Content uploaded by Erik Geslin
Author content
All content in this area was uploaded by Erik Geslin on Jul 05, 2018
Content may be subject to copyright.
WiiMedia: motion analysis methods and applications using a consumer video
game controller
Akihiko Shirai∗
ENSAM Presence & Innovation / JSPS
Erik Geslin
ESCIN Laval
Simon Richir
ENSAM Presence & Innovation
Abstract
”WiiMedia” is a study using the WiiRemote, a new consumer video
game controller from Nintendo’s, for media art, pedagogical appli-
cations, scientific research and innovative unprecedented entertain-
ment systems. Normally, consumer hardwares, like standard con-
trollers of new video game platforms, are closed to public develop-
ers. The Nintendo’s WiiRemote however can be connected easily
to an ordinary PC thanks to a BlueTooth adapter. Thus, public de-
velopers can access to the WiiRemote’s acceleration and IR sensors
via this wireless connection. We think it might enlarge the non-
professional game development environment with a new innovative
game controller. However, when we tried to develop our projects
with the WiiRemote, we encountered many difficulties because the
only data that can be captured are basic data and not the full player’s
motion. Through the WiiMedia project, with the development of a
few applications, we developed some motion analysis methods us-
ing the WiiRemote. This paper describes case studies that include
states of the arts and several motion analysis methods.
CR Categories: H.5.2 [Information Interfaces and Presentation]:
Interfaces—Input devices and strategies
Keywords: entertainment, video game, WiiRemote, media study
1 Introduction
Innovation and reform of game platforms occur every five years.
The three new consumer game platforms, Microsoft Xbox360,
Sony Playstation3 and Nintendo Wii have been released at the end
of 2006. In “the last warring state period” (2001), the main topic
was computer graphics performances. In this warring state period,
the main topic seems to be network services and human interfaces
of game controllers. Especially Nintendo Wii proposed a new at-
tractive idea the “WiiRemote” (a.k.a Wiimote or RVL-003), that
symbolizes a game play revolution.
From Nintendofs expectations, the new controller will attract “lost
markets” or “new users” taking woman and elderly into the new
product designs. Moreover some featured titles with motion inter-
faces, like “WiiSports”, may create some new market in the e-sports
field.
However, entertainment systems have a destiny that will be tired
by monotonicity. To prevent it, the innovative user interface will
need to give a new amazing experience to the consumer through
new titles to continue to offer a new entertainment.
∗e-mail: shirai@mail.com
2 Concepts of “WiiMedia”
Wii will be “a mechanical emulator platform” that will support old
brilliant games, if Wii were not to release more brilliant expec-
tations and experiences than the video preview of GDC2005 and
launch titles. They must create the opportunity for the WiiRemote
to have an impact through the game titles by producers continually.
Concretely, if game designers would like to give badly-developed
impression to the players they can use the WiiRemote as “an input
device for timing” like...
•A substitute of buttons
•Just swinging actions
•Violent actions that may give a heavy load
•Highly difficult timing input
These “badly-developed utilization” will give an impression that
“Skillful players always can win”, a similar formula to the one
which has been used for the past 20 years. As a result, “lost users”
will be tired of this formula for the new games to come.
The important points for new users are “Do not need to get a win”
and also “Don’t play a game infinitely”. Probably, the objective of
game for them may be “enjoy a better life with getting experiences
using computers”. Now they just go out of the their way to play
with a new controller.
The main motivation for them is “get the new experiences”. Fol-
lowing this hypothesis, if game developers would like to attract
the users more effectively, they should focus on exploring “artis-
tic experiences” or “law of nature sciences” instead of “violence
and killing worlds”. Because current computer games are repre-
senting too lopsided worlds from the human life, people who have
interests for art and sciences cannot find satisfaction in games. For
such users, a badly-developed interaction is just a badly-thought
toy. “Creating new games by themselves” will be an easier way to
get a new experience than playing with this one.
Now we named this different computer entertainment activity as
”WiiMedia”. It views games as a media that can realize cre-
ative ideas, opening and suggesting new technologies to encour-
age to create various applications, allowing non-profit developers
like Sunday painters or authors to each realize “wellness”. In other
words, it brings us closer to media-arts and conventional games for
future entertainment.
It may be hard for the current game industry to accept the thought
of ”WiiMedia”. However, there are some common techniques, ap-
proaches and essences that are possible to apply to video games
directly. In fact, there are some cases in which media-arts affected
the video games industry.
In this article, we report some case studies of ”WiiMedia” with
some concrete applications as a first report. Concerning the tech-
nology, we describe usage, electronic circuit, motion analysis meth-
ods and contents development environment of the WiiRemote for
non-profit developers.
3 WiiMedia:Lead-up
3.1 Using WiiRemote on PC
After establishing a connection with the PC, we can test the WiiRe-
mote manipulation with some developed applications. The ”Wiin-
Remote” software is already supporting buttons, acceleration and
IR sensors being thus already enough to use the WiiRemote as a
presentation pointer [WiinRemote ]. ”GlovePIE” is an emulation
software for VR devices. It has a powerful script environment that
can describe emulation behaviors for keyboard, mouse, joystick and
MIDI [GlovePIE ]. It is useful if the user does not require other
programming. In the other case, there are some code examples and
applications running on MacOSX and Linux, based on the work
gathered on a website ”WiiLi.org”, which also has a big forum for
sharing such knowledges [WiiLi.ORG ].
Figure 1: Bluetooth connection of WiiRemote on DELL M2010
3.2 Possibilities and Limitations of Development
At programming level, full function meaning vibration, LEDs,
speaker and external port (including Nunchuk controller and inter-
nal memory) are possible to access through the HID driver. There
are no SDKs but a lot of hackers are trying to reveal WiiRemote
hardware making it more transparent than before. After release of
some utilities, the number of developers are increasing day by day.
The only worry is the gray legal zone in which such actions take
place. Nintendo might block the BlueTooth connection if someone
makes a lot of cheating methods to the game by using the same
method. However, main objectives of WiiMedia are not cheating.
The objective of cheating is “get a win easier” in stead of “create
new applications with the new controller”.
3.3 WiiMedia:“USB Sensorbar”
As the first step of the WiiMedia study, we developed a ”USB Sen-
sorbar” to separate ourselves from the proprietary video game con-
sole,the Wii. Originally, WiiRemote detects the ”Sensor Bar’” un-
der the TV screen to define an absolute position. Nintendo’s offi-
cial ”Sensor Bar” seems to be a sensing device but in fact it is a set
of Infrared LEDs whose wavelengths are around 900 nm, without
modulation. If we can take the power source (DC5V) to illuminate
IR-LEDs from an USB port, this will be sufficient for the WiiMe-
dia project and cost effective, instead of in the Wii console. Fig.2
shows an example of our USB Sensorbar.
3.4 WiiMedia:API
To access the WiiRemote easily, we developed WiiMedia API func-
tions. They are categorized as basic and utility functions. Basic
functions provide Open/Close to HID (Human Interface Device) of
Figure 2: “USB Sensorbar”
Bluetooth, thread management and some basic data handling func-
tions. Utility functions provide higher layer functions like data log-
ger, speculated gravity direction, posture of the WiiRemote, fre-
quency of capture, a motion classifier and some experimental eval-
uation functions. All codes are written in class oriented C language
for compatibility.
Figure 3: WiiMedia APIs
3.5 WiiMedia:Racing “AceSpeeder2Wii”
We tried to implement WiiMedia APIs in a racing game as a
first experimental application. AceSpeeder2 (AS2) is a typical
”Anti-Gravity” Sci-Fi racing game product running on PC plat-
form[Takahiro NAKATANI ]. We already used AS2 for scientific
research using new motion interfaces with computer vision tech-
niques like GPUVision and OpenCV[Shirai et al. 2006].
Figure 4: “AceSpeeder2”
3.6 Motion control for driving games
The original control of AS2 is very simple. Left and right handling
with a digital input and acceleration with a button. When a player
double clicks on the acceleration button it generates a boost.
We tried to realize a robust design for racing games without any
”button operation” thanks to the WiiRemote. Players would grab
the WiiRemote’s cross key by the right side of the controller, then
slant WiiRemote left and right to input the handling. To accelerate,
a player moves the WiiRemote to his chest. When the player moves
it as far away as possible, it brakes. When the WiiRemote is set
with its left side on a table, it makes the game to pause.
While in the game, all functions are implemented by WiiMedia
APIs. Especially for the boosting, we used the Magnitude func-
tion of our APIs to shake the WiiRemote. We also realized some
hardware effects with our APIs like showing left shield power with
WiiRemote’s LEDs and feedback with vibrators when the player
vehicle is getting damaged.
Figure 5: Controlling of WiiMedia:Racing “AceSpeeder2Wii”
3.7 Conflicting with game Physics
We made a public examination on Laval Virtual 2007, with hun-
dreds of people testing our controlling method. As a result, we
observed most players can play the game without detailed expla-
nations. We think the method is a good representation for Anti-
Gravity racing with precise information instead of digital pad or
joysticks.
However, this characteristic revealed a new problem. It is conflict-
ing with the game physics. For example, in the real world, car
drivers may not control their cars to make rounding left or right
turns the same way because when they try to turn the wheel hard
and suddenly, on the real car in real physics, they feel the physi-
cal limitations of the real world through their hands. However, in
our examination, there are some players who swung the WiiRemote
rapidly or turned it over 90 degrees.
A better sensibility of the WiiRemote in the application program
seems to be a solution to the problem, but it is not that simple. In
fact, the sensibility directly affects the game difficulties,the physical
parameters of each vehicles and the base modeling of game physics.
It may be a problem for the new generation of video game con-
soles using wireless motion controllers like WiiRemote. We think
it should be corrected with feed-backs of force, vibration and/or
sounds, to tell the difference in the player’s input and game physics.
4 WiiMedia:“Papier Poupee Painter”
4.1 Concept
We describe a drawing software, Papier-Poupee-Painter (PPP) as
our second case study. Development of drawing softwares like Kid-
Pix and Tux Paint are very interesting. Drawing, a typical creative
activity, is possible to do with traditional pigments. However if
we can create a drawing software which can provide a computer
augmented experiment and creativity, there is a significance that is
only realized with computer interactions. Also if an art media can
be composed of WiiRemote, a new video game controller, with at-
tractive pleasure representation, it may generate a new game genre
and increase the computer game majority.
However, there are quite a lot of requirements when focusing on
“for kids” applications. There are some previous works with com-
puter drawing creation in video games. Igarashi and his 3D graffiti
engine, “Teddy” is applied to “Graffiti Kingdom / Magic Pengel”
(TAITO, 2001) and “Amazing Island” (SEGA, 2004) to generate
player’s favorite creatures from their own drawings. It is an good
idea, but the analog pads needed for precise manipulation for draw-
ings made it quite difficult for smaller children.
In the examples on PC platforms, there are some human-interface
techniques that limit the functions for children to for example sim-
ple tools like a stamp tool or thick pens. However, the method
may form “careless painter kids” in the future. Because when we
use pastel or crayon on the actual paper, we need to manage draw-
ing force, a direction and adhesion of pigments instead of a mouse
click.
This project, WiiMedia:PPP supposed infancy users, an age group
of 3 to 5 years old. This generation can understand an object on the
screen in an interactive system as “something changeable by me”
but sometime they cannot understand the operation of mouse and
GUI pallets (varies between individuals). They mainly draw sym-
bols, coloring, numbers, alphabets and primitive shapes like vehi-
cles without scenography. Also the treatment of paper and pigments
are important but they often play with pigments when the adults do
not watch them.
4.2 Papier Poupee Painter
We propose the “Papier Poupee Painter”, art media drawing soft-
ware which can arouse children’s imaginations. The painting ma-
terial should be robust, enduring ,attractive and playful for children
as a interaction, but one can realize new interactions instead of us-
ing a mouse. Concretely, the painting material using WiiRemote
was dressed with “Poupee” (french for doll for children in a secu-
rity blanket). The drawing system has a realistic drawing stroke
simulator.
Figure 6: Drawing material of Papier Poupee Painter.
4.3 Realistic drawing stroke simulator
Murakami had developed a realistic drawing software that can
simulate paper texture with strokes of several pigments in real-
time[Kyoko Murakami 2006]. This technique originally came
from Non Photo-realistic Rendering (NPR) but it is a very realistic
method to represent an importance of paper and pigments utiliza-
tion. The method obtains a height map of each paper media from
12 different illuminated photographs. On the real-time stroking en-
gine, it selects which height map is used according to a drawing
direction, and pigments are deposited on the height map. It also
supports tilt, azimuth, pressure and position of brush from a high
end tablet. In the PPP project, we apply this technique to obtain the
position and pressure of player with WiiRemote’s acceleration and
IR sensors.
4.4 Position detection limitations of WiiRemote
The WiiRemote device has a two dimensional Infrared (IR) position
sensor. However, it has three issues to create drawing software.
First, the output coordinate of IR sensor is not an absolute 3D co-
ordinates. Normally, IR sensor of WiiRemote output relative 2D
coordinates (IrX, IrY). It means the center of gravity of IR intensity
instead of pixel image matrix. Ideally, it should be a linear mapping
like a laser pointers but it has a distortion. It also supports to detect
1-4 IR points (2 points with a stability). Thus we can estimate depth
(PZ) with geometry and distance of 2 groups of IR LEDs.
The second issue is a limitation of area by view of IR sensor. The IR
sensor of the WiiRemote is about 36 degrees by our measurement.
For instance, if we align WiiRemote to the two IR-LEDs directly,
there is a safe area that can detect 2 LEDs safely (Fig.8). On the
plane of Z=1200mm, the width of range is only 800mm (A 4 years
old child can take 900mm width for left to right hand in maximum).
Making it a big issue that should be resolved.
Lastly, there seems to be a specific issue of WiiMedia:PPP. It is
difficult to make an action of “drawing on the screen” with WiiRe-
mote as a pen. The reason is a location between WiiRemote and
USB sensor bar. Thus we solve this issue by installing IR-LEDs on
a ceiling or a floor; or reduce the distance between two IR-LEDs.
4.5 Position analysis with correction
Fig.8 shows the value of IR coordinates (IrX, IrY) change widely
with the depth (PosZ). For example, if the WiiRemote outputs “IrX”
= 768, the physical position (PosX, PosZ) might be (0,660), (-
330,1000), (-330,2310), (-660,3000). Also there are no guaranties
that the WiiRemote has no rotation on the Y-axis. In fact, IR sen-
Figure 7: Appearance of paper texture for real pastel strokes from
different directions (top). Murakami’s pigment adhesion model
with height map from a light source for a paper (middle). The ef-
fects of a pressure interpolation (bottom).
Figure 8: Measurement of WiiRemote’s Infrared coordinates on
Y=0 plane. A data line means same set of PosX. Unmarked points
mean the outside of sensor’s view range.
sors can possibly be used to obtain relative coordinates instead of
an absolute position directly. This may limit the interaction using
WiiRemote with the IR sensor.
Figure 9: Calibration for IR coordinates to absolute coordinates.
We tried to obtain more precise absolute positions with IR coor-
dinates (IrX1, IrY1, IrX2, IrY2) and known distances between a
couple of IR-LED groups (d=200mm). Fig.9 Left shows the dif-
ference of each IrX (IrZ = |IrX 1−IrX 1|). Then it shows a same
curve in several PosZ plane. It also shows an inverse proposition
(Z=k/X). To obtain the factor k, substitute a static value, 200mm as
a distance between LEDs (Fig.9 Right). Then we can say the factor
k is a linear function of PosZ. Finally, we can get the actual position
PosZ from the differences of IrX1 and IrX2 (Fig.10 Right).
Figure 10: Calibration result for depth.
The detected depth is used as the pressure on the pen. The stroking
engine supports other inputs like rotation and direction of pen
which should be tried in the future. To test the advantage of Wi-
iMedia:PPP, we have shown it on Laval Virtual 2007 to the general
public. To evaluate our several detection methods, we sometimes
switched the input modes as below.
1. Acceleration sensors only
2. Normal usage of WiiRemote
3. IR sensor facing to down
We watched how many seconds the players played with the meth-
ods. Normally if a method is attractive, the playing time will be
longer than for the other methods.
During the evaluation, infancy girls were more curious about it than
boys. However, some infancy children were scared of the new in-
teractive system. In the expo place, it was dark, making it a little
difficult to show a tangible toy for drawing.
As the result of the evaluation for input methods, (1) the accelera-
tion sensor was the most frequently played a long time. The reason
is simple and easy. Most players, including adults, normally do
not have any object to draw something. So to swing and see the
behavior of pigment is a sort of good play (Fig.11). It might be a
different result if we were to use an other condition like back pro-
jection screen or plasma display as tabletop. Also if the system has
some other tasks and interactions like a coloring, sound feedback
or narrative, the usage will be extend.
Figure 11: A result of an infancy child on “Papier Poupee Painter”
To conclude this chapter, we note another idea to enlarge the de-
tection area of IR sensor (Fig.12. It is a polar coordinate method
using gravity detection by acceleration sensors. This method is not
limited in the direct position, so the movement area is larger than
woth a normal usage of the WiiRemote. In fact, this method is spe-
cialized for WiiMedia:PPP with method (3). Because there is no
motion sensor axis on the WiiRemote that can detect this rotation
in the normal usage.
Figure 12: An new idea of new position detection with Infrared and
gravity detection acceleration sensors (right).
5 WiiMedia: JaWii’s Virtual Fencing
This chapter describes a case study of a sword fighting game with
Wiimedia’s acceleration and analysis studies.
Alfred JARRY (1873-1907) is a famous dramatist from Laval city
in France. He was also famous for his classic sword play. In 2007,
Laval city had an idea to show a creation of Alfread JARRY with
virtual reality techniques to mark the centennial of his death, result-
ing in us working with them to realize it.
The idea was to create a small game system to explain the classic
sword play to the public. In the beginning, we thought about using
a magnetic motion capture device for VR. But even if it is a precise
device, it is also fragile and expensive. Especially in past works
Figure 13: Alfred JARRY and his classic sword play
of our interactive systems, we had a sword play game system us-
ing a magnetic sensors but it limits players’ activity and also looks
breakable. Through such experiences, we decided to apply the Wi-
iRemote to this project. It would be very cost effective, attractive
for the public and lower the potentiality of breaking.
Figure 14: System diagram of “JaWii’s Virtual Fencing”
The game system is very simple. When a player swings the WiiRe-
mote, the player character can attack JaWii, an imitated character
of Alfred JARRY. The goal of the play is not to be defeated by him.
JaWii shows a lot of classic sword play motions during the play.
The animation is from a professional fencing player with Metamo-
tion’s Gypsy mechanical motion capture (Fig.16).
5.1 Motion capturing and analysis for sword play
To develop JaWii’s system, we had to compare several methods to
detect players’ motions. Finally there are only 3 motions which are
detected now, however in the beginning, there were other actions
like height differences allowing several heights of engagement.
The main priority of the analysis method is its response time and
focusing to the magnitude of acceleration. We compared several
algorithms to get better results. The best way for sword play is
creating an evaluation function to watch the magnitude of WiiRe-
mote variables and detect a representing target vector of a targeted
motion.
However, it is very hard work to define the target vectors for each
target motion. Also the output of WiiRemote’s acceleration has a
defect. Fig17 is an attacking motion in 3D vector graph with a
Figure 15: Screen shot of“JaWii’s Virtual Fencing”
Figure 16: Collaboration with professional fencing player on clas-
sic fencing motion archiving
Figure 17: Reconstructure of 3D position with acceleration sensors.
Blue arrows (lower acceleration in the end of one swing) are lost
parts of detection.
recorded acceleration data and estimated position by integral of ac-
celeration. In the beginning of the motion, the behavior is sharpen
but in the end of each swinging motion, it shows a lack part colored
by blue. We guess the force in this part is under the minimum de-
tection of WiiRemote or a noise cutting filter for gravity canceling
is cut from the value.
Also there are other issues for detection. Sometimes “the standard
handling” of WiiRemote is different for each player. Then if we
define a strict vector as a target, it may be difficult for other play-
ers. For example, an acceleration path of a professional sword-man
play player is very sharp, instead of smooth curve found with nor-
mal players. If the WiiRemote twists in the motion, the vector also
twists. That is also a reason why we should focus to see the magni-
tude instead of each acceleration coordinates.
Figure 18: Optical motion capture data by several swinging motions
with WiiRemote of the attack front, left and right. Red colored
arrows are defined by a evaluation function for a target vector.
In this case, we compared the data outputs to analyze a behavior of
WiiRemote position and acceleration with an optical motion cap-
ture (Fig.5.1). The captured motions are “Attack Front”, “Attack
Left” and “Attack Right” with the little differences of the path of
swinging. It was very difficult to classify only with WiiRemote.
But now we can design an evaluation functions with the path of
motion capture data. In our case of sword play, the allowance for
each motions are not strict. If the system misunderstands a “attack
left” motion as “attack front” in 10 percent for inputs, it may not be
critical. Because, an important action like “special attack” should
have a sharper magnitude then other actions even in normal life.
Estimating the magnitude is also important to define the level of
attack, but it is also important to integrate it with graphic anima-
tions. The synchronization feelings are better with the best fitted
animation playback speed to player’s continuous motions. A strong
magnitude means a faster motion or a longer stroke. If we do not
care about that, a player inputs a same motion continuously but the
animation behavior on the screen always seems to be late. A small
motion input should be assigned to a small animation to fit the du-
ration.
6 Conclusion
As a first study of WiiMedia, we developed three applications for
several user categories The first application is ”AceSpeeder2Wii”
for teens. It is a basic usage of the WiiMedia:APIs on an actual
game product. It also told us a new problem which is movements
conflicting with physics, with wireless game controllers.
The second one is ”Papier Poupee Painter”, an real-time paper and
stroke simulation drawing software with a doll interface. The input
methods for drawing will be useful in the game industry. It should
be further developed in order to establish a better way of interaction.
Also it is very interesting to improve the drawing software using a
competence of paper and stroke simulation engine and WiiRemote.
Figure 19: Player testing of “JaWii’s Virtual Fencing”
The third one is ”JaWii’s Virtual Fencing” for 7 to 12 years old
children. It is a sword play fighting action game with a famous
dramatist Alfred JARRY. Especially it was created with a city and
a professional fencing player to archive a classic sword play with
real-time animations.
There are a lot of problems in the motion detection with the WiiRe-
mote. In this article, it was described as precise information with
some solutions and ideas.
7 Future work
The WiiMedia project is just getting started. We will continue to
improve our techniques especially for motion and position detec-
tion. For better motion detection, some machine learning algorithm
will be useful. For example, Support Vector Machine (SVM) or
context learning like hand writing recognition techniques may be
applied.
On the other side, now we have some applications that use the Wi-
iRemote. We have basic APIs, utility APIs and Virtools building
block codes. They will be given as open source to share the knowl-
edge. In fact, a lot of students in our laboratory used WiiRemote for
their projects. I would be very happy if this article and knowledge
may help a project that is motivated with a new human interface in
entertainment like the WiiRemote.
8 Acknowledgement
This research was partially supported by JSPS (Japan Society for
the Promotion of Science) and city of Laval(Project JaWii). We
would like to express our thanks to Kyoko Murakami (Papier-
PoupeePainter), Takahiro Nakatani (AceSpeeder2), Benoit Pince-
maille (JaWii’s motion, supervising of escreme), E.Kako (HID
driver), Gill Van Herzele, Alexandre El Ayoubi and all students in
MNRV, especially Mehdi Saleh.
References
BLESER, T. W., SI BE RT, J. L. , AND MCGEE, J. P. 1988. Charcoal
sketching: returning control to the artist. ACM Trans. Graph. 7,
1, 76–81.
BLU ESOLE IL. http://www.bluesoleil.com/.
BOU GUILA , L. , SATO, M., HAS EGAWA, S., NAO KI , H., MAT-
SU MOT O, N. , AN D TOYAMA , A. 2002. A new step-in-place
locomotion interface for virtual environment with large display
system. In ACM SIGGRAPH 2002 Emerging Technologies.
CAI LLOIS , R. 1958. Les jeux et les hommes. Gallimard, ISBN2-
07-035125-4.
DUR AND, F., OSTROMOUKHOV, V., MILLER, M., DURAN LE AU,
F., AN D DORSEY, J. 2001. Decoupling strokes and high-level
attributes for interactive traditional drawing. In Proceedings
of the 12th Eurographics Workshop on Rendering Techniques,
Springer-Verlag, London, UK, 71–82.
E.KAKO. http://www.kako.com/.
GLOV EPIE. http://carl.kenner.googlepages.com/glovepie.
HUIZINGA, J. 1939. Homo ludens. Vom Ursprung der Kultur im
Spiel. Rowohlt Verlag, 1994, ISBN 3-499-55435-6.
HUIZINGA, J. 1955. Homo ludens: A study of the play-element in
culture. Boston, Beacon Press , ISB.
KAL NINS, R . D., MARKOSIAN, L., ME IE R, B. J., KO WALSKI,
M. A., LEE , J. C. , DAVID SO N, P. L., W EBB, M., HU GHES,
J. F., AN D FIN KEL ST EI N, A. 2002. Wysiwyg npr: drawing
strokes directly on 3d models. In SIGGRAPH ’02: Proceedings
of the 29th annual conference on Computer graphics and inter-
active techniques, ACM Press, New York, NY, USA, 755–762.
KYOK O MURAKAMI, REIJI TS URUNO , E. G. 2006. Natural-
looking strokes for drawing applications. The Visual Computer
Volume 22, Number 6, pp.415–423.
MURAKAMI, K., TSURU NO, R., A ND GE NDA , E. 2006. Natural-
looking strokes for drawing applications. Vis. Comput. 22, 6,
415–423.
RUDOLF, D., MO UL D, D., A ND NEUFE LD , E. 2005. A bidi-
rectional deposition model of wax crayons. Computer Graphics
Forum 24, 1, 27–39.
SALEN, K., AN D ZIMMERMAN, E. 2003. Rules of Play: Game
Design Fundamentals. ISBN: 0262240459.
SHIRAI, A., SATO, M., KUME, Y., A ND KUSAHARA, M.
1998. Foot interface: Fantastic phantom slipper. In ACM SIG-
GRAPH’98 Enhanced Realities, Orlando, Florida USA.
SHIRAI, A., HASEGAWA, S., SATO, M., A ND KOIKE, Y.
2002. “penguin hockey”:a virtual reality game system for
children. International Workshop on Entertainment Comput-
ing(IWEC2002) Workshop Note, Makuhari, Japan,, pp.439–446.
SHIRAI, A., KOBAYASHI, K., KAWAKITA, M.,
HASEGAWA, S., M ASAYUKI, M. N. , A ND SATO. 2004.
Entertainment Applications of Human-scale Virtual Reality
Systems. The fifth Pacific-Rim Conference on Multimedia
(PCM2004) Special Session:Human-scale VR and Interaction,
Tokyo, Japan, pp31-38,.
SHIRAI, A., DOMINJON, L., TAK AHASH I, M ., MIYATA, K.,
SATO, M., AND RICHIR, S. 2005. Robogamer: Develop-
ment of robotic tv game player using haptic interface and gpu
image recognition. Proceedings of SIGCHI International Con-
ference on Advances in Computer Entertainment Technology
(ACE2005), ACM digital library, pp.471–472.
SHIRAI, A., NAK ATANI, T., GESLIN, E., KIMURA , H. , TAKA-
HA SHI, M. , MI YATA , K. , AN D RICHIR, S. 2006. Physical
evaluation of new computer entertainment interfaces under natu-
ral play conditions. Sandbox: an ACM SIGGRAPH Video Game
Symposium, poster.
SHIRAI, A . 2004. Entertainment systems. Transaction of art and
science, Volume 3, No1,, pp.22–34,.
SOU SA, M. C., AN D BUC HAN AN , J. W. 2000. Observational
models of graphite pencil materials. Computer Graphics Forum
19, 1, 27–49.
TAKAG I, S ., NAKA JI MA , M., AND FU JI SH IRO , I. 1999. Volumet-
ric modeling of colored pencil drawing. In PG ’99: Proceedings
of the 7th Pacific Conference on Computer Graphics and Appli-
cations, IEEE Computer Society, Washington, DC, USA, 250.
TAKAHIRO NAKATANI, R. Acespeeder2,
http://www.raingraph.com/.
WII LI.ORG. http://www.wiili.org/.
WIINREM OTE. http://onakasuita.org/wii/.
