Overview

These SuperTux variants inspired by Super Mario Bros. demonstrate several cutting-edge types of AI:

1) Experience-driven Procedural Content Generation

Play a short SuperTux game and let the game generate an adapted level that is personalized according to your chosen emotional state. The content generator adopts player experience models trained on hundreds of players to generate content that maximizes your experience given your playing style.

Features

2) Player Experience Modeling

Play a variant of the SuperTux and let the game assess your emotional state using neuro models. The models are trained on hundreds of players to approximate the relationship between game content and player behavior. These models are then used to predict your emotional state based on your playing style and the level content.

Features

3) Evolution

Play an infinite number of SuperTux levels generated using grammatical evolution. The structure of the levels is specified using a design grammar that is adopted to generate endless playable levels by the use of grammatical evolution.

Features

Launch Website


Publications

Noor Shaker, Georgios Yannakakis and Julian Togelius. Digging deeper into platform game level design: session size and sequential features, in Proceedings of EvoGames: Applications of Evolutionary Computation, Lecture Notes on Computer Science, 2012.

Noor Shaker, Georgios Yannakakis and Julian Togelius. Feature Analysis for  Modeling Game Content Quality, in Proceedings of the 2011 IEEE Conference on  Computational Intelligence and Games, 2011.

Noor Shaker, Georgios Yannakakis and Julian Togelius. Towards Automatic Personalized Content Generation for Platform Games, in Proceedings of Artificial  Intelligence and Interactive Digital Entertainment (AIIDE 10), 2010.

Noor Shaker, Georgios Yannakakis, Julian Togelius, Miguel Nicolau and Michael O’Neill.Evolving Personalized Content for Super Mario Bros Using Grammatical Evolution, in Proceedings of Artificial Intelligence and Interactive Digital Entertainment (AIIDE12), 2012.

Noor Shaker, Miguel Nicolau, Georgios Yannakakis and Julian Togelius and Michael O’Neill.Evolving Levels for Super Mario Bros Using Grammatical Evolution, in Proceedings of the 2012 IEEE Conference on Computational Intelligence and Games (CIG 2012), 2012.