In the context of video games, Procedural Content Generation (PCG) refers to the automatic creation of contents, performed using algorithms and/or heuristics that are ‐ generally ‐ designed specifically for the game under development. PCG can be exploited to produce different contents, depending on the video game genre and on the peculiarities of the specific game, with the positive effects of reducing development time and increasing randomness of content and/or gameplay. In this invited speech, we will present the researches on PCG conducted in the PONG (Playlab fOr inNovation in Games) laboratory of the Department of Computer Science at University of Milan in the last years. We have exploited PCG potentialities with different approaches: -) ”classical” PCG for content creation: we have focused our attention on the proposal of procedural-based tools aimed at an optimization of the design process of levels and/or large virtual worlds. In particular, we have shown how PCG can be efficiently applied for the interactive generation of levels for Platformer games, exploiting also the relation between levels structure and musical rhythms. From the experimental evaluation, the resulting levels were considered fun to play, and the design tool allowed a satisfying control on the final desired level of complexity. We have also applied PCG for the generation of imaginary worlds for fantasy video games. We have based the procedural generation on an approximated simulation of the physical phenomena at the basis of the evolution of our planet, starting from the placement of physical elements like mountains, oceans, etc., and then adding inhabited areas to the map. In this stage, the tool has been designed in order to consider specific places created using tools to interactively help the production of narrative structures for stories to include in games. -)PCG based on evolutionary algorithms: we have focused our research on novel methods to exploit the characteristics of Genetic Algorithms (GAs) in the generation of game contents. In particular, we have presented an algorithm to explicitly address the need to introduce more variety and unpredictability in the monsters inside MMOs, in order to avoid that the players consider the game repetitive and less enjoyable after a long period of time spent playing. The main idea was to characterize each monster specie present in the game through its genome, and to generate new species by recombining their chromosomes, which represent a set of physical characteristics and skills. Each monster was represented by a chromosome composed by 53 genes, and the recombination process evaluates also the probability for the new monster to actually survive in the habitat in which it is born. More recently, we have presented a method for the automatic generation of realistic layered materials, based on the application of a GA. We have shown how a GA can be applied efficiently to evolve the parameters of a Bidirectional Scattering Distribution Function (BSDF), in order to generate different versions of a target material presenting a moderate amount of perceptual differences.
Procedural Content Generation Techniques applied to Game Design and Development (Keynote Speech) / D. Gadia. ((Intervento presentato al 21. convegno GAME‐ON : European Conference on Simulation and AI in Computer Games tenutosi a Aveiro nel 2020.
Procedural Content Generation Techniques applied to Game Design and Development (Keynote Speech)
D. GadiaPrimo
2020
Abstract
In the context of video games, Procedural Content Generation (PCG) refers to the automatic creation of contents, performed using algorithms and/or heuristics that are ‐ generally ‐ designed specifically for the game under development. PCG can be exploited to produce different contents, depending on the video game genre and on the peculiarities of the specific game, with the positive effects of reducing development time and increasing randomness of content and/or gameplay. In this invited speech, we will present the researches on PCG conducted in the PONG (Playlab fOr inNovation in Games) laboratory of the Department of Computer Science at University of Milan in the last years. We have exploited PCG potentialities with different approaches: -) ”classical” PCG for content creation: we have focused our attention on the proposal of procedural-based tools aimed at an optimization of the design process of levels and/or large virtual worlds. In particular, we have shown how PCG can be efficiently applied for the interactive generation of levels for Platformer games, exploiting also the relation between levels structure and musical rhythms. From the experimental evaluation, the resulting levels were considered fun to play, and the design tool allowed a satisfying control on the final desired level of complexity. We have also applied PCG for the generation of imaginary worlds for fantasy video games. We have based the procedural generation on an approximated simulation of the physical phenomena at the basis of the evolution of our planet, starting from the placement of physical elements like mountains, oceans, etc., and then adding inhabited areas to the map. In this stage, the tool has been designed in order to consider specific places created using tools to interactively help the production of narrative structures for stories to include in games. -)PCG based on evolutionary algorithms: we have focused our research on novel methods to exploit the characteristics of Genetic Algorithms (GAs) in the generation of game contents. In particular, we have presented an algorithm to explicitly address the need to introduce more variety and unpredictability in the monsters inside MMOs, in order to avoid that the players consider the game repetitive and less enjoyable after a long period of time spent playing. The main idea was to characterize each monster specie present in the game through its genome, and to generate new species by recombining their chromosomes, which represent a set of physical characteristics and skills. Each monster was represented by a chromosome composed by 53 genes, and the recombination process evaluates also the probability for the new monster to actually survive in the habitat in which it is born. More recently, we have presented a method for the automatic generation of realistic layered materials, based on the application of a GA. We have shown how a GA can be applied efficiently to evolve the parameters of a Bidirectional Scattering Distribution Function (BSDF), in order to generate different versions of a target material presenting a moderate amount of perceptual differences.Pubblicazioni consigliate
I documenti in IRIS sono protetti da copyright e tutti i diritti sono riservati, salvo diversa indicazione.