Mixed-Initiative Level Design with RL Brush

This paper introduces RL Brush, a level-editing tool for tile-based games designed for mixed-initiative co-creation. The tool uses reinforcement-learning-based models to augment manual human level-design through the addition of AI-generated suggestions. Here, we apply RL Brush to designing levels for the classic puzzle game Sokoban. We put the tool online and tested it with 39 different sessions. The results show that users using the AI suggestions stay around longer and their created levels on average are more playable and more complex than without.

Co-generation of game levels and game-playing agents

Open-endedness, primarily studied in the context of artificial life, is the ability of systems to generate potentially unbounded ontologies of increasing novelty and complexity. Engineering generative systems displaying at least some degree of this ability is a goal with clear applications to procedural content generation in games. The Paired Open-Ended Trailblazer (POET) algorithm, heretofore explored only in a biped walking domain, is a coevolutionary system that simultaneously generates environments and agents that can solve them. This paper introduces a POET-Inspired Neuroevolutionary System for KreativitY (PINSKY) in games, which co-generates levels for multiple video games and agents that play them. This system leverages the General Video Game Artificial Intelligence (GVGAI) framework to enable co-generation of levels and agents for the 2D Atari-style games Zelda and Solar Fox. Results demonstrate the ability of PINSKY to generate curricula of game levels, opening up a promising new avenue for research at the intersection of procedural content generation and artificial life. At the same time, results in these challenging game domains highlight the limitations of the current algorithm and opportunities for improvement.

Illuminating Mario Scenes in the Latent Space of a Generative Adversarial Network

Recent developments in machine learning techniques have allowed automatic generation of video game levels that are stylistically similar to human-designed examples. While the output of machine learning models such as generative adversarial networks (GANs) is notoriously hard to control, the recently proposed latent variable evolution (LVE) technique searches the space of GAN parameters to generate outputs that optimize some objective performance metric, such as level playability. However, the question remains on how to automatically generate a diverse range of high-quality solutions based on a prespecified set of desired characteristics. We introduce a new method called latent space illumination (LSI), which uses state-of-the-art quality diversity algorithms designed to optimize in continuous spaces, i.e., MAP-Elites with a directional variation operator and Covariance Matrix Adaptation MAP-Elites, to effectively search the parameter space of theGAN along a set of multiple level mechanics. We show the performance of LSI algorithms in three experiments in SuperMario Bros., a benchmark domain for procedural content generation. Results suggest that LSI generates sets of Mario levels that are reliably mechanically diverse as well as playable.

Multi-Stage Transfer Learning with an Application to Selection Process

In multi-stage processes, decisions happen in an ordered sequence of stages. Many of them have the structure of dual funnel problem: as the sample size decreases from one stage to the other, the information increases. A related example is a selection process, where applicants apply for a position, prize, or grant. In each stage, more applicants are evaluated and filtered out, and from the remaining ones, more information is collected. In the last stage, decision-makers use all available information to make their final decision. To train a classifier for each stage becomes impracticable as they can underfit due to the low dimensionality in early stages or overfit due to the small sample size in the latter stages. In this work, we proposed a \textit{Multi-StaGe Transfer Learning} (MSGTL) approach that uses knowledge from simple classifiers trained in early stages to improve the performance of classifiers in the latter stages. By transferring weights from simpler neural networks trained in larger datasets, we able to fine-tune more complex neural networks in the latter stages without overfitting due to the small sample size. We show that it is possible to control the trade-off between conserving knowledge and fine-tuning using a simple probabilistic map. Experiments using real-world data demonstrate the efficacy of our approach as it outperforms other state-of-the-art methods for transfer learning and regularization.

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 questions 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 algorithm such as Generative Adversarial Networks struggle in this domain and propose adaptations to improve their performance. In particular we augment the neighborhood of a Markov Random Fields approach to not only take local but also symmetric positional information into account. We conduct several empirical tests including a user study that show the improvements achieved by the proposed modifications, and obtain promising results.

Multi-Objective level generator generation with Marahel

This paper introduces a new system to design constructive level generators by searching the space of constructive level generators defined by Marahel language. We use NSGA-II, a multi-objective optimization algorithm, to search for generators for three different problems (Binary, Zelda, and Sokoban). We restrict the representation to a subset of Marahel language to push the evolution to find more efficient generators. The results show that the generated generators were able to achieve a good performance on most of the fitness functions over these three problems but on Zelda and Sokoban they tend to depend on the initial state than modifying the map.

Generating and Adapting to Diverse Ad-Hoc Cooperation Agents in Hanabi

Hanabi is a cooperative game that brings the problem of modeling other players to the forefront. In this game, coordinated groups of players can leverage pre-established conventions to great effect, but playing in an ad-hoc setting requires agents to adapt to its partner's strategies with no previous coordination. Evaluating an agent in this setting requires a diverse population of potential partners, but so far, the behavioral diversity of agents has not been considered in a systematic way. This paper proposes Quality Diversity algorithms as a promising class of algorithms to generate diverse populations for this purpose, and generates a population of diverse Hanabi agents using MAP-Elites. We also postulate that agents can benefit from a diverse population during training and implement a simple "meta-strategy" for adapting to an agent's perceived behavioral niche. We show this meta-strategy can work better than generalist strategies even outside the population it was trained with if its partner's behavioral niche can be correctly inferred, but in practice a partner's behavior depends and interferes with the meta-agent's own behavior, suggesting an avenue for future research in characterizing another agent's behavior during gameplay.

Evaluating the Rainbow DQN Agent in Hanabi with Unseen Partners

Hanabi is a cooperative game that challenges exist-ing AI techniques due to its focus on modeling the mental states ofother players to interpret and predict their behavior. While thereare agents that can achieve near-perfect scores in the game byagreeing on some shared strategy, comparatively little progresshas been made in ad-hoc cooperation settings, where partnersand strategies are not known in advance. In this paper, we showthat agents trained through self-play using the popular RainbowDQN architecture fail to cooperate well with simple rule-basedagents that were not seen during training and, conversely, whenthese agents are trained to play with any individual rule-basedagent, or even a mix of these agents, they fail to achieve goodself-play scores.

Towards Designer Modeling through Design Style Clustering

We propose modeling designer style in mixed-initiative game content creation tools as archetypical design traces. These design traces are formulated as transitions between design styles; these design styles are in turn found through clustering all intermediate designs along the way to making a complete design. This method is implemented in the Evolutionary Dungeon Designer, a prototype mixed-initiative system for roguelike games. We present results both in the form of design styles for rooms, which can be analyzed to better understand the kind of rooms designed by users, and in the form of archetypical sequences between these rooms. We further discuss how the results here can be used to create style-sensitive suggestions. Such suggestions would allow the system to be one step ahead of the designer, offering suggestions for the next phase, assuming that the designer will follow one of the archetypical design traces.

Unified Multi-Domain Learning and Data Imputation using Adversarial Autoencoder

We present a novel framework that can combine multi-domain learning (MDL), data imputation (DI) and multi-task learning (MTL) to improve performance for classification and regression tasks in different domains. The core of our method is an adversarial autoencoder that can: (1) learn to produce domain-invariant embeddings to reduce the difference between domains; (2) learn the data distribution for each domain and correctly perform data imputation on missing data. For MDL, we use the Maximum Mean Discrepancy (MMD) measure to align the domain distributions. For DI, we use an adversarial approach where a generator fill in information for missing data and a discriminator tries to distinguish between real and imputed values. Finally, using the universal feature representation in the embeddings, we train a classifier using MTL that given input from any domain, can predict labels for all domains. We demonstrate the superior performance of our approach compared to other state-of-art methods in three distinct settings, DG-DI in image recognition with unstructured data, MTL-DI in grade estimation with structured data and MDMTL-DI in a selection process using mixed data.