Capturing Local and Global Patterns in Procedural Content Generation via Machine Learning

Recent procedural content generation via machine learning (PCGML) methods allow learning from existing content to produce similar content automatically. While these approaches are able to generate content for different games (e.g. Super Mario Bros., DOOM, Zelda, and Kid Icarus), it is an open question how well these approaches can capture large-scale visual patterns such as symmetry. In this paper, we propose match-three games as a domain to test PCGML algorithms regarding their ability to generate suitable patterns. We demonstrate that popular algorithms such as Generative Adversarial Networks struggle in this domain and propose adaptations to improve their performance. In particular, we augment the neighbourhood of a Markov Random Fields approach to take into account not only local but also symmetric positional information. We conduct several empirical tests, including a user study that show the improvements achieved by the proposed modifications and obtain promising results.

[1]  Julian Togelius,et al.  Bootstrapping Conditional GANs for Video Game Level Generation , 2019, 2020 IEEE Conference on Games (CoG).

[2]  Anurag Sarkar,et al.  Controllable Level Blending between Games using Variational Autoencoders , 2020, ArXiv.

[3]  Julian Togelius,et al.  Procedural Content Generation via Machine Learning (PCGML) , 2017, IEEE Transactions on Games.

[4]  Simon M. Lucas,et al.  Evolving mario levels in the latent space of a deep convolutional generative adversarial network , 2018, GECCO.

[5]  E. James Whitehead,et al.  Design patterns in FPS levels , 2010, FDG.

[6]  Philipp Eisen,et al.  Simulating Human Game Play for Level Difficulty Estimation with Convolutional Neural Networks , 2017 .

[7]  Matthew Guzdial,et al.  Game Level Generation from Gameplay Videos , 2021, AIIDE.

[8]  Matthew Guzdial,et al.  Explainable PCGML via Game Design Patterns , 2018, AIIDE Workshops.

[9]  Simon M. Lucas,et al.  Tile pattern KL-divergence for analysing and evolving game levels , 2019, GECCO.

[10]  Jacob Schrum,et al.  CPPN2GAN: combining compositional pattern producing networks and GANs for large-scale pattern generation , 2020, GECCO.

[11]  Kenneth O. Stanley,et al.  Compositional Pattern Producing Networks : A Novel Abstraction of Development , 2007 .

[12]  Amy K. Hoover.  Composing Video Game Levels with Music Metaphors through Functional Scaffolding , 2015 .

[13]  Julian Togelius,et al.  Linear levels through n-grams , 2014, MindTrek.

[14]  Peter Clifford,et al.  Markov Random Fields in Statistics , 2012 .

[15]  Zhen Zhang,et al.  An Adaptive Markov Random Field for Structured Compressive Sensing , 2018, IEEE Transactions on Image Processing.

[16]  Noor Shaker,et al.  Alone We Can Do So Little, Together We Can Do So Much: A Combinatorial Approach for Generating Game Content , 2014, AIIDE.

[17]  Julian Togelius,et al.  Predicting Resource Locations in Game Maps Using Deep Convolutional Neural Networks , 2021, Proceedings of the AAAI Conference on Artificial Intelligence and Interactive Digital Entertainment.

[18]  Jacob Schrum,et al.  Generative Adversarial Network Rooms in Generative Graph Grammar Dungeons for The Legend of Zelda , 2020, 2020 IEEE Congress on Evolutionary Computation (CEC).

[19]  Santiago Ontañón,et al.  Controllable Procedural Content Generation via Constrained Multi-Dimensional Markov Chain Sampling , 2016, IJCAI.

[20]  Michael Mateas,et al.  Sampling Hyrule: Sampling Probabilistic Machine Learning for Level Generation , 2015 .

[21]  Michael Mateas,et al.  MCMCTS PCG 4 SMB: Monte Carlo Tree Search to Guide Platformer Level Generation , 2021, Proceedings of the AAAI Conference on Artificial Intelligence and Interactive Digital Entertainment.

[22]  Han Zhang,et al.  Self-Attention Generative Adversarial Networks , 2018, ICML.

[23]  Santiago Ontañón,et al.  A Hierarchical MdMC Approach to 2D Video Game Map Generation , 2015, AIIDE.

[24]  Yoshua Bengio,et al.  Generative Adversarial Nets , 2014, NIPS.

[25]  M. Hassner,et al.  The use of Markov Random Fields as models of texture , 1980 .

[26]  Julian Togelius,et al.  Patterns and procedural content generation: revisiting Mario in world 1 level 1 , 2012, DPG '12.

[27]  Simon Osindero,et al.  Conditional Generative Adversarial Nets , 2014, ArXiv.

[28]  Daniel P. Huttenlocher,et al.  Sparse Long-Range Random Field and Its Application to Image Denoising , 2008, ECCV.

[29]  Jason Yosinski,et al.  An Intriguing Failing of Convolutional Neural Networks and the CoordConv Solution , 2018, NeurIPS.

[30]  Risto Miikkulainen,et al.  A Taxonomy for Artificial Embryogeny , 2003, Artificial Life.

[31]  Santiago Ontañón,et al.  Learning to Generate Video Game Maps Using Markov Models , 2017, IEEE Transactions on Computational Intelligence and AI in Games.

[32]  Marshall F. Tappen,et al.  Learning non-local range Markov Random field for image restoration , 2011, CVPR 2011.

[33]  Michael Mateas,et al.  Super Mario as a String: Platformer Level Generation Via LSTMs , 2016, DiGRA/FDG.

[34]  Adam M. Smith,et al.  WaveFunctionCollapse is constraint solving in the wild , 2017, FDG.

[35]  Soumith Chintala,et al.  Unsupervised Representation Learning with Deep Convolutional Generative Adversarial Networks , 2015, ICLR.

[36]  Staffan Björk,et al.  Patterns In Game Design , 2004 .

[37]  P. Bickel,et al.  Texture synthesis and nonparametric resampling of random fields , 2006, math/0611258.