Georgios Miaoulis

Intelligent Computer Graphics 2011

Realistic Skin Rendering on GPU.- Affective States in Behavior Networks.- Information Theory Tools for Viewpoint Selection, Mesh Saliency and Geometry Simplification.- Classifying Volume Datasets Based on Intensities and Geometric Features.- Light Source Storage and Interpolation for Global Illumination: A Neural Solution.- An Intelligent System for Overlaying Texts on Background Images Based on Computational Aesthetics.- Parallel Coordinates: Intelligent Multidimensional Visualization.- An Adjectival Interface for Procedural Content Generation.- An SVM/GA Hybrid Framework for Qualitative Knowledge Aided 3D Scene Synthesis.- Machine Learning and Pattern Analysis Methods for Profiling in a Declarative Collaorative Framework.- AURAL: An Evolutionary Interface for a Robotic Sonification Process.

Active learning of user's preferences estimation towards a personalized 3D navigation of geo-referenced scenes

The current technological evolutions enter 3D geo-informatics into their digital age, enabling new potential applications in the field of virtual tourism, pleasure, entertainment and cultural heritage. It is argued that 3D information provides the natural way of navigation. However, personalization is a key aspect in a navigation system, since a route that incorporates user preferences is ultimately more suitable than the route with the shortest distance or travel time. Usually, user’s preferences are expressed as a set of weights that regulate the degree of importance of the scene metadata on the route selection process. These weights, however, are defined by the users, setting the complexity to the user’s side, which makes personalization an arduous task. In this paper, we propose an alternative approach in which metadata weights are estimated implicitly and transparently to the users, transferring the complexity to the system side. This is achieved by introducing a relevance feedback on-line learning strategy which automatically adjusts metadata weights by exploiting information fed back to the system about the relevance of user’s preferences judgments given in a form of pair-wise comparisons. Practically implementing a relevance feedback algorithm presents the limitation that several pair-wise comparisons (samples) are required to converge to a set of reliable metadata weights. For this reason, we propose in this paper a weight rectification strategy that improves weight estimation by exploiting metadata interrelations defined through an ontology. In the sequel, a genetic optimization algorithm is incorporated to select the most user preferred routes based on a multi-criteria minimization approach. To increase the degree of personalization in 3D navigation, we have also introduced an efficient algorithm for estimating 3D trajectories around objects of interest by merging best selected 2D projected views that contain faces which are mostly preferred by the users. We have conducted simulations and comparisons with other approaches either in the field of on-line learning or route selection using objective metrics in terms of precision and recall values. The results indicate that our system yields on average a 13.76xa0% improvement of precision as regards the learning strategy and an improvement of 8.75xa0% regarding route selection. In addition, we conclude that the ontology driven weight rectification strategy can reduce the number of samples (pair-wise comparisons) required of 76xa0% to achieve the same precision. Qualitative comparisons have been also performed using a use case route scenario in the city of Athens.

Intelligent Scene Modelling Information Systems

This book is dedicated to intelligent scene modeling information systems, that is information systems using Artificial Intelligence techniques to design scenes. Declarative scene modeling techniques are presented, as well as their implementation in an intelligent information system. In order to improve efficiency of declarative modeling based scene modeling information systems, various techniques are proposed in the book: coupling of a declarative modeler with a classical geometric modeler. Use of machine-learning Artificial Intelligent techniques, allowing the system to learn the user preferences; introduction of high level concepts, especially the concept of style in architectural design; introduction of collaborative declarative modeling techniques.

Visual Complexity and Intelligent Computer Graphics Techniques Enhancements

Its coming again, the new collection that this site has. To complete your curiosity, we offer the favorite visual complexity and intelligent computer graphics techniques enhancements book as the choice today. This is a book that will show you even new to old thing. Forget it; it will be right for you. Well, when you are really dying of visual complexity and intelligent computer graphics techniques enhancements, just pick it. You know, this book is always making the fans to be dizzy if not to find.

8 User Profiling from Imbalanced Data in a Declarative Scene Modelling Environment

Declarative Modelling is an early-phase design technique allowing the user to describe an object or an environment in abstract terms, closer to human intuition. The geometric solutions automatically yielded for such a description are evaluated by the user and may be subsequently used for the construction of a computational model of his/her preferences. Due to the physical limitations of the human evaluator, and the large number of the representations produced, only a subset of the latter are actually evaluated by the user and eventually a small number of them are approved, leading to imbalanced datasets in regard to the learning mechanism invoked. In the current work we discuss and assess the capability of a mechanism adopted for user modelling in a declarative design environment to handle this imbalance. The experimental results in this context indicate considerable efficiency in the prediction for the under-represented class.

Using Visual Representation for Decision Support in Institutional Research Evaluation

Higher Education Institutes worldwide are facing an increased demand to strengthen their capacities for research and innovation. This study introduces an ontology-based software system architecture that supports research policy evaluation processes and decision-making strategies, using visual analytics. A knowledge modeling technique drawing on multi criteria analysis and data visualisation is proposed. In addition, the paper presents a prototype built on Protege, Pellet reasoner and Java Technologies, which is friendly to the user and capable of interactive synthesis of institutional decision support criteria. In this work we make a transition from knowledge to visual web-based decision support systems with different kinds of visualisations. The developed system enables research managers to evaluate key aspects of academic research activity in the context of specific policies and criteria, correlate strategic goals with research performance and make informed decisions on the establishment of research strategies.

A Collaborative Framework for Virtual Globe Based 3D City Modeling

Urban planning and city modeling have gradually shifted, during the last decade, from highly demanding, in terms of required skills and supporting information, to tasks now supported by efficient, widely available applications, thus becoming popular and largely accessible. There are many GIS systems nowadays that offer a freely navigable three dimensional environment. This evolution of GIS systems has, in turn, led to the requirement for and creation of virtual 3D models of the ground and buildings. Online communities have created and distributed over the Internet libraries of georeferenced 3D models. The public is encouraged to participate in the design of 3D scenes and many companies offer free tools to facilitate the design of 3D models, specialized in buildings. In this paper, we present a collaborative approach for the construction of a city model, and its implementation through a prototype environment, employing freely available design tools. The prototype system comprises a collaborative database, supported by a web-based interface. Users are able to create and upload their models to the common database over the web, thus constructing a realistic 3D city model in a given area in a collaborative manner.

Machine Learning and Pattern Analysis Methods for Profiling in a Declarative Collaorative Framework

One important issue for a collaborative design framework is the intuitive manner designers describe scenes. To address the humans’ subjective description of a scene, personalization mechanisms are incorporated in the proposed architecture. In this paper, we examine different pattern analysis and machine intelligence algorithms for profiling in such framework. Two main approaches are described; the single (independent) profile estimation and the collaborative (dependent) case. Experimental results are presented to illustrate the efficiency of the proposal user’ profile methodologies.

Verbalization of 3D Scenes Based on Natural Language Generation Techniques

Basic ideas and requirements can be outlined by a design through ambiguous terms in order to define a desirable scene. Declarative modeling approach receives a rudimentary description and produces a set of scenes that are close to designer view. The reverse declarative modeling paradigm supports the designer to distinguish a set of scenes, accommodate further the pre-selected scenes to his needs, and produces a new enriched declarative description which initiates a new forward declarative design cycle for new promising scenes. The aim of the present work is to enhance the communication between the designer and machine, in such a way to increase the designer understanding and perception, by structuring a description in textual mode, reflecting all necessary semantic and geometric information, whenever the designer alters the pre-selected scenes.

Elicitation of User Preferences via Incremental Learning in a Declarative Modelling Environment

Declarative Modelling environments exhibit an idiosyncrasy that demands specialised machine learning methodologies. The particular characteristics of the datasets, their irregularity in terms of class representation, volume, availability as well as user induced inconsistency further impede the learning potential of any employed mechanism, thus leading to the need for adaptation and adoption of custom approaches, expected to address these issues. In the current work we present the problems encountered in the effort to acquire and apply user profiles in such an environment, the modified boosting learning algorithm adopted and the corresponding experimental results.