AGOCG logo
Graphics Multimedia VR Visualisation Contents
Training Reports Workshops Briefings Index

IMAGINA 97
19 - 21 February 1997, Monte Carlo

The helicopter transferring Imagina delegates between Nice and Monte Carlo approached the landing stage after making a swift, picturesque journey between the two cities. As it made its final descent, the engine noise became louder, and the swirling air swept the ground for another group of delegates. I was not impressed at all, for I was not on board. I was driving a Nissan Micra, and as it bounced down the last treacherous hill, with all three cylinders firing stochastically, and the exhaust singing a throaty song - nothing like a Ferrari Tostestorone, or is it Testerosser?

I parked it outside Monaco's Casino. I gave the gold-clad parking attendant a handful of Irish and UK coinage; he thanked me for my generosity, joked that BSE was not a problem any more, saluted, and I was ready for another Imagina conference.

I made my way to the Beach Plaza Hotel where I had stayed for Imagina '84. After registering, I enjoyed my first fruits-de-mere meal, and then collapsed into bed. The dawn chorus of hammering, pneumatic drills, and general major civil engineering noises woke me at 7.30 a.m. Yes, it was good to be back in Monte Carlo - the best-kept building site in Europe! It was remarkable to see lorries washed from bonnet to wheels, taking away rubbish and rubble before starting their journey across the streets of Monte Carlo. The British habit where a carefully placed sign predicting "Mud on Road" permits lorries to spread their cargo over our roads - rather than taking it to a land-fill site - is just not accepted in Monaco, and rightly so. Anyway, after complaining to the manager about the four-star noise I could hear from my bed, we compromised on a new daily tariff, and I promised not to write a single word about the inconvenience I had suffered during my stay at the Beach Plaza. My word is my bond! And I'll say nothing about the missing towels and flannels.

The theme of this year's conference was "Meta-mondes" or in English

"Meta-verses". Yes, I know, I could not recall this word either, but after reading Philippe Queau's introduction to the conference, all became clear. Philippe is responsible for organising the conference programme, and describes "Meta-verses" as "realms arising from the marriage of networks and three-dimensional images". I, personally, would have used the term "Meta-worlds" or "Meta-universes".

There were six major conference topics: Narration.Interaction, Virtual Communities and Networked Video Games, From Image to Model, Motion, Understanding 3D, and Subtle Special Effects. There were 35 presentations in all, so I will only report on those that I found interesting.

The conference began with Narration.Interaction, with six papers mainly from the USA. The first was given by Greg Roach who proposed a simple way of analysing interactive narrative properties. Andy Cameron, founder member of Anitrom, then developed the theme of interactive stories from our understanding of linear narrative, the multi-linear novels of the 1930s, to interactive games such as Tomb Raider.

The next speaker was Troy Bolotnick, President of LightSpeed Media, Inc. Bolotnick was concerned with Episodic Web Sites, which are sites that maintain audience loyalty by attracting them back, day-after-day, episode-after-episode. Personally, I can't see how it can ever work - my version of Netscape crashes every 30 minutes, making episodic browsing virtually impossible! But if the WWW replaces television in the next millennium, perhaps there will be a need for such sites.

Chris Crawford then gave us a primer on interactive storytelling, and introduced us to inclination equations, where it is possible to express boy meets girl scenarios as:

Inclination[SpurnKiss] = Loyalty[Gal for her Significant Other]

Inclination[Kiss] = Affection[Gal for Guy] + Trust[Gal for Guy]

Inclination[RedHotKiss] = Libido[Gal] + EventHappened [Gal for Guy]

Give me a good book anytime!

The second session was much more practical, and was about networked video games. To begin, the President of Sega Digital Communications, Yuzo Naritomi, described how the Sega Saturn could communicate with other units using a modem and IC card. Admittedly, the IC card is cheap ($18), but who pays for the telephone bill? NEXT!

Gurminder Singh, from the National University of Singapore, gave an interesting presentation on "KidSpace - Interactive Discovery Learning". It was good to see networked multimedia being put to such good use.

The third session was called From Image to Model, and the first paper was by Takeo Kanade, from the Robotics Institute at Carnegie Mellon University. Kanade described a technique called Virtualized Reality, where a person is placed inside a geodisic structure fitted with 51 video cameras at the vertices. Using depth maps, Kanade showed that it was possible to view the person inside the dome from any virtual position. He describes this process as "Virtualized Reality", and said he had a machine working at video frame rates. The next stage was to see how users could interact with the virtual environment. For me, this was an exciting project.

Thomas Vetter, from the Max-Planck-Institut in Germany described how an automatic morphing technique could be used to provide arbitrary views of human heads when only one 2D image was available. The system draws upon knowledge gained about the topology of human heads using faces of other people, and applies it to a single 2D image to create any 3D view.

Anyone who has modelled a large real-world, terrain database that includes buildings and other man-made structures, will know that such projects are very time consuming. In recent years, considerable research has been undertaken to develop semi-automatic procedures to capture geometry from the real world. Luc Robert's presentation "Realistic Scene Models from Image Sequences" showed how he could capture a wide range of textured objects from stereo-pairs. The technique looks very promising, and has some excellent applications in car simulators.

The fourth session was "Setting in Motion" with Hal Bertram from Jim Henson's Creature Shop, as Chairman. He spoke about the techniques used to create realistic motion, and how they could be transferred to real-time puppeteering and animation. Ken Perlin's paper "IMPROV: A System for the Creation of Real-Time Behaviour-Based Animated Actors" described an animation engine that could be used to associate realistic behaviours to virtual dancers and musicians. The technique employs a script to control layered, continuous, non-repetitive motions and interpolation to provide believable behaviours. One of the goals is to build a virtual band with full MIDI support!

Michiel Van de Panne, from the University of Toronto, spoke about the problem of animating human walking. Even today, with all of the software tools available to animators, walking remains a difficult process. The slightest imperfection in any animated walk attracts our attention immediately. Van de Panne described a closed-loop control process that enables him to control the walking behaviour of an animated person walking along arbitrary paths, and at different speeds.

As cloning is currently in the news, it seemed fitting that someone should address the problem of creating and animating virtual clones of real humans. This was the subject of Agnes Saulnier and Pierre-Emannuel Chaut from INA. Their paper brought us up to date with their research results, where they showed some excellent images of animated 3D heads.

The fifth session was "Knowing and Understanding in 3D". The first paper was by Fabio Pettinati from Apple Computer, Inc. who confirmed that it was still difficult to find information on the Internet. He blamed it on today's search engines, and went on to describe the new "HotSauce" product from Apple. HotSauce employs a 3D metaphor for navigating the Internet, and uses a Meta-Content Format (MCF) which describes information in files known as XSpaces. The HotSauce plug-in replaces a Web browser's standard Web page display with a window containing a graphical depiction of a particular XSpace. I can't wait to sell my PC and buy an Apple Mac!

There were then two interesting presentations by Jack Lancaster and Didier Laroche. Lancaster described the "BrainMap" project that allows users to navigate through a human brain exploring the various regions. The Web site database now contains data from 203 technical papers, 695 experiments, and 7221 locations. This is an excellent example of how knowledge can be disseminated over the Web. Laroche's paper explored a totally different set of geometry: the Marmaria site in Delphes, Greece. The research team behind this project modelled the Greek temple at Delphes using computer graphics, and integrated the virtual building with real-world terrain to create some exciting animations.

One way to stop delegates leaving at lunch time on the last day is to place an interesting subject for the final afternoon. At Imagina, this is always Film Special Effects. Jan Kounen and Rudolphe Chabrier from Mac Guff Ligne, started the session by describing their in-house software tools Trukor and Symbor. They used these, together with Alias/Wavefront to create some excellent effects for their latest film "Dobermann".

Kelley Ray, from Sony Pictures Imageworks, then described the subtle effects used in the film "The Craft". For me, the best effects are those that are not noticed as being synthetic, and Kelley Ray showed how they used 3D Studio to create and animate hundreds of very realistic butterflies.

Next there was the obligatory "Twister" presentation by ILM's Stefen Fangmeier. The credits for the film show an incredible number of people working on the special effects, which is why Bournemouth

University is launching an MA/MSc in Digital Special Effects this year! If you can't beat them, join them!

Finally, our own Mike Boudry from the Computer Film Company concluded the session by showing how he had created a wealth of effects in several major films. Thanks to people like Mike, the UK still maintains a serious reputation in film special effects, and rightly so!

There we are; another Imagina conference over; another fruits-de-mere, and I collapsed into bed. The following day I collected my car from the Casino, only to find the gold-clad parking attendant waiting to hand back the assorted Irish and UK coins I had accidentally given him. I apologised using the very best French verbs I could muster, such as the passive, conditional form of avoir, and the grovelling, past tense of desole, and handed him 6 trillion lire! (UK equivalent 47p). After loading my luggage, conference proceedings, etc, he pushed my trusty Nissan Micra in the direction of Nice. As I left the blue skies of Nice behind, little did I realise that a new strain of flu was waiting for me at Heathrow!

John Vince
Bournemouth University