Catálogo de publicaciones - libros

The present work presents a novel framework for 3D content-based search and retrieval. On contrary to most state-of-the-art approaches, the query model can be not only an existing object from a database but also a model manually generated by the user using sketches. In the context of the proposed framework, three interfaces to the sketch-based 3D search application were tested and comparative results were extracted according to usability and efficiency criteria.

Designing affective Human Computer-Interfaces such as Embodied Conversational Agents requires modeling the relations between spontaneous emotions and behaviors in several modalities. There have been a lot of psychological researches on emotion and nonverbal communication. Yet, these studies were based mostly on acted basic emotions. This paper explores how manual annotation and image processing might cooperate towards the representation of spontaneous emotional behavior in low resolution videos from TV. We describe a corpus of TV interviews and the manual annotations that have been defined. We explain the image processing algorithms that have been designed for the automatic estimation of movement quantity. Finally, we explore several ways to compare the manual annotations and the cues extracted by image processing.

MPEG-4 facial animation parameters are used in order to evaluate theoretical predictions for intermediate expressions of a given emotion episode, based on Scherer’s appraisal theory. MPEG-4 FAPs and action units are combined in modelling the effects of appraisal checks on facial expressions and temporal evolution issues of facial expressions are investigated. The results of the synthesis process can then be applied to Embodied Conversational Agents (ECAs), rendering their interaction with humans, or other ECAs, more affective.

Ambient Intelligence research is about ubiquitous computing and about social and intelligent characteristics of computer-supported environments. These characteristics aim at providing inhabitants or visitors of these environments with support in their activities. Activities include interactions between inhabitants and between inhabitants and (semi-) autonomous agents, including mobile robots, virtual humans and other smart objects in the environment. To provide real-time support requires understanding of behavior and activities. Clearly, being able to provide real-time support also allows us to provide off-line support, that is, intelligent offline retrieval, summarizing, browsing and even replay, possibly in a transformed way, of stored information. Real-time remote access to these computer-supported environments also allows participation in activities and such participation as well can profit from the real-time capturing and interpretation of behavior and activities performed supported by ambient intelligence technology. In this paper we illustrate and support these observations by looking at results obtained in several European and US projects on meeting technology. In particular we look at the Augmented Multiparty Interaction (AMI) project in which we are involved.

In this paper we present a web based tool for authoring adaptive assessments based on IMS QTI, IMS LIP and IEEE PAPI learning standards. The tool allows authors of assessments to customize the applied adaptation strategies to meet their own teaching experiences and goals. During the assessment procedure the system accumulates knowledge about learners and adapts the exposure of items based on this knowledge and on the adaptable rule model. Design decisions and the employed user, domain and rule models are analyzed and discussed. Results from a usability evaluation are discussed at the end of the paper.

In this paper, we use adaptable instructional digital images. The images used are the result of screen capture. These images illustrate the specific use of a spreadsheet software package in order to complete a given task. This task consists from several parts, and for these parts the corresponding images are given in different versions. Each version is using a different amount of details for the same part of the task. The task is accessible to the students through a single html file hosted on a web site, available through an Intranet. Via this web site students choose their preferable “knowledge path” (that is: the choice of image files they make in order to view the whole task). In the experimental course set, there are 3 steps where students choose between 2 versions for each, giving altogether 8 different choices (knowledge paths) that a student can choose from.

This paper presents e-Class Personalized (e-CP), a new extension of the widely available open source Learning Content Management System e-Class. e-CP monitors interaction of the users of e-Class with its content and services and adapts the services to better suite the users’ interests and tasks. E-CP has been tested for over a year at the University of Patras e-Class server with over 20,000 users. It drew positive response by the user population that were exposed to this version of the leaning Content Management System. In this paper we discuss architectural decisions and evaluation results of e-CP.

When we trace the history of mind and learning theories we clearly see a transitioning course that traverses Cartesianism, Behaviorism, and finally Functionalism. Current advances in computer scanning technologies reinforce the view that learning should also be examined under the prism of brain-centered materialist theories. Adaptive learning systems are instructional technologies that try to minimize the mismatch between learner needs and the learning environment. Currently, they try to elicit the learner needs with performance measures but they ignore learner differences at the brain level. This paper offers a shift of viewpoint in thinking about future adaptive learning systems. If we want education to be precisely tailored to the needs of learners then instructional technologies must take advantage of known individual differences in brain processing. The paper offers the justification of such an approach, analyzes its implications, proposes an implementation model, reviews related work, and outlines future challenges.

Users of hypermedia systems are heterogeneous, came from different backgrounds and have different level of knowledge in the use of information systems. This paper discusses a multi-layered approach for developing personalized e-books. The layers regard both Human Computer Interaction/Navigation level and the Content of the e-book. These layers are defined and described in detail. The tei-Book architecture is also presented in the paper and it is proposed as suitable for providing personalized e-book views to the various e-book readers. This is done by an E-Book manager that manage to provide the appropriate e-book view according to the User Profiles and Book Profile Components that dynamically are developed by the tei-Book. Future work includes the evaluation of the personalized e-book views and also the evaluation of system’s usability.

We present a conditional rewrite system for arithmetic and membership univariate constraints over real numbers, designed for computer assisted learning (CAL) in elementary math. Two fundamental principles guided the design of the proposed rewrite rules: (emulating steps students should take) and , aiming that step-by-step solutions to problems look like ones carried out by students. In order to gain more flexibility to modify rules, add new ones and customize solvers, the rules are written in a specification language and then compiled to Prolog. The rewrite system is complete for a relevant subset of problems found in high-school math textbooks.