algorithmically, has been a longtime goal within the technical games research field. However, existing autonomous game design systems have relied in large part on human-authoring for game design knowledge, such as fitness functions in search-based methods. In this paper, we describe an experiment to attempt to learn a human-like fitness function for autonomous game design in an adversarial manner. While our experimental work did not meet our expectations, we present an analysis of our system and results that we hope will be informative to future autonomous game design research.

show abstract

Section: Procedural Content Generation Via Machine Learningmentioning

confidence: 99%

“…traditional PCG, and can create high quality content [24]. However, there has been relatively little work on using PCGML for generating entire games.…”

mentioning

confidence: 99%

See 1 more Smart Citation

Adversarial Random Forest Classifier for Automated Game Design

Maurer

Guzdial

2021

The 16th International Conference on the Foundations of Digital Games (FDG) 2021

Self Cite

View full text Add to dashboard Cite

show abstract

“…This paper explores how machine learning can be used for procedural content generation as a surrogate model, indirectly influencing the fitness function of a search-based PCG algorithm [33]. So far, machine learned models are primarily used to directly manipulate game content [32]. For instance, neural networks have mostly been used to learn level patterns which are then applied directly to the level.…”

Section: Related Workmentioning

confidence: 99%

“…While machine learning has a long history in procedural generation [32], the proposed framework uses its learned model indirectly (as a surrogate model to guide evolution) rather than directly. More importantly, it follows earlier research [17] in merging game rules (in the form of class parameters) and level properties as inputs, in order to learn how their interrelations affect gameplay outcomes.…”

Section: Related Workmentioning

confidence: 99%

Pairing character classes in a deathmatch shooter game via a deep-learning surrogate model

Karavolos

Liapis

Yannakakis

2018

Proceedings of the 13th International Conference on the Foundations of Digital Games

View full text Add to dashboard Cite

This paper introduces a surrogate model of gameplay that learns the mapping between different game facets, and applies it to a generative system which designs new content in one of these facets. Focusing on the shooter game genre, the paper explores how deep learning can help build a model which combines the game level structure and the game's character class parameters as input and the gameplay outcomes as output. The model is trained on a large corpus of game data from simulations with artificial agents in random sets of levels and class parameters. The model is then used to generate classes for specific levels and for a desired game outcome, such as balanced matches of short duration. Findings in this paper show that the system can be expressive and can generate classes for both computer generated and human authored levels. CCS CONCEPTS • Applied computing → Computer games; • Theory of computation → Evolutionary algorithms; • Computing methodologies → Neural networks;

show abstract

Evolving mario levels in the latent space of a deep convolutional generative adversarial network

Volz

Schrum²,

Liu

et al. 2018

Proceedings of the Genetic and Evolutionary Computation Conference

207

247

View full text Add to dashboard Cite

Generative Adversarial Networks (GANs) are a machine learning approach capable of generating novel example outputs across a space of provided training examples. Procedural Content Generation (PCG) of levels for video games could benefit from such models, especially for games where there is a pre-existing corpus of levels to emulate. This paper trains a GAN to generate levels for Super Mario Bros using a level from the Video Game Level Corpus. The approach successfully generates a variety of levels similar to one in the original corpus, but is further improved by application of the Covariance Matrix Adaptation Evolution Strategy (CMA-ES). Specifically, various fitness functions are used to discover levels within the latent space of the GAN that maximize desired properties. Simple static properties are optimized, such as a given distribution of tile types. Additionally, the champion A* agent from the 2009 Mario AI competition is used to assess whether a level is playable, and how many jumping actions are required to beat it. These fitness functions allow for the discovery of levels that exist within the space of examples designed by experts, and also guide the search towards levels that fulfill one or more specified objectives.

show abstract

Procedural Content Generation via Machine Learning (PCGML)

Cited by 17 publications

References 0 publications

Adversarial Random Forest Classifier for Automated Game Design

Adversarial Random Forest Classifier for Automated Game Design

Pairing character classes in a deathmatch shooter game via a deep-learning surrogate model

Evolving mario levels in the latent space of a deep convolutional generative adversarial network

Contact Info

Product

Resources

About