Computer Graphics International 2002, 1-5 July, Bradford, UK

"What is Your Relationship with Your Information Space?"

Jim J. Thomas, David R. McGee, Olga A. Kuchar, Judith W. Graybeal, Dennis L. McQuerry, and Pamela L. Novak
Pacific Northwest National Laboratory1
Richland, WA 99352
{firstname.lastname}@pnl.gov , www.pnl.gov/infoviz

Abstract
This invited paper describes a vision for a fundamentally new approach to finding the unexpected and verifying the expected in massive information spaces. Rather than communicate with our information spaces using abstractions, our relationship and interaction with our information spaces are that of a master to its slave. Today, we specify searches and our information resources respond to our specifically worded queries. However, when dealing with massive information spaces, determining how to construct the queries themselves is daunting. Instead, we argue that the information spaces themselves must be given sufficient latitude to support a human-information discourse, by (1) developing its own initiative and thereby supporting a more equal communication style, (2) presenting information within a context that can itself be relied upon as an artifact of communication, (3) while creating a two way dialogue for query and thought refinement. We will motivate the change from human-computer interaction into human-information interaction, discuss higher order interactions with information spaces, and address the technical challenges in achieving this vision.

"Problems and solutions for 3D accurate functional modeling of the hip and the shoulder"

Nadia Magnenat-Thalmann, MyungJin Kang, Taro Goto
MIRALab, CUI, University of Geneva 24, rue du General Dufour, Geneva, Switzerland

Abstract

This paper presents the state-of-the-art on 3D functional modeling of the hip and the shoulder and analyses critical aspects and solutions. For the accurate human hip and shoulder model, we need to give a careful consideration for three-dimensional surface modeling, correct material properties about organs (bones, cartilages, ligaments, muscles, and skins) and joint mechanics based on the correct information. Finally, as one solution among many others, we suggest a method to model 3D individual articulations.

Keywords: 3D functional modeling, Hip, Shoulder, 3D reconstruction, Magnetic Resonance Imaging, Deformable model, Registration

J. Jiang,
Department of Electronic Imaging & Media Communications University of Bradford, United Kingdom

Abstract

As millions of images are being created and stored in computer networks and more and more images are already presented in compressed format at the source, it becomes increasingly important to consider direct image indexing and retrieval in compressed domain rather than in pixel domain. In this paper, we describe a range of such algorithms out of our recent research work to provide a platform for further research and further development of robust, efficient and effective software tools for many applications in the community of IT and computer science. As worldwide efforts for practical image compression are represented by JPEG international standardization activities, this presentation will focus on automatic image indexing directly in the compressed domain via: (a) DCT-based JPEG; (b) wavelets-based JPEG-2000; and (c) prediction-based JPEG-LS.

Since human interpretation of image content is characterized by high level activities, attempts by the community of image processing and computer vision in extracting low-level features in pixel domain invites many questions about the accuracy of image indexing and retrieval. To this end, feature-extraction approach or automatic signatures of images often fail to produce satisfactory solution for retrieving images based on their content. As a result, pixel values are often desired for such content analysis or visual inspections. Therefore, one section out of this paper is contributed to the description of our work towards extracting not only features, but also a complete image for such content access. Compared with full decompression, the image extraction technology described features in low computing cost and low complexity, which essentially bridges the gap between the compressed domain and the pixel domain.

M. Jern, S. Palmberg, M. Ranlöf ITN,
University of Linköping Norrköping, 601 74 Sweden

Abstract

"Collaborative Visualization" refers to development of scientific understanding, which is mediated by scientific visualization tools in a collaborative context. A "Collaboratory" is a visual computing environment allowing project teams to collaborate and share data and insight while distributed over a network, using intuitive visual navigation techniques "Visual Conference Calls". The distributed architecture is based on Application Component Sharing and a network abstraction system for games providing real-time data interactivity. The innovation can be summarised as the integration of network technology used in collaborative games with 3D data navigation techniques and based on a low-level, fine-grained "atomic" component infrastructure, implemented and validated in real-world environments.

In this paper we focus on 3D medical imaging tools describing a collaboratory implementation of a "VolumeViewer". A Web based application component for 2D and 3D visualizations of MRI, CT and scientific 3D volume data that supports peer-to-peer collaborative sessions on PC desktops. The EC funded research project SMARTDOC proposes how collaborative visual data navigation technology, with Web-enabled application sharing, can be used routinely. The SMARTDOC project is also a blueprint to inform researchers and software engineers about the possibility to extend existing visualisation applications to support collaboration, e.g. making it easier to "collaboratize" visualization applications.

Keywords: collaborative visualization, volume visualization, visual user interface

Abstract

Simulating a human society is a very complex interdisciplinary problem. In this paper, we try to show some important aspects to create inhabited worlds where virtual people can interact, co-operate, perceive the world and the society. Main aspects are flexible motion control, perception of the real and the virtual world, and high-level behavior. Concepts are illustrated in the case-study of emergent crowds.

There have been two major leading and enabling technologies making this possible, semiconductors and the deployment of high-bandwidth; and there have been two trailing technologies, content, and rights protection.