The Believability Gene in Virtual Bots - Archive ouverte HAL Accéder directement au contenu
Communication Dans Un Congrès Année : 2016

The Believability Gene in Virtual Bots

Résumé

Video game development is not only one of the most profitable entertainment industries but also represents a very interesting field of research. Particularly in the area of Artificial Intelligence (AI), it provides many interesting (and hard to tackle) challenges. One of them consists in creating artificial bots (i.e., game characters not controlled by a human) that mimic human behavior or, at least, show a believable behavior. This paper deals with this issue by describing two very different approaches that were proposed for creating believable bots and applied, with some degree of success, in the context of a first-person shooter (FPS) game. One approach is based on the idea of imitating human player behavior, and the other one consists of automatically creating bots via an interactive evolutionary algorithm. The paper analyzes the performance of these proposals and introduces different forms of hybridizing both approaches for future applications.
Fichier non déposé

Dates et versions

hal-01315402 , version 1 (13-05-2016)

Identifiants

  • HAL Id : hal-01315402 , version 1

Citer

Mihai Polceanu, Antonio M. Mora, Jose L. Jimenez, Cédric Buche, J. Fernandez-Leiva. The Believability Gene in Virtual Bots. Twenty-Ninth International Florida Artificial Intelligence Research Society Conference, May 2016, Miami, United States. pp.346-349. ⟨hal-01315402⟩
94 Consultations
0 Téléchargements

Partager

Gmail Facebook X LinkedIn More