Cyril Brom

Pogamut 3 Can Assist Developers in Building AI (Not Only) for Their Videogame Agents

Many research projects oriented on control mechanisms of virtual agents in videogames have emerged in recent years. However, this boost has not been accompanied with the emergence of toolkits supporting development of these projects, slowing down the progress in the field. Here, we present Pogamut 3, an open source platform for rapid development of behaviour for virtual agents embodied in a 3D environment of the Unreal Tournament 2004 videogame. Pogamut 3 is designed to support research as well as educational projects. The paper also briefly touches extensions of Pogamut 3; the ACT-R integration, the emotional model ALMA integration, support for control of avatars at the level of gestures, and a toolkit for developing educational scenarios concerning orientation in urban areas. These extensions make Pogamut 3 applicable beyond the domain of computer games.

What does your actor remember? towards characters with a full episodic memory

A typical present-day virtual actor is able to store episodes in an ad hoc manner, which does not allow for reconstructing the actors personal stories. This paper proposes a virtual RPG actor with a full episodic memory, which allows for this reconstruction. The paper presents the memory architecture, overviews the prototype implementation, presents a benchmark for the efficiency of the memory measurement, and details the conducted tests.

Implementing digital game-based learning in schools: augmented learning environment of `Europe 2045'

It is widely agreed that the traditional process of schooling can benefit from the usage of computers as supportive tools. Of various approaches using computers in education over the last decade, e-learning and edutainment have become the most prominent. Recently, a number of authors have criticised these approaches arguing that they conserve traditional ‘drill and practice’ behaviouristic methods of teaching instead of enhancing and augmenting them. It has been proposed that a ‘paradigm shift’ is needed and that this shift may come through utilizing all the advantages of full-fledged video games, so-called digital game-based learning (DGBL). However, several case studies reported serious problems with the DGBL. Among the most notable issues are the lack of acceptance of games as an educational tool, problems with integration of games into formal schooling environments, and the so-called transfer problem, which is the problem of the inherent tension between game play and learning objectives, the tension that mitigates the ability of students to transfer knowledge gained in the video game to the real-world context. Here, we present a framework for an augmented learning environment (ALE), which verbalises one way of how these problems can be challenged. The ALE framework has been constructed based on our experience with the educational game, Europe 2045, which we developed and which has been implemented in a number of secondary schools in the Czech Republic during 2008. The key feature of this game is that it combines principles of on-line multi-player computer games with social, role-playing games. The evaluation which we present in this paper indicates the successful integration of the game and its acceptance by teachers and students. The ALE framework isolates key principles of the game contributing to this success, abstracts them into theoretical entities we call action-based spaces and causal and grounding links, and condenses them in a coherent methodological structure, which paves the way for further exploitation of the DGBL by educational game researchers and designers.

Flow, social interaction anxiety and salivary cortisol responses in serious games

Serious games are supposed to instigate engagement and, in turn, improve learning. High engagement is frequently connected with a positive affective state and a high flow state. However, the alleged link between a learners affective state, his/her flow state and learning outcomes has not been investigated in detail in the context of serious games. Even less information is available on how serious games may influence markers of physiological arousal. To fill this gap, participants of this exploratory study (N?=?171) played one of the six different serious game-based treatments, while we measured their affect, flow, cortisol secretion and learning achievement. The treatments were supposed to generate different levels of engagement and cortisol responses, because some of them were designed for a single user, while others were team-based, featuring so-called social-evaluative threat (ST) components. Our results revealed that flow was positively related to positive affect and negatively to negative affect. While flow and positive affect were related to learning gains, almost no relationship between either of these three variables and cortisol levels was found. Negative affect and cortisol were elevated in social interaction anxious males in team-based conditions. This study contributes to the limited body of research on the relationship between engagement and learning in serious games. We provide new perspectives on the relationships between flow, positive/negative affect and cortisol. Our findings highlight the fact that team-based serious games with ST components may have adverse effects on learners, particularly males, with high social interaction anxiety. We studied the link between affect, flow, cortisol secretion and learning achievement.Learners participated in one of the six different serious game-based treatments.Flow was positively related to positive affect and negatively to negative affect.Flow/positive affect was related to achievement, but not to cortisol levels.Cortisol was elevated in social interaction anxious males in team-based conditions.

Designing an educational game: case study of 'Europe 2045'

This paper presents a theoretical framework, which has been adopted in designing an on-line multi-player strategy game Europe 2045. Europe 2045 is an educational tool for high school social science courses, aimed at familiarizing students with political, economic, and social issues in contemporary Europe. Apart from learning facts, players develop a range of key skills: discussion ability, negotiation, teamwork, and group decision-making. The presented theoretical framework is based on a critical analysis of crucial issues, which seem to determine the success or failure of development and implementation of an educational game in the formal school environment. It demonstrates key approaches the authors of Europe 2045 have adopted in order to overcome already known problems related to game-based learning. On a general level this paper discusses issues related to formal fact learning in educational systems and the possible role of educational games in enhancing these systems.

Simulation Level of Detail for Virtual Humans

Graphical level of detail (LOD) is a set of techniques for coping with the issue of limited computational resources by reducing the graphical detail of the scene far from the observer. Simulation LOD reduces quality of the simulation at the places unseen. Contrary to graphical LOD, simulation LOD has been almost unstudied. As a part of our on-going effort on a large virtual-storytelling game populated by tens of complex virtual humans, we have developed and implemented a set of simulation LOD algorithms for simplifying virtual space and behaviour of virtual humans. The main feature of our technique is that it allows for several degrees of detail, i.e. for gradualvarying of simulation quality. In this paper, we summarise the main lessons learned, introduce the prototype implementation called IVE and discuss the possibility of scaling our technique to other applications featuring virtual humans.

Episodic memory for human-like agents and human-like agents for episodic memory

Episodic memory has been approached from many levels of analysis and many of its facets have been modeled computationally. Recently, several models of episodic memory have emerged in the domain of intelligent virtual agents (IVAs). Compared to neuro-/psychological models, their plausibility is limited. On the other hand, they can store representations of large environments and other complex memories over long-time intervals. This paper presents one such model and discusses the possibility of using IVAs as a test-bed to investigate neuro-/psychological models. The conclusion is that IVAs and their virtual environments can constitute an ecologically plausible framework allowing for study and integration of the neuro-/psychological models.

Story manager in 'Europe 2045' uses Petri nets

Europe 2045 is an on-line multi-player strategy game aimed at education of high-school students in economics, politics, and media studies. The essential feature of the game is that players face various simulated scenarios and crises addressing contemporary key issues of the unified Europe. These scenarios are branching and can evolve in a parallel manner. In this paper, we present a technique for specifying plots of these scenarios, which underpins the story manager of Europe 2045. The technique is based on a modification of Petri Nets. We also detail one particular scenario concerning the current crisis in Darfur. On a general level this paper discusses the strengths and weaknesses of implementation of Petri Nets in virtual storytelling.

A virtual reality task based on animal research – spatial learning and memory in patients after the first episode of schizophrenia

Objectives: Cognitive deficit is considered to be a characteristic feature of schizophrenia disorder. A similar cognitive dysfunction was demonstrated in animal models of schizophrenia. However, the poor comparability of methods used to assess cognition in animals and humans could be responsible for low predictive validity of current animal models. In order to assess spatial abilities in schizophrenia and compare our results with the data obtained in animal models, we designed a virtual analog of the Morris water maze (MWM), the virtual Four Goals Navigation (vFGN) task. Methods: Twenty-nine patients after the first psychotic episode with schizophrenia symptoms and a matched group of healthy volunteers performed the vFGN task. They were required to find and remember four hidden goal positions in an enclosed virtual arena. The task consisted of two parts. The Reference memory (RM) session with a stable goal position was designed to test spatial learning. The Delayed-matching-to-place (DMP) session presented a modified working memory protocol designed to test the ability to remember a sequence of three hidden goal positions. Results: Data obtained in the RM session show impaired spatial learning in schizophrenia patients compared to the healthy controls in pointing and navigation accuracy. The DMP session showed impaired spatial memory in schizophrenia during the recall of spatial sequence and a similar deficit in spatial bias in the probe trials. The pointing accuracy and the quadrant preference showed higher sensitivity toward the cognitive deficit than the navigation accuracy. Direct navigation to the goal was affected by sex and age of the tested subjects. The age affected spatial performance only in healthy controls. Conclusions: Despite some limitations of the study, our results correspond well with the previous studies in animal models of schizophrenia and support the decline of spatial cognition in schizophrenia, indicating the usefulness of the vFGN task in comparative research.

Towards Fast Prototyping of IVAs Behavior: Pogamut 2

We present the platform for IVAs development in the human like environment of the first-person shooter game Unreal Tournament 2004. This environment is extendible and supported by vast community of users. Based on our previous experience the problem of fast verification of models of artificial intelligence or IVAs is in implementation issues. The developer spends most of his time solving technical environment dependent issues and malfunctions, which drives him away from his goals. Therefore our modular platform provides a tool, which helps solving those problems and the developer can spend saved time by solving another AI based issues and model verification. The platform is aimed for research and educational purposes.