Catálogo de publicaciones - libros

COMPASS2008 is a general service platform developed to be utilized as the tourist and city explorers assistant within the information services for the Beijing Olympic Games 2008. The main goals of COMPASS2008 are to help foreigners to overcome language barriers in Beijing and assist them in finding information anywhere and anytime they need it. Novel strategies have been developed to exploit the interaction of multimodality, multilinguality and cross-linguality for intelligent information service access and information presentation via mobile devices.

The rapid success of the Internet has spurred a continuous development of web-based technologies. Several applications also feature multimedia environments which are very appealing and can effectively convey knowledge, also in a life-long learning perspective. This paper presents aims and developments of the ChiKho EU project. ChiKho has designed a web-distributed educational game which allows players to share and improve their knowledge about the heritage of European cities and countries. The paper also describes the ChiKho’s logical structure, the interaction modalities and the technical architecture. We finally presents preliminary usability results from tests with final users which we performed in the four ChiKho exhibition sites (London, Genoa, Plovdiv and Kedainiai), where we launched a prototype version of the web game.

This paper describes Squidball, a new large-scale motion capture based game. It was tested on up to 4000 player audiences last summer at SIGGRAPH 2004. It required the construction of the world’s largest motion capture space at the time, and many other challenges in technology, production, game play, and study of group behavior. Our aim was to entertain the SIGGRAPH Electronic Theater audience with a cooperative and energetic game that is played by the entire audience together, controlling real-time graphics and audio by bouncing and batting multiple large helium-filled balloons across the entire theater space. We detail in this paper the lessons learned.

Many computer games use custom scripts to control the ambient behaviors of non-player characters (NPCs). Therefore, a story writer must write fragments of computer code for the hundreds or thousands of NPCs in the game world. The challenge is to create entertaining and non-repetitive behaviors for the NPCs without investing substantial programming effort to write custom non-trivial scripts for each NPC. Current computer games have simplistic ambient behaviors for NPCs; it is rare for NPCs to interact with each other. In this paper, we describe how generative behavior patterns can be used to quickly and reliably generate ambient behavior scripts that are believable, entertaining and non-repetitive, even for the more difficult case of interacting NPCs. We demonstrate this approach using BioWare’s Neverwinter Nights game.

First person games are computer games, in which the user experiences the virtual game world from an avatar’s view. This avatar is the user’s alter ego in the game. In this paper, we present a telepresence interface for the first person game Quake III Arena, which gives the user the impression of presence in the game and thus leads to identification with his avatar. This is achieved by tracking the user’s motion and using this motion data as control input for the avatar. As the user is wearing a head-mounted display and he perceives his actions affecting the virtual environment, he fully immerses into the target environment. Without further processing of the user’s motion data, the virtual environment would be limited to the size of the user’s real environment, which is not desirable. The use of Motion Compression, however, allows exploring an arbitrarily large virtual environment while the user is actually moving in an environment of limited size.

Presentations on public information systems, like a large screen in a museum, usually cannot support heterogeneous user groups appropriately, since they offer just a single channel of information. In order to support these groups with mixed interests, a more complex presentation method needs to be used. The method proposed in this paper combines a large stationary presentation system with several Personal Digital Assistants (PDAs), one for each user. The basic idea is to ”overwrite” presentation parts on the large screen, which are of little interest to a particular user, with a personalized presentation on the PDA. We performed an empirical study with adult participants to examine the overall performance of such a system (i.e. How well is the information delivered to the users and how high is the impact of the cognitive load?). The results show, that after an initial phase of getting used to the new presentation method, subjects’ performance during parallel presentations was on par with performance during standard presentations. A crucial moment within these presentations is whenever the user needs to switch his attentional focus from one device to another. We compared two different methods to warn the user of an upcoming device switch (a virtual character ”jumping” from one device to another and an animated symbol) with a version, where we did not warn the users at all. Objective measures did not favour either method. However, subjective measures show a clear preference for the character version.

In this paper we address the problem of performing references to wall mounted physical objects. The concept behind our solution is based on virtual characters. These characters are capable of performing reasonable combinations of motion, gestures and speech in order to disambiguate references to real world objects. The new idea of our work is to allow characters to migrate between displays to find an optimal position for the reference task. We have developed a rule-based system that, depending on the individual situation in which the reference is performed, determines the most appropriate reference method and technology from a number of different alternatives. The described technology has been integrated in a museum guide prototype combining mobile- and stationary devices.

In this paper, we introduce a novel approach to the use of AI technologies to support user experience in Virtual Reality Art installations. The underlying idea is to use semantic representations for interaction events, so as to modify the course of actions to create specific impressions in the user. The system is based on a game engine ported to a CAVE-like immersive display, and uses the engine’s event system to integrate AI-based simulation into the user real-time interaction loop. The combination of a set of transformation operators and heuristic search provides a powerful mechanism to generate chain of events. The work is illustrated by the development of an actual VR Art installation inspired from Lewis Caroll’s work, and we illustrate the system performance on several examples from the actual installation.

Humor is an aspect of human behavior considered essential for inter-personal communication. Despite this fact, research in human-computer interaction has almost completely neglected aspects concerned with the automatic recognition or generation of humor. In this paper, we investigate the problem of , and bring empirical evidence that computational approaches can be successfully applied to this task. Through experiments performed on very large data sets, we show that automatic classification techniques can be effectively used to distinguish between humorous and non-humorous texts, with significant improvements observed over apriori known baselines.

We propose an AmbientBrowser system that helps people acquire knowledge during everyday activities. It continuously searches Web pages using both system-defined and user-defined keywords, and displays summarized text obtained from pages found by searches. The system’s sensors detect users’ and environmental conditions and control the system’s behavior such as knowledge selection or a style of presentation. Thus, the user can encounter a wide variety of knowledge without active operations. A pilot study showed that peripherally displayed knowledge could be read and could engage a user’s interest. We implemented the system using a random information retrieval mechanism, an automatic kinetic typography composer, and easy methods of interaction using sensors.