Keynote Speeches:

We are pleased to announce two IMMERSCOM keynote speeches:

 Keynote Speech I:
	Title: Towards Immersive Interactive Media
	Speaker:Prof. Marc Cavazza,University of Teesside
	Date:9:00-10:00 Oct 10, Wednesday 


Keynote Speech II:
	Title: Camera,Projector,and Microphone Arrays in Immersive Videoconferencing
	Speaker: Dr. Harlyn Baker, Hewlett-Packard Laboratories
	Date:9:00-10:00 Oct 11, Thursday


The abstracts of the speeches and the bibliographies of the speakers are 
below.

Towards Immersive Interactive Media Prof. Marc Cavazza  University of Teesside U.K. 9:00-10:00 Oct 10, Wednesday

Abstract:

Narrative and Virtual Reality are the two Media alternatives to everyday reality. The prospect of their convergence, illustrated for instance by immersive storytelling (epitomised by the sci-fi device known as “Holodeck”), brings new opportunities for the development of future media. Based on our recent research in Virtual Reality Art and in immersive storytelling, I will discuss the potential development of immersive media. It is now exactly 10 years since Altman and Nakatsu have hypothesised the convergence of telecommunications, interactivity, VR and media that would create the context for that evolution: even if like most technological predictions its realisation may be delayed, it is essential that content is available as soon as the infrastructure is in place. The elements of immersive media can be described through two dimensions, which are content, accounting for aesthetic dimensions, and interactivity, which brings user experience to new levels. Our research in Virtual Reality Art has sought to re incorporate elements of interaction in the aesthetic content itself, by investigating how elements of physical reality (laws of Physics, causality) could be redefined as part of the artistic work itself.

More recently, we have explored the technical feasibility of immersive interactive story telling by implementing a first complete prototype. The system puts the user at the centre of an immersive narrative played by virtual actors: he can interact by means of physical interaction and speech, acting as one character in the story. We have demonstrated its potential by creating an immersive interactive version of a classical XIXth century French novel, Madame Bovary. The system operates in a CAVE-like immersive projection system powered by an 8-node PC cluster. Real-time visualization and interaction is supported by game engine technology, which has been adapted to the context of real-time stereoscopic displays using a specific software package, CaveUT developed in collaboration with several other research groups. The narrative engine, which supports the dynamic response of the story to real-time user intervention, is based on technology we have developed in previous research and has been used in desktop systems as well as Mixed Reality ones. This system is not only a first illustration of what immersive interactive storytelling could be, it also provides the technical elements to transfer more traditional media to an immersive setting, thereby supporting new explorations of the user experience.

Biography:

Marc Cavazza is a Professor in the School of Computing of the University of Teesside ( United Kingdom). His research group in Intelligent Virtual Environment specialises in Artificial Intelligence research in the context of Virtual and Mixed Reality. The group is recognised internationally for its research in interactive storytelling, which has received the best paper prize at both the first (2001) and the second (2003) International Conference on Virtual Storytelling and is amongst the most cited works in the field.

The group has attracted research funding of 2M€ over the past years and its research has featured regularly in the national and international press, including The Daily Telegraph (UK), The Times (UK), Le Monde > (France), Il sole 24 Ore (Italy), the Danish and French radio broadcasting corporations, RTBF (Belgian Television), and the ACM news.

Marc Cavazza has published over 170 refereed papers in the field of Intelligence User Interfaces, Virtual Reality and Entertainment Computing, and has served on the programme committee of over 50 conferences, including the ACM Autonomous Agents and Multi-Agent Systems, ACM Advanced in Computer Entertainment Technologies and the Intelligent Virtual Agents conference.

He’s been an invited speaker at Imagina 2003, Imagina 2005, ACM Virtual Reality Software and Technology 2003, and has given the opening keynote at the Conference on Technologies for Interactive Digital Storytelling and Entertainment (TIDSE) 2006 as well as the Artificial Intelligence and Simulation of Behaviour (AISB) Conference 2007 (Newcastle UK). He his co-chair of the 2007 International Conference on Virtual Storytelling.

Camera, Projector, and Microphone Arrays in Immersive Videoconferencing Dr. Harlyn Baker  Hewlett-Packard Laboratories U.S.A. 9:00-10:00 Oct 11, Thursday

Abstract:

Videoconferencing has a long and uninspiring history—from the Bell System’s early Picturephone to current PC and laptop CU-SeeMe/Skype-type offerings. Failure to attain widespread acceptance can be attributed to a variety of causes, including small images  with poor quality, low frame rate, high latencies, difficulty in setup, and audio-video mismatch with lack of lip-sync.  Despite this history, HP boldly moved toward the video conferencing market in 2004 when it joined with Dreamworks to develop a life-sized high-quality collaboration system called Halo. Its goal has been to bring people together in an environment that looks, sounds, and feels like being across the table. Exploiting the aesthetic principles that Dreamworks practices in bringing realism to movies, Halo is now a successful product with deployments in over 100 sites around the globe.  Our laboratories are looking over the horizon for this technology since, while Halo is as effective as the best of current technology permits, it isn’t yet what it aims to be—a near equivalent to “being there.”

In our laboratories, we are investigating ways to bring the interaction closer yet to reality.  These include producing vastly higher resolution and immersive displays (through ganging together large numbers of displays for increased brightness and contrast, with shaping to provide—for example—surround viewing), developing very high resolution and arbitrary-format panoramic cameras (again, through integration of large numbers of individual imagers for mosaicked capture), delivering spatial sound with multi-path echo cancellation (by modeling room structure and dynamics, and adapting the audio field for this), and studying multi-viewpoint and 3D display.  While maintaining the quality of the Halo experience, our intention here is to move toward eliminating the barrier that distance presents, bringing people together in a flexible environment that meets—or exceeds—the functionality of reality.  We want clarity of display to meet human needs; immersive visual and auditory contexts that rival co-presence; the gaze alignment and consistency, eye contact, and gesturing that contribute so much to non-verbal communication; and we want to do all of this through scalable solutions providing flexible formats, with availability for everyone using single PCs and commodity components.  In this address I will show what Halo does now, and present what we are doing to work toward truly immersive remote collaborative experiences.

Dr. Harlyn Baker has worked in the field of computer vision for about three decades.  From his early studies in Edinburgh, addressing 3D stereo modeling and recognition, through his dynamic-programming stereo work at Stanford, to his developments at SRI, where he co-invented Epipolar Plane Image (EPI) Analysis and created the first spatiotemporal manifold representation process, he has been an innovator in multiple-image analysis. Applying his imaging insights more broadly, he has analyzed the structural implications of fossil data (Lucy and her contemporaries), and performed the earliest simulated surgical procedures/surface manipulations from acquired imagery. While at SRI, Dr. Baker received one of the first government contracts to model anatomy from the NIH’s Visible Human Dataset. Her has served on journal editorial boards, numerous international conference program committees, been the recipient of many international speaking invitations, and been published in major journals of his field.  On leaving Interval Research Corporation, Dr. Baker was a co-founder of a real-time stereo ranging company (TYZX).  A Senior Scientist in the Mobile and Media Systems Laboratory at Hewlett-Packard Laboratories in Palo Alto California since 2000, he was technical lead on an augmented-reality multi-participant videoconferencing technology (Coliseum), and developed camera systems in support of this and more recent multi-viewpoint imaging systems.  His current focus is on exploiting massively redundant imaging for omnidirectional and multidimensional visualization and spatial understanding.