Interactive evolution and exploration within latent level-design space of generative adversarial networks

Generative Adversarial Networks (GANs) are an emerging form of indirect encoding. The GAN is trained to induce a latent space on training data, and a real-valued evolutionary algorithm can search that latent space. Such Latent Variable Evolution (LVE) has recently been applied to game levels. However, it is hard for objective scores to capture level features that are appealing to players. Therefore, this paper introduces a tool for interactive LVE of tile-based levels for games. The tool also allows for direct exploration of the latent dimensions, and allows users to play discovered levels. The tool works for a variety of GAN models trained for both Super Mario Bros. and The Legend of Zelda, and is easily generalizable to other games. A user study shows that both the evolution and latent space exploration features are appreciated, with a slight preference for direct exploration, but combining these features allows users to discover even better levels. User feedback also indicates how this system could eventually grow into a commercial design tool, with the addition of a few enhancements.

[1]  Julian Togelius,et al.  The 2010 Mario AI Championship: Level Generation Track , 2011, IEEE Transactions on Computational Intelligence and AI in Games.

[2]  Kalyanmoy Deb,et al.  Simulated Binary Crossover for Continuous Search Space , 1995, Complex Syst..

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

[4]  Alexandru Iosup,et al.  Procedural content generation for games: A survey , 2013, TOMCCAP.

[5]  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).

[6]  Julian Togelius,et al.  Sentient Sketchbook: Computer-aided game level authoring , 2013, FDG.

[7]  Jean-Baptiste Mouret,et al.  Illuminating search spaces by mapping elites , 2015, ArXiv.

[8]  A. E. Eiben,et al.  Interactive Evolutionary Algorithms , 2015 .

[9]  Rafael Bidarra,et al.  Procedural Generation of Dungeons , 2014, IEEE Transactions on Computational Intelligence and AI in Games.

[10]  Sebastian Risi,et al.  Accelerating the Evolution of Cognitive Behaviors Through Human-Computer Collaboration , 2016, GECCO.

[11]  Vanessa Volz,et al.  Uncertainty Handling in Surrogate Assisted Optimisation of Games , 2019, KI - Künstliche Intelligenz.

[12]  Julian Togelius,et al.  Deep Interactive Evolution , 2018, EvoMUSART.

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

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

[15]  Kenneth O. Stanley,et al.  Picbreeder: A Case Study in Collaborative Evolutionary Exploration of Design Space , 2011, Evolutionary Computation.

[16]  Kenneth O. Stanley,et al.  A novel human-computer collaboration: combining novelty search with interactive evolution , 2014, GECCO.

[17]  Jeff Clune,et al.  Improving HybrID: How to best combine indirect and direct encoding in evolutionary algorithms , 2017, PloS one.

[18]  Kenneth O. Stanley,et al.  Abandoning Objectives: Evolution Through the Search for Novelty Alone , 2011, Evolutionary Computation.

[19]  Michael Mendler,et al.  Automatic Generation of Dungeons for Computer Games , 2002 .

[20]  Julian Togelius,et al.  DeepMasterPrint: Generating Fingerprints for Presentation Attacks , 2017, ArXiv.

[21]  Mads Haahr,et al.  Procedural Puzzle Generation: A Survey , 2020, IEEE Transactions on Games.

[22]  Pascal Kerschke,et al.  Single- and multi-objective game-benchmark for evolutionary algorithms , 2019, GECCO.

[23]  Sebastian Risi,et al.  Interactive evolution of levels for a competitive multiplayer FPS , 2015, 2015 IEEE Congress on Evolutionary Computation (CEC).

[24]  Navdeep Jaitly,et al.  Adversarial Autoencoders , 2015, ArXiv.

[25]  Edoardo Giacomello,et al.  Searching the Latent Space of a Generative Adversarial Network to Generate DOOM Levels , 2019, 2019 IEEE Conference on Games (CoG).

[26]  Léon Bottou,et al.  Wasserstein Generative Adversarial Networks , 2017, ICML.

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

[28]  Julian Togelius,et al.  Artificial Intelligence and Games , 2018, Springer International Publishing.

[29]  Julian Togelius,et al.  Autoencoder and Evolutionary Algorithm for Level Generation in Lode Runner , 2019, 2019 IEEE Conference on Games (CoG).

[30]  Julian Togelius,et al.  Procedural Personas as Critics for Dungeon Generation , 2015, EvoApplications.

[31]  Kenneth O. Stanley,et al.  Evolving content in the Galactic Arms Race video game , 2009, 2009 IEEE Symposium on Computational Intelligence and Games.

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

[33]  Kenneth O. Stanley,et al.  Abstract , 1998, Clinical Neurology and Neurosurgery.

[34]  Joel Lehman,et al.  Petalz: Search-Based Procedural Content Generation for the Casual Gamer , 2016, IEEE Transactions on Computational Intelligence and AI in Games.

[35]  Nathan Brewer Computerized Dungeons and Randomly Generated Worlds: From Rogue to Minecraft , 2017, Proc. IEEE.

[36]  Aaron C. Courville,et al.  Improved Training of Wasserstein GANs , 2017, NIPS.

[37]  Yoshua Bengio,et al.  Plug & Play Generative Networks: Conditional Iterative Generation of Images in Latent Space , 2016, 2017 IEEE Conference on Computer Vision and Pattern Recognition (CVPR).

[38]  Julian Togelius,et al.  Evolving levels for Super Mario Bros using grammatical evolution , 2012, 2012 IEEE Conference on Computational Intelligence and Games (CIG).

[39]  Hideyuki Takagi,et al.  Interactive evolutionary computation: fusion of the capabilities of EC optimization and human evaluation , 2001, Proc. IEEE.

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

[41]  Julian Togelius,et al.  Procedural Content Generation in Games , 2016, Computational Synthesis and Creative Systems.

[42]  Morteza Behrooz,et al.  The Learning of Zelda: Data-Driven Learning of Level Topology , 2015 .

[43]  Santiago Ontañón,et al.  The VGLC: The Video Game Level Corpus , 2016, ArXiv.

[44]  Simon Colton,et al.  General Analytical Techniques For Parameter-Based Procedural Content Generators , 2019, 2019 IEEE Conference on Games (CoG).

[45]  Joris Dormans,et al.  Adventures in level design: generating missions and spaces for action adventure games , 2010, PCGames@FDG.

[46]  Kristy Elizabeth Boyer,et al.  Generating Educational Game Levels with Multistep Deep Convolutional Generative Adversarial Networks , 2019, 2019 IEEE Conference on Games (CoG).

[47]  Sebastian Risi,et al.  CG-GAN: An Interactive Evolutionary GAN-based Approach for Facial Composite Generation) , 2019, AAAI.

[48]  D. E. Goldberg,et al.  Simple Genetic Algorithms and the Minimal, Deceptive Problem , 1987 .

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