ISGW '97
August 17-19, 1997
Stevenson, Washington

Workshop Panels


ISGW '97 Home Page

PANEL 1 - Sunday August 17th - 4:45 PM

PANEL TITLE Tele-immersion: Developing a Long-term Research Agenda

PANEL ORGANIZERS Maxine Brown (Electronic Visualization Laboratory, University of Illinois at Chicago) and Greg R. Luecke (Iowa State University)

PANEL DESCRIPTION

"Tele-Immersion" -- collaborative virtual reality (VR) over networks -- is the extension of the "human/computer interaction" paradigm to "human/computer/human collaboration," with the computer providing real-time data in shared, collaborative environments, to enable computational science and engineering researchers to interact with each other (the "tele-conferencing" paradigm) as well as their computational models, over distance, using advanced real-time 3D immersive interfaces. This panel provides an opportunity for participants to present, compare, contrast, and prioritize their various research agendas to enable virtual collaboration. Is VR is a viable technology for scientific problem-solving, regardless of distance? We are striving to make VR an integral part of the country's advanced computational infrastructure for the 21st century, whether the researchers are in the same room or spread across the nation. Core research foci include, but are not limited to:

POSSIBLE PANELISTS

PANEL 2 - Monday, August 18th - 8:45 AM

PANEL TITLE The Role of Interactive Systems in Universal Access

PANEL ORGANIZER Alan Biermann

PANEL DESCRIPTION

Multimedia information technology is now at the stage where it may be able to create a new infrastructure to facilitate the deployment of knowledge in service to society. Such an infrastructure must be based on new research into methods of information representation that are display and modality independent. Such methods may involve ontologies, frames, semantic networks, annotation techniques, combinations of these, and other newer approaches for representing, storing, and transmitting information. New theories of information measurement, summarization, expression, and co-expression must be developed and serve as the foundation for such methods. The success of such efforts will be measured by the degree to which all citizens may both provide and access information and thus share knowledge regardless of ability or technology and independent of place or time. The National Science Foundation in carrying out its strategic plan must now push the envelope of science and technology to explore these possibilities and to open up science and technology to all segments of society. The panelists are asked to address the following issues: 1) What are the key scientific research issues, the understanding of which would enable universal citizen access to information and global knowledge without danger of information caste formation? 2) Is there a form of understanding or representation that is modality independent, whereby one can be assured that people who access information via different modalities are getting the same information? Are there pairs of modalities for which this is true? Are there types of information for which this is true? 3) To what extent can factors such as usability, learnability, confidence, and believability be both scientifically understood and promoted in information access environments such as digital libraries and digital government interfaces? 4) What is the role of dialogue and adaptive interactivity (beyond query-based information retrieval) in achieving universal access and the goal of every citizen involvement in the information society?

POSSIBLE PANELISTS



PANEL 3 - Monday, August 18th - 10:30 AM

PANEL TITLE Modeling the way humans produce and perceive speech: Advances, challenges, and applications

PANEL ORGANIZERS Abeer Alwan and Michael Hoffman

PANEL DESCRIPTION

Quantitative models of human speech production and sound perception provide important insights into our cognitive abilities and have led to high-quality computer synthesis of speech, robust automatic speech recognition (ASR), and efficient speech and audio coders. Recently, with rapid advances in hearing research, imaging, non-linear analysis, and computer technology, more sophisticated speech production and perception models are being proposed. In this panel, we will discuss recent advances in signal processing techniques that exploit what we know about speech production and sound perception modeling. In addition, we will summarize the challenges that lie ahead in developing robust, integrated interfaces to facilitate reliable and useful human-computer interaction.

POSSIBLE PANELISTS

PANEL 4 - Monday, August 18th - 6:00 PM

PANEL TITLE Toward Conversational Agents

PANEL ORGANIZER Dominic Massaro

PANEL DESCRIPTION

This panel will address the challenges faced in the development of artificial conversational agents-- simulations of human(oid) heads that are able to engage in "conversations" with people.
  1. challenges (& rewards) of integrating non-verbal devices into conversational agents;
  2. the integration of auditory speech recognition, visual speech recognition and gesture recognition, auditory speech generation, and visual speech generation (talking face with appropriate expression of emotion);
  3. synthesizing Conversation Between Human-Like Cooperative Agents; and
  4. Communication, Co-articulation, and Dialog Gesture in Facial Animation

POSSIBLE PANELISTS

PANEL 5 - Tuesday, August 19th - 7:30 AM

PANEL TITLE Information Overload Tools -- Getting Past Hyperbole to Comparison

PANEL ORGANIZER Joe Konstan

A great deal of research is being conducted around the general problem of finding valuable information in a sea of data. The CHI community sees regular "debates" between agent and direct manipulation advocates, but there is little work underway to seriously compare technologies and research approaches on neutral turf. The purpose of this panel is twofold: to have panelists _briefly_ describe the problem domains in which they think their approach is most promising, and to have the panel as a whole discuss what steps are needed to establish metrics, build corpora and testing facilities, and commence meaningful evaluation that matches technological approaches to problems.

POSSIBLE PANELISTS:

Tim Carmell : Last modified: August 6, 1997