|
|
Coevolution of AI and level generation for Super Mario game
Flimmel, Július ; Černý, Vojtěch (advisor) ; Pilát, Martin (referee)
Procedural Content Generation is now used in many games to generate a wide variety of content. It often uses players controlled by Artificial Intelligence for its evaluation. PCG content can also be used when training AI players to achieve better generalization. In both of these fields, evolutionary algorithms are employed, but they are rarely used together. In this thesis, we use the coevolution of AI players and level generators for platformer game Super Mario. Coevolution's benefit is, that the AI players are evaluated by adapting level generators, and vice versa, level generators are evaluated by adapting AI players. This approach has two results. The first one is a creation of multiple level generators, each generating levels of gradually increased difficulty. Levels generated using a sequence of these generators also mirror the learning curve of the AI player. This can be useful also for human players playing the game for the first time. The second result is an AI player, which was evolved on gradually more difficult levels. Making it learn progressively may yield better results. Using the coevolution also doesn't require any training data set.
|
|
|
Coevolution of AI and level generation for Super Mario game
Flimmel, Július ; Černý, Vojtěch (advisor) ; Pilát, Martin (referee)
Procedural Content Generation is now used in many games to generate a wide variety of content. It often uses players controlled by Artificial Intelligence for its evaluation. PCG content can also be used when training AI players to achieve better generalization. In both of these fields, evolutionary algorithms are employed, but they are rarely used together. In this thesis, we use the coevolution of AI players and level generators for platformer game Super Mario. Coevolution's benefit is, that the AI players are evaluated by adapting level generators, and vice versa, level generators are evaluated by adapting AI players. This approach has two results. The first one is a creation of multiple level generators, each generating levels of gradually increased difficulty. Levels generated using a sequence of these generators also mirror the learning curve of the AI player. This can be useful also for human players playing the game for the first time. The second result is an AI player, which was evolved on gradually more difficult levels. Making it learn progressively may yield better results. Using the coevolution also doesn't require any training data set.
|
|
|
Procedural Level Generator with Unity Integration
Nepožitek, Ondřej ; Gemrot, Jakub (advisor) ; Černý, Vojtěch (referee)
Procedural content generation is a method that is sometimes used in video games to increase their replayability. In our previous work (Nepožitek, 2018), we implemented an algorithm for procedural generation of 2D dungeons, with the main focus on giving game designers complete control over the structure of generated levels. The algorithm takes a set of user-defined building blocks as input and produces levels that all follow the structure of a specified level connectivity graph. In the first part of the thesis, we address some shortcomings of our previous work. We improve the algorithm with several new features such as better support for corridors between rooms or the possibility to generate platformer levels. We also propose several performance improvements and analyze the speed of the algorithm on various inputs. In the second part of the thesis, we present an integration of our algorithm into the Unity game engine. In the final part of the thesis, we demonstrate that our generator is able to produce levels that are similar to what we can see in two popular games - Enter the Gungeon and Dead Cells. The resulting algorithm is much faster than the previous version, contains new features and is ready to be used in the Unity game engine.
|
|
|
Procedural Content Generation for Video Games using Open Data
Tuncel, Merve ; Gemrot, Jakub (advisor) ; Kratochvíl, Miroslav (referee)
Games get boring when they start repeating themselves and do not offer players new content. Procedural content generation (PCG) is increasingly used to generate this content. PCG-based game design decreases the need to have a human designer or a writer to generate the content. Algorithmic creation of game content can augment the creativity of human designers and this makes it possible small so-called indie teams to create the content for their game without the big resources. In this work, the field of PCG is introduced. Application of PCG is shown through a mobile game implementation. The implementation details of the mobile game Rush Hour will be presented that makes use of Foursquare, Twitter and Mapbox APIs, which eases the content creation using open data as the input of PCG.
|
|
|
Methods for Procedural Generation of Skill Trees for Computer Games
Jaroschy, Petr ; Gemrot, Jakub (advisor) ; Pilát, Martin (referee)
Game developers often face the issue of having well balanced game in terms of difficulty, especially in role-playing games. Skill tree is a game element which contributes to solve this issue by giving the player more power in-game. Many game elements can be procedurally generated to save time and money to developers, but can skill trees be procedurally generated too? And how do we validate, that generated skill trees are well suited for the game? That is the goal of this work. We have made a simple turn-based game. Then we made several variations of generators of skill trees for the game. We took the best trees and validated their performance using artificial players based on data collected during their gameplay. Then we compared the trees and concluded that skill trees can be generated by our suggested method and their variations.
|
|
|
Procedural Generation of Endless Runner Type of Video Games
Černý, Vojtěch ; Gemrot, Jakub (advisor) ; Pilát, Martin (referee)
Procedural content generation (PCG) is increasingly used to generate many aspects in a variety of games. AI players, both hand scripted or also generated (by AI methods), are used to evaluate this content. Comparatively little effort is invested in using PCG to generate the whole game, including its rules. In this thesis, we use evolutionary algorithms to generate the game rules, its content and the evaluating AI player on a narrow, but flourishing, genre of endless runners - games where the player is constantly running. For this purpose, we have implemented a framework for creating endless runner games. Our approach could provide more efficiency for game designers, explore completely new game concepts in endless runners, platformer games, and be further generalized to other game genres.
|
|
|
Procedural 2D Map Generation for Computer Games
Nepožitek, Ondřej ; Gemrot, Jakub (advisor) ; Holan, Tomáš (referee)
In some video games, levels are procedurally generated to increase game's replayability. However, such levels may often feel too random, unbalanced and lacking an overall structure. Ma et al. (2014) proposed an algorithm to solve this problem. Their method takes a set of user-defined building blocks as an input and produces layouts that all follow the structure of a specified level connectivity graph. The algorithm is based on two main concepts. The first one is that the input graph is decomposed into smaller chains and these are laid out one at a time. The second one is that configuration spaces are used to define valid relative positions of building blocks. In this thesis, we present an implementation of this method in a context of 2D tile-based maps. We enhance the algorithm with several new features, one of them being a mode to quickly add short corridors between neighbouring rooms. We also propose speed improvements, including a smarter decomposition of the input graph and tweaks of the stochastic method that is used to lay out individual chains. The resulting algorithm is able to quickly produce diverse layouts, which is demonstrated on a variety of input graphs and building blocks sets. Benchmarks of our algorithm show that it can achieve up to two orders of magnitude speedup compared to the original...
|
|
|
Procedural generation of cities in 3D
Krabec, Miroslav ; Gemrot, Jakub (advisor) ; Beneš, Jan (referee)
Title: Procedural generation of cities in 3D Author: Miroslav Krabec Department: Department of Software and Computer Science Education Supervisor: Mgr. Jakub Gemrot Abstract: During development of computer games great amount of time is spent on creating game environment. For that reason there is an effort to generate this environment procedurally. One of the interesting areas is generation of cities. Algorithm (Weber et al., 2009) is fully geometrical and offers considerable freedom in parametrization of the city, however there has not been an open implementation of this algorithm. Our work offers such an implementation and includes the possibility of 3D visualization in Unity 3D, a tool designed for development of computer games. Here we emphasize the ease of using custom 3D models of buildings, roads and crossroads. Resulting software can help game developers to generate cities for their games. However it can generate only relatively small cities (several thousands of buildings), but in context of computer games this is usually sufficient. Resulting city is very parameter sensitive and it is not trivial to achieve desired outcome. Further research would be needed for evaluation of generated cities. Keywords: procedural generation, cities, 3D
|
guest ::
