Kahaner Report: Abstracts presented at VRST'94, Singapore, 8/94
Rick Schlichting
Office of Naval Research-Asia (ONR Asia) in Tokyo from NIST. The
following is the professional opinion of David Kahaner and in no
way has the blessing of the US Government or any agency of it. All
information is dated and of limited life time. This disclaimer should
be noted on ANY attribution.]
[Copies of previous reports written by Kahaner can be obtained using
anonymous FTP from host cs.arizona.edu, directory japan/kahaner.reports.]
From:
Dr. David K. Kahaner
US Office of Naval Research Asia
(From outside US): 23-17, 7-chome, Roppongi, Minato-ku, Tokyo 106 Japan
(From within US): Unit 45002, APO AP 96337-0007
Tel: +81 3 3401-8924, Fax: +81 3 3403-9670
Email: kah...@cs.titech.ac.jp
Re: Abstracts presented at VRST'94, Singapore, 8/94
10/10/94 (MM/DD/YY)
This file is named "vrst-abs.94"
ABSTRACT. Contains the abstracts of papers presented at Virtual Reality
& Software Technology (VRST'94), held at the Institute for Systems
Science (ISS), Singapore, 22-26 August 1994. An accompanying report
(vrst.94) contains a summary of this conference.
The technical program of the conference was divided into a series of
paper, panel and demonstration/video sessions. Abstracts of all
presented papers, demonstrations/videos and panel session opening
remarks are attached below. They are collected by session title and
appear in the order presented.
Abstracts of Presented Papers.
Following is a complete set of abstracts for all papers,
demonstrations/videos and panel session opening remarks presented at
VRST '94. They are collected by session title and in the order presented
at the conference.
Opening Session:
Keynote Address
FACING THE CHALLENGE: DELIVERING VIRTUAL REALITY
Michael F. Deering
Sun Microsystems Computer Corporation 2250 Garcia Avenue
Mountain View, CA 94043, USA
michael...@eng.sun.com
EXTENDED ABSTRACT
With more than half a decade of concept demonstrations behind it, it is
time for Virtual Reality to start facing up to the challenge of
delivering real systems. We have to be brave enough to state
quantitatively what makes a system Virtual Reality, as opposed to
advanced multimedia or just 3D graphics. The field has to find a way to
be inclusive while still stating that frame rates and lags that were
known to be unacceptable more than 200 years ago are still unacceptable
today.
By definition, to grow beyond a niche market, technical appalications of
Virtual Reality must be found with broad market appeal. To be fiscally
viable, the gains such applications bring in end-user productivity must
exceed any extra hardware costs.
There is room for optimism as well, however. Several recent technical
advances within Sun's VR effort will be described. These include: nearly
no-lag head tracking, HoloCam: a 32-channel stereo camera for live
telepresence, some recent dramatic advances in low cost high speed
rendering hardware, and HoloSketch: a completely VR based 3D sketching
animation authoring system. HoloSketch is an example of a potential
high volume application for technical Virtual Reality: a general purpose
tool for non-programmers who need to easily create and manipulate 3D
shapes and worlds.
Paper Session: Gestures
NATURAL GESTURE IN VIRTUAL ENVIRONMENTS
ALAN WEXELBLAT
MIT Media Laboratory 20 Ames St., Cantbridge, MA 02139 USA
w...@media.mit.edu
I present an experiment and the resulting prototype on a new method for
interacting with virtual environments. This method involves the capture
and use of natural empty-hand gestures. Users are allowed to gesture in
their normal continuous manner, rather than being restricted to a small
set of discrete gestural commands. Gestures are captured and analyzed
into a higher-level description. This description can be used by an
application-specific interpreter to understand the gestural input in
its proper context. An analyzer of this sort enables natural gesture
input to any appropriate application.
Toward a Vision-Based Hand Gesture Interface
Francis K.H. Quek
Electrical Engineering and Computer Science Department
The University of Illinois at Chicago
1007 SEO, 851 S. Morgan, Chicago, IL 60607
Email: qu...@eecs.uic.edu
May 11, 1994
ABSTRACT
We present a two-pronged approach to vision-based gesture recognition of
an unencumbered hand. Our goal is a gesture interface that makes sense
from the standpoint of human usage, and which can be implemented by
machine vision. We first explore gesture interpretation and usage
among humans. We glean rules and principles which permit us to define a
representative gesture taxonomy and provide information on gesture
segmentation and usage. This analysis indicates that the dynamics of
hand gestures are of critical importance in their segmentation and
interpretation. Second, we apply vector field computation to gesture
interpretation.
We extract vectors in a three-stage process. First, we extract moving
edges using a motion-enhancing edge detector. Second, we compute an
initial vector field estimate. Finally, we smooth this vector field to
produce a locally aligned field which minimizes noise. Results of our
system indicate that the approach is suitable for such interpretation.
A LOGICAL HAND DEVICE IN VIRTUAL ENVIRONMENTS
S. AUGUSTINE SU
Department of Computer Science, University of Maryland
College Park, Maryland 20742, USA
Email: S...@BUSH.CS.TAMU.EDU
and
RICHARD FURUTA
Department of Computer Science, Texas A&M University College Station,
Texas 77843-3112, USA
ABSTRACT
The human hands are the major means through which we gain our primary
connection to the world. The modeling of human hands is a very
important issue in virtual environments, however little research has
been done to support higher levels of abstraction of using hands beyond
that of just capturing raw data.
In this paper we present an alternative view of hand modeling, i.e., a
point-based hand model, and then investigate 3D static hand gestures in
detail. Thereby, we develop a device-independent and general-purpose
logical hand device, which supports the use of comprehensive 3D gestural
input in virtual environments. Based on our logical hand device, not
only can the implementation of "point, reach, and grab" interaction be
facilitated, but also American-Sign-Language-like static gestures can be
conceived easily.
Paper Session: Modeling & Animation
STEPS AND LADDERS IN VIRTUAL REALITY
MEL SLATER, MARTIN USOH, ANTHONY STEED
Department of Computer Science and London Parallel Applications Centre
Queen Mary and Westfield College, University of London,
Mile End Road, London El 4NS, UK.
Email: m...@dcs.qmw.ac.uk, big...@dcs.qmw.ac.uk, st...@dcs.qmw.ac.uk
ABSTRACT This paper describes a technique for walking through, and
climbing or descending steps and ladders in virtual reality. The idea is
that human participants in the virtual reality (VR) carry out a whole
body gesture similar to walking, by walking in place. A pattern analyser
distinguishes between walking in place behaviour, and any other
behaviour, and moves the participant through the environment when it
detects the walking in place. Such "walking" while on steps or ladders
similarly causes appropriate virtual movement. We discuss this in the
context of a paradigm for interaction called "body centred interaction
". This attempts to maximise the match between the mental body model
formed as a result of proprioceptive information generated by limb and
body movements, and the sensory data displayed by the VR system, within
the constraints imposed by limited tracking information. We argue that
the sense of presence in the VR is enhanced by such body centred
interaction techniques, and we present experimental evidence in support
of this claim.
FUZZY AGGREGATION OF MOTION FACTORS FOR HUMAN MOTION GENERATION
JIYUN LEE
Strategic Planning Team
Korea Telecom Research Laboratories
17, Su-Cho Ku, Woo-Myun Dong, Seoul, Korea
and
KWANGYEON WOHN
Dept. of Computer Science
Korea Advanced Institutc of Science and Technology
373-1, Yu-Sung Ku, Gu-Sung Dong, Taejon, Korea.
wo...@cs.kaist.ac.kr
ABSTRACT
Human motion generation is one of the important issues in computer
animation. To generate a realistic human motion, various information
of body joints are needed, and dynamics gives the solution to this.
However, formalizing complex human motions and solving these equations
are computationally expensive. To get an interactive and realistic human
motion, we made use of dynamic property which allows to use several key
human motion factors. In this paper, we select four key human motion
factors and propose an aggregation method using fuzzy logic. We
implemented an animation system based on this method, and numerous
experimental results showed that the proposed method generates an
interactive and realistic human motion. Such a human motion generation
can be easily adapted in virtual reality systems that make use of human
animation within virtual world.
A Geometric Modeling and Animation System for Virtual Reality
Sean Halliday, Mark Green
Department of Computing Science
University of Alberta, Edmonton, Alberta, Canada T6G 2H1
{halliday,mark}@cs.ualberta.ca
ABSTRACT
Very few virtual environment design tools have been developed for
nonprogrammers. Most non-programmers rely on programmers to produce
their environments, or produce static environments using traditional
geometrical modelets. Non-programmers, including artists and designers,
have a lot to contribute to the development of virtual environments. In
addition, tools developed to assist this user group will also benefit
other virtual environment developers. A highly interactive 3D geometry
modeling system and animation editor (JDCAD+) is presented in this
paper. The motivation behind this project is to develop interactive
tools for creating animated objects and importing them into virtual
environments. Without such tools, creating animated objects requires
coding the geometry and behavior of the object manually. Such models
are usually very time consuming to code and lack detail. Coding
behaviors for animation can also be very difficult due to complicated
equations of motion that need to be solved. Using key frame animation
has resulted in faster and easier behavior generation.
Demonstrations & Videos
UNIVERSITY OF ALBERTA 3D MODELING TOOLS
Mark Green, Sean Halliday, Jiandong Liang, Chris Shaw
Department of Computing Science, University of Alberta
Edmonton, Alberta, T6G 2H1, Canada
Email: MA...@CS.UALBERTA.CA
ABSTRACT
Over the past few years a number of 3D geometrical modeling tools have
been developed at the University of Alberta. The main thesis behind
these tools is that 3D design can best be performed using a 3D user
interface with 3D input devices. The geometrical modelers presented in
this demonstration will illustrate this idea and convince you that it is
reasonable. The first modeler, JDCAD, is a CSG modeler for the
mechanical engineering and industrial design communities. The second
modeler, JDCAD+, is an early version of JDCAD that includes interactive
behavior specification. The third modeler, Thred, is an experimental two
handed free form surface editor.
Interactive Visualization of a Beating Heart:
A Medical Application of the Virtual Workbench
S. Mciyappan and Heng Pheng Ann
Centre for Information-Enhanced Medicine
Institute of Systems Science, Heng Mui Keng Terrace, Singapore 0511
{meiyap, pheng}@iss.nus.sg
ABSTRACT
An interactive system for visualizing and manipulating beating heart
data has been enhanced by using the Virtual Workbench interface. The new
system allows intuitive visualization of time and space stacks of the
dynamic cardiac MRI data and it also provides a 3D interface for
manipulating such 4D data.
THE VIRTUAL ENVIRONMENT THEATER (TM) AS AN ARCHITECTURAL AND URBAN
DESIGN REVIEW SYSTEM
Peter C.C. Wong, Chet Dagit, Robert Jacobson
Worldesign Inc. 5348-1/2 Ballard Avenue NW
Seattle, WA 98107-4009 USA +1-206-781-5253 phone, -5254 fax
Email: info@worldesign. com
In June 1994, for the annual A/E/C SYSTEMS'94 conference - the building
industry's "SIGGRAPH" - Worldesign designed and built a Virtual
Environment Theater, or VETTM, as a prototypical virtual worlds-based
architectural and urban design review system. Over 3,000
professionals in the fields of architecture, engineering, and
construction (about a tenth of the attendees) experienced in the VET
several different potential versions of the Port of Seattle's "Central
Waterfront Development Project", as if they were involved in an actual
environmental impact review. The idea behind the VET is to
"virtualize", or make more immediately accessible, the vast amount of
information contained in the Port of Seattle's dense environmental
impact statement (EIS). In the United States, issuance of an EIS usually
precedes any major construction or development. The EIS then becomes an
object of debate between development proponents and opponents. The
public, which is supposed to be the prime beneficiary from this process,
is often excluded from the review since the information contained in the
EIS is presented in arcane language and without graphics or sound. The
VET permits representatives of the public to better understand what is
being proposed for a particular development and the implications for its
surrounding natural and built environments. Information can be imbedded
in visual and acoustic objects, so that buildings can "speak" their
prospective history and effects when called upon. During our showing of
the VET at A/E/C SYSTEMS, however, we discovered that the VET has a more
essential use, to help align the visions of the various parties - not
least, the clients - collaborating in the design, building, and use of a
new development or construction. Often the failure to align visions
causes the greatest disappointment (and expense) down the line, when the
designers, builders, and clients suddenly discover they do not have the
same vision of their common enterprise. In a shared virtual world, it
is easier to actively involve everyone in the review and, when
necessary, reformulation of basic plans, from the beginning. The value
of this consensus-building property of the virtual environment cannot be
overstated. The physical part of the VET is a large, multiscreen,
shared interactive virtual environment. Up to 20 persons can step into
the 270-degree FOV space defined by three 4.25 x 3.25-meter
rear-projection screens. Each screen is lit by intensely bright
Hughes-JVC 310 light-valve video projectors. In turn, each of the three
projectors is driven by a Digital AXP workstation paired with a Kubota
Denali graphics subsystem; a fourth Digital workstation synchs up the
network. The Worldspace), the virtual world running within the VET, is
constructed of models built in 3D Studio, a world framework provided by
Sense8's WorldToolKit, and Worldesign's own proprietary Reality Works)
software. For our installation, Visual Synthesis Inc. provided
spatialized sound and Kintek Stereo the sound-surround apparatus.
Funding for the VET was provided by ARPA, the U.S. Defense Department
research wing, and by Environmental Systems Research Institute Inc., the
leading GIS (geographic information systems) vendor. In the future,
Worldesign will bring the VET home to Seattle for installation at the
Port. After several months of upgrading the basic Worldspace - using
newly available CAD models for the buildings, rather than the crude
sketches in the current EIS - the VET will be ready to import analytical
algorithms governing traffic congestion, economic return of different
types of development, polltion and energy profiles of buildings, and so
forth. Over time it will become a highly useful system for design review
well-integrated with the way things are done today, but offering many
new opportunities for innovation and experimentation. Ultimately, VETs
will be networked to provide shared environments that transcend space
and time.
Dancing with the Virtual Dervish: Virtual Bodies
DIANE GROMALA
School of Communications, DS-40, The University of Washington Seattle,
Washington 98195 USA
with
YACOV SHARIR
Department of Theatre + Dance, The University of Texas
Austin, Texas 78712 USA
ABSTRACT
Dancing with the Virtual Dervish: Virtual Bodies is a collaborative
project in virtual reality (VR) by visual artist and designer Diane
Gromala, choreographer Yacov Sharir, and composer Russell Pinkston. It
was funded by a major grant from the Cultural Initiatives Program of the
Department of Communications Canada through a two-year residency at the
Banff Centre for the Arts in Canada. This project and the six others in
the Art and Virtual Environments Project, selected from among an
international pool of artists working in virtual reality, represented a
remarkable opportunity. The Banff Centre provided each of the
collaborative groups with a team of engineers, a highly technologically
literate art assistant, and access to a high-end Silicon Graphics
machine, head-mounted displays, data-gloves, trackers, and
state-of-the-art television, film, and sound studios. Through intensive
periods of research and testing, Dancing with the Virtual Dervish:
Virtual Bodies resulted in a dance performance where the dancer and
audience members perform and interact with a virtual environment in
real-time. Large-scale video projections of what each interactor
experienced created another level of VR in the performance space, and
turther encouraged participation. The opportunities and limitations of
the technology were embraced and pushed, resulting in new creative
strategies and directions for further technological development.
Paper Session: Architectures
A USER-DEFINED VIRTUAL ENVIRONMENT DIALOGUE ARCHITECTURE
Anthony Steed and Mel Slater
Department of Computer Science, Queen Mary and Westfield College
Mile End Road, London, E1 4NS, England
st...@dcs.qmw.ac.uk, m...@dcs.qmw.ac.uk
ABSTRACT
This paper describes a manipulable dialogue architecture that allows the
participant in a virtual environment to change how their gestures
affect the virtual environment. The dialogue structure has a visual
representation within the virtual environment that can be manipulated to
alter the gestures that are recognised, or to create new gestures. An
implementation is described and the benefits that this ability to change
the interaction gestures from within the environment are discussed.
DESIGNING IN VIRTUAL REALITY:
IMPLEMENTING PERCEPTION-ACTION COUPLING WITH AFFORDANCES
Gerda J. F. Smets, Pieter Jan Stappers, Kees Overbeeke
Faculty of Industrial Design Engineering
p.j.st...@io.tudelft.nl
and
Charles van der Mast
Faculty of Informatics, Delft University of Technology
Jaffallan 9, NL-2628BX Delft, The Netherlands
ABSTRACT
In this paper we describe a package we are developing for computer aided
design (CAD) in virtual reality (VR). We present both the version of tim
design package that has been realized and the final mode of interaction
that we are aiming for. We believe that designing in virtual reality has
some advantages over designing with an ordinary computer system. We
describe these advantages in terms of the ecological approach to
perception, focusing on three of the consequences of this approach. The
first advantage is that human-product interactions can be evaluated.
The second concerns the role of perception-action coupling in producing
true direct manipulation, allowing designers to interact with their
computer on a more intuitive behavioural level than by selecting
commands or pushing buttons. The third consequence is the desirability
of providing perceptual information about the affordances in the design
environment.
The Virtual Workbench: Dextrous VR
Timothy Poston and Luis Serra
Centre for Information-Enhanced Medicine
Institute of Systems Science
National University of Singapore, Heng Mui Keng Terrace, Singapore 0511
(t...@iss.nus. sg, lu...@iss.nus.sg)
ABSTRACT
Skilled work in a VR environment requires dexterity. Dexterity is more
easily achieved by embedding the virtual world in the user's natural
work volume than by immersing the user in a beyond-arm's-reach virtual
space. We describe a cost-effective means of achieving this, with easily
calibrated hand-eye coordination; a stereo display one looks and
reaches into. We discuss the hardware and software environments, the
calibration required, a test for the support of dexterity, and a
preliminary application in the manipulation of medical images.
Panel Session: Enabling Technologies for VR
Chair: Daniel Thalmann
Swiss Federal Institute of Technology, Switzerland
thal...@di.epfl.ch
Peter Astheimer, Fraunhofer-Institute for Computer Graphics (IGD) Germany
asth...@igd.fug.de
Luis Serra, National University of Singapore, Singapore
lu...@iss.nus.sg
Mel Slater, Queen Mary College, UK
m...@dcs.qmw.ac.uk
Ideally a virtual reality system and application enables humans to be
totally immersed in pseudo-real or imaginary virtual worlds. This is
achieved technically by stimulating all human senses (see, hear, touch,
taste, smell) and responding to human actions. An application can be
qualified as totally immersive when the user is not able to distinguish,
whether the world he temporarily inhabitates is virtual or real. Todays
software and hardware technology is far from a realization of such an
idealized system. The graphics part of VR-systems - the generation and
display of multiple frames per second - is well understood and suitable
hardware and software systems generally satisfy our needs. Audio
hardware and device controlling software is also available now
(multimedia- workstations, Midi-equipment, signal processor based
computing units, special devices), although mostly not directly designed
and applicable for VR purposes. Force feedback subsystems and tactile
displays are slowly being developed and customized. The stimulation of
the olfactory and gustatory senses with program controllable
subsystems is still subject to experimentation. With audio hardware and
system software readily available it is now possible to take the next
evolutionary step and use acoustic simulations to enhance virtual
worlds. Acoustic simulation comprises the generation of the sound
itself, the calculation of sound propagation in an environment and the
subsequent auralization and spatialization of the computed acoustic
parameters. Up to now not all effects experienced in our real world can
be simulated in realtime like the computation of sound propagation and
subsequent 3d auralization in arbitrary environments. For virtual
reality applications powerful workstations or signal processing hardware
is needed. An audiovisual system addresses two important human senses at
the same time and increases the level of realism towards an ideal
immersive system significantly. The incorporation of sound is a first
step towards a complete perception system which addresses all human
senses. People communicating in a shared virtual environments may be
one of the most important applications of Virtual Reality. Using full
teleconferencing is expensive, and only suitable where the meeting
participants do not need to share the same virtual space. For example,
ISDN rates, even with coding, are relatively slow and it is expensive to
produce full colour moving images at 30 frames per second. An
alternative may be to transmit only the (changes to) geometry which are
elaborated by independent programs at the network nodes. This will
enable people to have face-to-face meetings, carry out tasks jointly,
and exchange information. It requires not only sufficient bandwidth, but
suitable and active representations of the participants in such
virtual meetings. Such representations should display not just overall
body posture, but also ideally information about facial expression.
Enabling technology required for virtual meetings therefore includes
both rapid internetworking with suitable protocols for the transfer of
geometric data, but also sensors that enable the measurement of
behavioural and emotional state. For an interaction between a virtual
human and a real one, there is no possibility of transferring data
structures, and image understanding methods are required to provide the
virtual human with a perception of the real human's behaviour. True
interaction between the virtual and the real humans requires a two-way
communication between them at the geometric level, at the physical
level, and at the behavioral level. At the geometric level, 3D devices
like a DataGlove allow the real human to communicate any geometric
information to the virtual one. At the physical level, using a force
transducer, a force or a torque may be communicated to a virtual human
who can apply a force that may be felt by the real human using a force
feedback device. It is for example possible to simulate the VR scene
where the animator and the virtual human tug on the two ends of a rope.
At the behavioral level, we consider emotional communication between the
virtual human and the real one, for this kind of emotional exchange
mainly based on recognition of facial expressions, only video input
seems appropriate. Finally, for the believability of many VR
applications, there is a need for autonomous virtual people, reacting
to environments and making decisions based on perception systems, memory
and reasoning. A pair of virtual humans interacting is a closed system
which can be developed by equipping the virtual humans with
complementary behaviours. In order to support interaction and
communication, virtual humans should be equipped with the ability to
'recognize' other virtual humans and 'perceive' their facial
expressions, gestures and postures. In this case, there is no need for
real recognition or perception, of course, because information from the
data structures that define these behaviours in one virtual human can be
passed directly to a second virtual human.
Invited Talk:
Virtual Reality Teclmologies and its Applications to Industrial Use
Junji Nomura
Virtual Reality R&D Group, Information System Center
Matsushita Electric Works, Ltd., 1048, Kadoma, Osaka 571, Japan
Email: NOM...@AI.MEW.CO.JP
ABSTRACT
Virtual reality, a new paradigm for relationship between humans and
computers has been recently well-known and currently investigated for
practical use in the various industrial fields. Using three-dimensional
computer graphics, interactive devices, and high-resolution display, a
virtual world can be realized in which one can pick up imaginary objects
as if they were physical world. Using this technology, Matsushita
Electric Works, Ltd. has been developing several application systems
Jor industrial use since 1990. This paper details three VR application
systems operating in the real world: Virtual Space Decision Support
System employing Kansei Engineering which is applied for production and
sales mainly in the system kitchen business, a telepresence robot system
employing semi-autonomous mobile function which is utilized for
security field and a low-cost VR system employing physiological
feedback mechanism which is used for health care field.
Paper Session: Novel 3D Techniques & Devices
MULTISENSORY SCIENTlFIC DATA SENSUALIZATION THROUGH VR TECHNOLOGY
TETSURO OGI
Advanced Science Department, Mitsubishi Research Institute
2-3-6, Otemachi, Chiyoda-ku, Tokyo 100, Japan
Email: te...@mri.co.jp
and
MICHITAKA HIROSE
Faculty of Engineering, University of Tokyo
7-3-1, Hongo, Bunkyo-ku, Tokyo 113, Japan
Email: HIR...@IHL.T.U-TOKYO.AC.JP
ABSTRACT
Rapid advances in the computational ability of scientific computing
systems has given rise to the existence of vast amounts of data.
However, understanding complex phenomena represented by data through
conventional data visualization methods alone utilizes only the user's
visual sensations. Hence, in this study, virtual reality technology was
applied to develop a multisensory data sensualization method using
visual, acoustic, and touch sensations to display scientific data. In
particular, a prototype wind sensation display system was developed to
generate touch sensations for scientific data. Scientific data was
represented in various forms of physical stimuli such as color, sound
frequency, and wind magnitude. In multisensory data sensualization,
accurate and efficient transmission of data from the computer to the
user was achieved through the effective integration of several
sensations. In this study, we conducted several experiments on the
perception of scaler data and vector data in three dimensional space
using multisensory data sensualization methods and verified their
effectiveness in giving the user accurate perceptions of data. Results
indicated that multisensory data sensualization methods can be applied
to effectively represent and display scientific data in several
engineering fields. In this paper, applications of data sensualization
to structural dynamics and fluid dynamics were also illustrated.
AN INERTIAL HEAD-ORIENTATION TRACKER WITH AUTOMATIC DRIFT COMPENSATION
FOR USE WITH HMD'S
ERIC FOXLIN, NATHANIEL DURLACH
Research Laboratory of Electronics, Massachusetts Institute of Technology
Cambridge, MA 02139, USA
emf...@irts.mit.edu
ABSTRACT
Current head-tracking technologies suffer from limitations such as
delay, limited range, vulnerability to interference, line-of-sight
requirements and high cost. In principle, the methods of Inertial
Navigation Systems (INS), applied to head-tracking, could overcome these
problems. However, inertial head-tracking has been largely neglected due
to the difficulty of making a small, light INS that does not drift too
much. In order to evaluate the suitability of inertial sensors for use
in virtual environment and teleoperator head-tracking applications, an
inertial head-orientation tracker has been built and bench-tested for
accuracy, resolution, noise, and latency. Yaw, pitch and roll of the
head are computed by Euler integration of the outputs of three
orthogonal angular rate sensors. Drift compensation is accomplished by
making use of natural pauses in head motion to obtain stable readings
from a two-axis fluid inclinometer and a fluxgate compass. The system
achieves 0.1 ms lag, 0.008 degree angular resolution, and an
unrestricted working volume. The pitch and roll axes, using a fluid
inclinometer for drift compensation, achieve 1degree angular accuracy.
The drift compensation of the yaw axis using a compass has not yet been
evaluated. The results indicate excellent potential for the use of
inertial technology in head-tracking, and work is under way to extend
the system to 6 degrees of freedom.
A NEW ULTRASONIC POSITIONING PRINCIPLE YIELDING PSEUDO-ABSOLUTE LOCATION
HUGH L. APPLEWHITE
Piltdown Inc, 4470 SW Hall Boulevard, M/S Beaverton, OR 97005 USA
hu...@teleport.com
ABSTRACT
Modulated Phase Coherence (MPC) is a new VR positioning principle. The
advantages of MPC include pseudo-absolute position determination and
direct measurement of velocity. MPC uses temporal or spatial
modulation of a carrier wave to remedy one of the primary drawbacks of
Phase Coherent (PC) systems, the lack of absolute position
determination. Basic MPC positioning principles are described along
with Doppler effect compensation. Three implementation approaches are
analyzed. The positive features of the PC approach, high data rates over
large volumes, low lag, and good environmental noise immunity, are
retained. The intended audience is knowledgable in systems and
software, but not necessarily in signal processing or communications.
Emphasis is on the interaction of geometry and signal properties.
A STUDY ON THE SYNTHESIS OF ENVIRONMENTAL SOUNDS
MICHITAKA HIROSE, SHIN'YA KOMORI
Department of Mechano-Informatics, Faculty of Engineering, University of Tokyo
7-3-1, Hongo, Bunkyo-ku, Tokyo 113, Japan
Email: HIR...@IHL.T.U-TOKYO.AC.JP
and
TOSHIKI NAGUMO
Computer and Communication Research Center, Tokyo Electric Power Co., Inc.
1-4-10, Irifune, Chuo-ku, Tokyo 104, Japan
ABSTRACT
HRTF(Head Related Transfer Function)-based acoustic display systems have
been studied in parallel with the development of virtual reality
technology. However, their initial intended function was only to display
the point sound sources to the user in three-dimensional virtual space.
Hence, this provided motivation to design and develop a new sound field
display system, which could reproduce complex and widespread
Environmental Sound Fields (ESF) with a high degree of realism. To
develop this system, the understanding of the behaviour of sound waves
is crucial. Thus, an ESF display system along with the application of
sound and noise principles was developed and tested. The simulation
results were used to generate theoretical diagrams of the sound waves
generated by point sound sources. Secondly, an experimental prototype
speaker system was developed and accurately tested in terms of auditory
localization. Furthermore, the ESF display system was implemented into
two virtual reality devices. Creation of these graphs and experimental
results provided better understanding into the behaviour of the ESF
display system and the subjects with respect to integration with
conventional virtual reality devices.
Invited Talk:
Virtual Environments in Scientific Visualization
STEVE BRYSON
Computer Sciences Coporation, Numerical Aerodynamic Simulation Systems Division
NASA Ames Research Center, Moffett Field, Ca.
bry...@nas.nasa.gov
ABSTRACT
The use of virtual environment (virtual reality) techniques for
scientific visualization are surveyed. Lessons learned from vanous
existing applications are stated, with regard to implementation areas
such as computation, graphics, data management, and user interface.
The fruitful interplay of virtual reality and scientific visualization
is stressed.
Paper Session: Designing Effective Virtual Worlds
MANAGING MUTUAL AWARENESS IN COLLABORATIVE VIRTUAL ENVIRONMENTS
STEVE BENFORD
Department of Computer Science, The University of Nottingham,
Nottingham, NG7 2RD, UK
JOHN BOWERS
Department of Psychology, The University of Manchester, Manchester, UK
LENNART E. FAHLEN
The Swedish Institute of Computer Science (SICS), S-16428
Kista-Stockholm, Sweden
l...@sics.se
CHRIS GREENHALGH
Department of Computer Science, The University of Nottingham,
Nottingham, NG7 2RD, UK
ABSTRACT
This paper introduces a spatial model of interaction which aims to
support groups of people in using their natural communication skills in
distributed virtual environments. First, we outline our motivation for
undertaking this work in terms of the social significance of space in
supporting co-operative work. Next, we summarise the model's main
concepts, namely aura, awareness, focus, nimbus, adapters and
boundaries. Following this, we describe routes to implementing the
model. The first is an approach suited to realising the model as an
application of existing VR platforms. The second is a general
implementanon of the model, based on its own specially tailored
distributed architecture. Finally, we present our initial observations
from use in the laboratory setting and outline issues for future work.
REAL TIME GRAPHICS AND VIRTUAL REALITY, FOR DRIVING SIMULATION IN URBAN
ENVIRONMENTS
SALVADOR BAYARRI, MARCOS FERNANDEZ, MARIANO PEREZ, FRANCISCO ROSICH
LISITT, University of Valencia, Hugo de Moncada 4, Valencia 46010, Spain
{bayarri, marcos, mpm, rosich}@glup.eleinf.uv.es
ABSTRACT
The rising interest in the development of medium-size driving simulators
with high performance level is leading to the research in Virtual
Reality techniques incorporating this technology. The spatial and
temporal complexity of the urban environment has not been broached by
existing driving simulators, and it offers a good test field to the
behavioral simulation algorithms, real time graphics management and VR
interaction. This paper presents a method to manage an urban
environment complex database, making use of hierarchical elision
techniques over a spatial connectivity network. It is also shown how to
adapt and refine traffic microsimulation algorithms to provide visual
quality in a local area (SMAL algorithm) and it is exposed the
connection performed between SMAL and a macroscopic qualitative
simulator (QS). Different techniques to achieve a real time 3D rear
mirror view capability are explained.
ENVISIONMENTS-CONSTRUCTING DRAMATIC VIRTUAL WORLDS
MAGGIE COOPER, IVOR BENJAMIN
Centre for HCI Design, Dept. of Business Computing, City University,
Northampton Square, London ECIV OHB, United Kingdom
m.co...@city.ac.uk
ABSTRACT
If virtual worlds are to fulfil their promise as stages for
entertainment, education and artistic endeavourr, with who or what shall
we people them? And what methods are available to us to design the
dramatic elements of these worlds? Both drama and virtual realities are
multi-modal experiences. By drawing on crossover research between the
disciplines of theatre, HCI and system design, we put forward an
innovarive basic framework for the design of virtual worlds - an
envisionment. This paper will examine the nature of interactive
experience in relation to user viewpoint and identity, and will develop
a frame of reference for the structure and composition of virtual worlds
in dramatic terms: location and environment, content, actions,
time-scale, events, metaphysical structures and characters. We shall
draw attention to the significance of constraints and potentials to
envisionment and to techniques drawn from object-oriented analysis and
design to model them. We hope that this paper will stimulate a
discussion of the issues raised and future directions for research.
Keywords: Virtual Reality, Virtual Worlds, Drama,
Human-Computer-Interaction, System Design.
Panel Session: Hands Off My VR: The Role of Gestures in VR
Chair: Mark Green
University of Alberta, Canada
Steve Bryson, NASA Ames, USA
Timothy Poston, National University of Singapore, Singapore
Alan Wexelblat, MIT Media Lab, USA
1. Introduction
What is the role of hands in VR? We use our hands to interact with
objects in the real-world, suggesting that hands are the natural input
device for VR. But, is this the case? Most input devices require the use
of at least one hand, the main issue is how the hand is used to interact
with the device. In the case of a glove, the input device attempts to
mimic the actions of the user's hand, so the user can use his or her
hand in a natural way. For the purpose of this panel session we will
call this gesture based input. Other types of input devices, such as
buttons mounted on a 3D tracker, can be viewed as tools that the user
manipulates with his or her hand. In this case the input device doesn't
attempt to mimic a hand. For the purposes of this panel session we will
call this tool based input. Now comes the question, which is best,
gesture based or tool based input? There have been no detailed studies
done on either style of interaction and we still have very limited
experience with the use of these devices, thus this is an excellent
topic for a panel session.
Paper Session: Collision Detection
Efficient Virtual Collision Detection for Multiple Users in Large
Virtual Spaces
KIM MICHAEL FAIRCHILD and TIMOTHY POSTON
Institute of Systems Science, National University of Singapore
Heng Mui Keng Terrace, Kent Ridge, Singapore 0511
{fair, tim}@iss.nus.sg
WILLIAM BRICKEN
Human Interface Technology Laboratory, University of Washington, FJ-15
Seattle, WA USA 98195
wil...@hitl.washington.edu
"Science Fiction writes about technology that is still in the Goofy
Prototype stage, this is why Virtual Reality is so popular now."
Bruce Sterling, Armadillo Con, 1990, Austin, Texas.
Abstract
Virtual reality researchers are developing technologies to deal with
current device limitations. But even with the accompanying advances in
raw computer power, VR application designers will always want just a bit
more power to make their virtual objects even more convincing. What we
need, along with the device advances, is software advances that 'solve'
the computationally expensive problems. One is efficiently determining
if two objects are touching, which is known as the collision detection
problem. A prototype system has been built that provides support for
collision detection in real-world applications. The Virtual Collision
detection algorithm scales up to handle tens of thousands of virtual
objects and uses order n memory and CPU resources. It supports teams of
users in worlds containing objects accessible at various resolutions.
The algorithm can also be directly extended to solve other problems such
as view management.
This paper describes the algorithms from a conceptual and implementation
standpoint. Simulation results are then presented.
Virtual and Real Object Collisions in a Merged Environment
DANIEL G. ALIAGA
Department of Computer Science, University of North Carolina at Chapel
Hill Chapel Hill, NC 27599, USA
ABSTRACT
See-through head-mounted display capability is becoming an important
part of Virtual Environment applications. In such applications, it may
be desirable to model the physical behavior of the virtual objects and
their interaction with the real objects. This paper describes a software
system which integrates interactive collision detection, collision
response and see-through headmounted displays. The system employs a
static model of the real world environment and allows for arbitrary
convex virtual objects to be placed in the environment. The user may
control the positions and velocities of the virtual objects. An
approximately constant time collision detection algorithm and a
Newtonian Mechanics based single point contact collision response is
used to model the apparent physical interaction of the virtual and real
objects for moderately complex environments.
Keywords: Virtual Reality, See-through Head-Mounted Displays, Collision
Detection, Dynamics, Parallel.
LEVEL-OF-DETAIL GENERATION AND ITS APPLICATION IN VIRTUAL REALITY
PETER ASTHEIMER, MARIA-LUISE POCHE
Fraunhofer-Institute for Computer Graphics (IGD)
Wilhelminenstr. 7, 64283 Darmstadt, Germany
email: asth...@igd.fhg.de
ABSTRACT
Virtual worlds have to be presented in realtime in order to enable
interactive control, viewing and object manipulations. The performance
of graphics workstations is limited in the maximum number of polygons
which can be processed in realtime. Complex world models (which are of
prime interest) exceed this limit easily. A set of rendering techniques
allow to handle and conquer complex worlds, where level-of-detail
techniques prove to be most promising and successful. This paper
introduces an interactive tool with a set of methods for the generation
of multiple levels-of-detail of objects, demonstrates several examples
and the application within IGD's proprietary VRSystem "Virtual
Design".
------------------------------END OF REPORT------------------------------