HOMERE: a Multimodal System for Visually Impaired People

to Explore Virtual Environments

∗ ∗ ∗
Anatole Lécuyer Pascal Mobuchon Christine Mégard
CEA List ONDIM CEA List
[email protected] [email protected] [email protected]

Jérôme Perret Claude Andriot Jean-Pierre Colinot

HAPTION CEA List PSA Peugeot Citroën
[email protected] [email protected] [email protected]

Abstract HOMERE - Haptic and audiO Multimodality to

Explore and Recognize the Environment - is a VR system
This paper describes the HOMERE system: a focused on three main applications for blind and visually
multimodal system dedicated to visually impaired people impaired people:
to explore and navigate inside virtual environments. The • Preparation to the visit of an existing real site
system addresses three main applications: the preparation (touristic site, administrative building, airport, train
to the visit of an existing site, the training for the use of station, etc),
the blind cane, and the ludic exploration of virtual • Training for the use of the blind cane within a
worlds.∗ reeducation system in “safe” conditions,
The HOMERE system provides the user with different • Ludic exploration of virtual worlds.
sensations when navigating inside a virtual world: a force
feedback corresponding to the manipulation of a virtual To reach this aim, the HOMERE system proposes an
blind cane, a thermal feedback corresponding to the approach based on multimodal immersion of the user
simulation of a virtual sun, and an auditory feedback in inside a virtual environment on scale 1. HOMERE
spatialized conditions corresponding to the ambient addresses different modalities of the haptic sense - with a
atmosphere and specific events in the simulation. A visual force feedback corresponding to the manipulation of a
feedback of the scene is also provided to enable sighted virtual blind cane and a thermal feedback corresponding
people to follow the navigation of the main user. to the simulation of a virtual sun in the simulation - and of
HOMERE has been tested by several visually impaired the auditory sense - with various sounds corresponding to
people who were all confident about the potential of this the ambient atmosphere and to specific events in the
prototype. simulation.

1. Introduction This paper begins with a description of previous work

in the field of the use of VR for visually impaired people.
Virtual Reality (VR) systems are generally assimilated Then it describes the different parts of the HOMERE
to the use of a head mounted display or another immersive system. It gives the results of an informal evaluation of the
visual interface. This definition - based essentially on prototype made by several visually impaired people. It
visual restitution - disables visually impaired people from ends with a conclusion and an extension to future work.
using VR systems and prevents them from immersing
themselves into virtual worlds. 2. Previous Work
However, new kinds of VR interfaces have recently
appeared, addressing new human sensory modalities other In everyday life, how do blind people experience the
than vision. These are haptic interfaces [6] [12] and information treated with vision by sighted people when
auditory ones [26]. They imply other means of exploring they navigate? The haptic channel is exploited via the
Virtual Environments (VE) and virtual objects which can white cane to scan the environment, detect obstacles and
be used by visually impaired people [32] [7]. identify distances between him/her and the obstacles [21].
It is also exploited via the palm and fingers to recognize
∗ the shape and texture of objects with precision and via the
Corresponding Authors: Anatole Lécuyer or Christine Mégard. CEA
legs for the ground surface [21]. The proprioceptive sense
LIST, DTSI/SRSI/LCI, BP 6, F92265 Fontenay-aux-Roses Cedex,
France. allows perceiving information about displacement of the
body and position of the limbs [31]. The auditory channel

Proceedings of the IEEE Virtual Reality 2003 (VR’03)

1087-8270/03 $17.00 © 2003 IEEE
 provides information about external events, presence of feedback joystick [21] or a pen-based device [19] [31].
other people, material of objects and distances in the The haptic device is used to display the “realistic” contact
surrounding space [21]. Sounds may also be used to give with the geometric surface of the objects [31] or a
spatial orientation, alert, detect and avoid obstacles [16]. “simplified” effect - for example a vibration when
The blind people have maybe a “sixth sense” - the sense crossing a virtual wall [35]. Researchers [31] [21]
of presence - which is the ability to be aware of the observed that haptic feedback and haptic exploration
presence of an object in the environment [29]. enabled and helped blind people to develop cognitive
Multimodality (i.e. the combination of several sensory maps of the environment.
modalities) generally provides both blind and sighted With the help of these exploratory studies, researchers
people with the best perception of the environment and made several remarks about the development of haptic-
allows the successful accomplishment of everyday tasks based VR systems for blind people. It seems important to
[18] [36]. display reference points to the user [8]. This system of
Nowadays, a wide number of systems based on haptic reference must remain stable and easy to find [33]. The
and/or auditory feedback are developed and problem of understanding the scale factor between the
commercialized to guide blind people when they navigate virtual and the real environment may occur [31]. At last,
in the real world [1] [3] [11]. As an example, different the interaction with the haptic interface must be intuitive
types of “augmented” or “smart” canes offer the and easy to understand [8]. Even in this case, the persons
possibility to detect obstacles in the real environment and may have different mental representations of this
to generate audio or haptic feedback [1]. interaction [8].
Virtual reality technologies were applied from the very
beginning to the assistance of people with special needs
3. The HOMERE System
[22] [34]. Virtual reality systems were developed for them
to: develop new skills (learn how to cross streets [16] or
The user of the HOMERE system is immerged inside
communicate more easily [22]), train for specific tasks
the virtual model of the Cité des Sciences et de l’Industrie
(use of the wheel chair [15]) or simply for fun (interactive
of La Villette in Paris (see Figure 1). A virtual guide is
games [22]).
present and assists orally the user during the navigation.
The recent diffusion of haptic [6] [12] and auditory
The user navigates inside the virtual model on a pre-
[37] interfaces enabled developments of VR systems for
defined path. This pre-defined path corresponds to the
blind people [7]. Haptic feedback has been proposed to
path actually dedicated to the guidance of blind people
blind people for: reading (tactile stimulation of Braille text
through the real Cité des Sciences et de l’Industrie. In the
[27]), drawing and artistic creation [20], the display of
Cité, this path is indicated by a specific surface on the
“visual information” (haptic “translation” of 3D or 2D
ground which blind people can easily recognize. The
images [18] [32]), education (haptic display of maths and
choice of a pre-defined path system was made by experts
graphs [33] [38]) and Computer Human Interface (tactile
of the Laboratory of Ergonomic Computer Science of
access to the web or other hypermedia by translating
University of Paris V and specialists of the Cité. This
traditional GUI into haptic ones [25] [13]). Haptic
choice was made first to prevent blind users from getting
Interfaces have also been used for the study of blind
lost in the virtual mock-up, and second because the aim of
people’s perception (studies on haptic perception of space
the HOMERE prototype is to test the possibility for a
[21] and objects [17] [8]), re-education and training
blind people to learn easily one pathway inside a virtual
(mobility training [21], training for the use of the blind
environment.
cane [35]) and finally gaming (memory game [33] or
The user sits on a stool in the center of the system (see
battleship game [32]). In most of these applications, the
Figure 2). He/she manipulates a virtual blind cane with the
haptic feedback is combined with an auditory one.
dominant hand and controls the navigation
Auditory feedback is specifically used for reading,
(forward/stop/backward) with the non-dominant hand.
displaying visual information in a verbal or non-verbal
During the navigation, the system provides the user
manner [10], navigation in hypermedia and/or virtual
with 4 different types of sensory feedback:
environments [37] [26], or gaming (auditory Quake [2]).
The use of haptic interfaces by blind people to • a force feedback corresponding to the manipulation
navigate in virtual environments has been recently studied of a blind cane
[21] [19] [18] [35] [31]. Virtual objects or virtual worlds • a thermal feedback corresponding to the simulation
are sometimes represented in 3D [21], but more generally of the sun
in 2D [19] [18] [35] [31] - for example by displaying a 2D • an auditory feedback of the ambient atmosphere and
cutting-plane of a building [19] [35]. The VE is directly of other specific events
explored with the hand. The haptic interface used is: a
tactile matrix [18], an exoskeleton glove [35], a force

Proceedings of the IEEE Virtual Reality 2003 (VR’03)

1087-8270/03 $17.00 © 2003 IEEE
 • a visual feedback which is proposed to partially- 3.1. Force Feedback
sighted people or to sighted people who want to
follow the navigation of the main user in the VE The force feedback of the HOMERE system consists in
providing the user with the sensation of manipulating a
The different types of sensory feedback and their blind cane. With the virtual cane, the user can detect
implementation are now detailed separately in the obstacles and appreciate the textures on the ground or on
following paragraphs. some objects of the VE.
To reach this aim, the system uses a force feedback
arm: the VIRTUOSE 3D [12] (see Figure 3). This
interface is a 6 DOF in input and 3 DOF in output device,
commercialized by the HAPTION company. The
workspace of the extremity of the VIRTUOSE 3D is of
42x49x92cm3. The maximum peak force of the device is
of 34N, and its maximum continuous force is of 11N.
In order to provide the user with a very intuitive
interface, the “prop1” of a blind cane has been placed at
the extremity of the VIRTUOSE (see Figure 3). The
length of the cane prop is of 50cm - i.e. shorter than a real
blind cane. This prop is also useful to make up for the lack
of torque feedback of the VIRTUOSE 3D. Thanks to the
lever-effect that is naturally induced by the use of a cane,
the force feedback of the device implies a partial torque
feedback.
The different types of haptic effects simulated with the
Figure 1. Virtual Scene: VIRTUOSE are:
the Cité des Sciences of La Villette • Collision Feedback: simulates a shock between the
virtual cane and another object of the VE.
• Texture Feedback: simulates the texture of the
ground or of an object in the VE.
• Navigation Effects: haptic effects related to the
motion of the user during the navigation (i.e.
forward, backward and turn effects). The haptic
navigation effects correspond to a force feedback
applied in the direction corresponding to the user’s
motion. When the user rotates, the cane is slightly
pulled to the right or left depending on the direction
of the turn. When the user advances or when he/she
goes back, the cane is pulled forwards or backwards.

The haptic rendering of the contact of the blind cane

with an object of the VE is performed in 4 steps: (1)
collision detection, (2) dynamic simulation of the VE, (3)
computation of the force feedback, (4) perturbation of the
force feedback due to the texture of the object.
The computation and control of the force feedback
(steps 1, 2 and 3) refer to previous work in the field of
collision rendering [23] [5]. It uses the model of a virtual

1
Hinckley et al. [14] have proposed the notion of “passive interface
prop” for the design of 3D user-interfaces. This represents a “physical
manipulation of familiar real-world objects in the user’s real
Figure 2. The HOMERE System environment” [14].

Proceedings of the IEEE Virtual Reality 2003 (VR’03)

1087-8270/03 $17.00 © 2003 IEEE
 link - made of a spring and a damper - between the haptic The frequency of the force feedback of the system is of
device and the virtual object representing the blind cane. approximately 500Hz.

3.2. Thermal Feedback

The HOMERE system provides the user with a thermal

feedback corresponding to a “virtual sun2”. This thermal
feedback depends on the orientation of the user according
to the sun in the simulation. It may thus be used as a
thermal compass.
The thermal feedback of HOMERE is generated by 12
infrared lamps of 150W each. The lamps are located on a
circular structure of a 2-meter diameter which surrounds
the user at a 2-meter altitude (see Figure 2). The lamps are
activated by a RS232 controlled relay bank (National
Instruments Field Point).
During the simulation, only one lamp is activated at a
time. This implies that the virtual sun’s location has a
resolution of 30 degrees. When the user walks inside a
building, the lamp is turned off. The constant altitude of
the lamps implies that the sun is always simulated at the
Figure 3. Prop-Based Interaction with the same time of the day.
Haptic Device via a Blind Cane
3.3. Auditory Feedback
The perturbation of the force feedback due to the
texture of the object (step 4) is calculated by referring to When the user navigates inside the VE, the system also
the techniques described in a previous work by Basdogan provides the user with an auditory feedback in spatialized
et al. [4]. This consists in applying a graphic bump- conditions.
mapping technique to the haptic rendering. The force The HOMERE system uses a “Home Cinema” 5.1
vector is modified according to the local geometry at the system: a SONY amplificatory with a BOSE set of 4
level of the contact point - this geometry being calculated speakers positioned around the user in a symmetrical
with the gray level of the corresponding texel [4]. When manner (see Figure 2).
compared with the techniques described in Basdogan et al. The auditory feedback consists in 4 different types of
[4], the algorithm proposed in the HOMERE system sound:
disturbs only the tangential component of the reaction
• Ambient atmosphere: children playing, people
force, since it has been previously shown that the
talking, street sounds, elevators, etc. The variation of
tangential component dominates the normal one in the
this sound provides information about the user’s
perception of macroscopic texture [28] [24]. The Figure 4
motion.
shows several images of texture that were mapped onto
the objects of our haptic database. • Oral comments of the virtual guide: “mind the
stairs”, “there is an elevator on your left”, etc. These
sounds are activated in some specific places of the
virtual site.
• Footstep of the virtual guide: when the user is in
motion, a sound of footstep is activated. This sound
corresponds to the footstep of the virtual guide who
is located 4 meters ahead of the user in the
simulation. This sound represents an indicator of the

2
The presence of a “virtual sun” to facilitate the wayfinding has been
proposed by Darken and Sibert [9] within a VE based exclusively on
visual feedback. This solution has apparently provided good results [9].
However, another experiment made by Rudle et al. [30] showed that the
Figure 4. Images Used for presence of a visual compass - a symbolic one - in a VE did not
Bump-Mapped Haptic Textures ameliorate the spatial knowledge of this VE.

Proceedings of the IEEE Virtual Reality 2003 (VR’03)

1087-8270/03 $17.00 © 2003 IEEE
 speed of motion as well as of the direction of the The frame rate of the visual feedback is of
turns on the pre-defined path (if the sound goes to approximately 20Hz.
the left, it indicates that a turn to the left is
approaching). 3.5. Interaction and Navigation
• Haptic related sounds: these are collision and texture
sounds. They are activated when the virtual cane The main user navigates inside the virtual model along
collides and when it is rubbed along the ground or a pre-defined path. The user keeps the possibility to go
along an object. These sounds depend on the forwards, to stop or to go backwards along this path.
material of the encountered object and on the To enter the navigation command (forward, stop or
characteristics of the physical contact (speed, backward), he/she uses a wireless gamepad: the
strength, etc). Guillemot Thrustmaster Gamepad. This gamepad is
manipulated with the non-dominant hand (see Figure 2).
The sounds were mainly recorded on the real site of the
Three buttons of the gamepad are used to enter the three
Cité des Sciences et de l’Industrie with a DAT recorder
possible choices.
system. The oral comments of the virtual guide were
During the motion, the speed of navigation remains
composed with the help of the Laboratory of Ergonomic
constant. After some preliminary testing, it was arbitrarily
Computer Science of University of Paris V.
set at 1 meter/second.
The sampling rate of the auditory feedback is of 44.1
kHz. Thus, the user of the HOMERE system uses both
his/her hands to explore the virtual model: the dominant
3.4. Visual Feedback hand to catch the virtual blind cane and the non-dominant
hand to control his/her navigation inside the environment.
A visual feedback is also available when the user is not
totally blind or when a sighted person wants to follow the As an illustration of the multimodality of HOMERE, a
navigation of the main user visually. left turn is represented successively by: a change in the
spatialization of the sound of the ambient atmosphere and
of the footstep, a pull of the blind cane to the left, and a
change of position of the virtual sun.

3.6. Software Platform

The software architecture of the HOMERE system

relies mainly on two separated applications: one “scenario
application” and one “haptic application” (see Figure 6).

The scenario application is in charge of the course of

the scenario and of the rendering of the visual, auditory
and thermal feedback. The haptic application is in charge
of the collision detection, the dynamic simulation of the
VE and the computation of the force feedback command.
Both applications are run on the same PC (Bi-Processor
PC).
A second PC (PC-286) is used to control the force
Figure 5. Visual Feedback feedback of the VIRTUOSE. It runs a control loop
which uses, as input, the haptic command sent by the
The visual feedback displays the avatars of both the “haptic application”.
main user and the virtual guide (see Figure 5). In the event
of a contact, the texture of the touched object is displayed The HOMERE system relies on the WorldToolKit
on the upper right side of the screen (see Figure 5) with (WTK) API of Sense8 for the graphic rendering. It
the corresponding impact point. also uses the VORTEX 1.5 software of CM-Labs for
The visual feedback is provided in monoscopic the collision detection and the dynamic simulation of the
conditions by using a simple workstation screen (see VE.
Figure 2).

Proceedings of the IEEE Virtual Reality 2003 (VR’03)

1087-8270/03 $17.00 © 2003 IEEE
 GamePad VIRTUOSE 3D 4.2. Results

The results of this informal evaluation are presented

and discussed hereafter in a descriptive manner.
Navigation
Scenario Application Processing

Haptic Application
4.2.1. The System

Database
Scenario
Engine
- Collision
- Physics
Global Usefulness
Usefulness of the system is obvious for most subjects.
Thermal
Rendering
Preparing a navigation in an unknown environment is the
Haptic
Rendering
Haptic
Controller
major application cited. The rehabilitation teachers who
Sound
Rendering
were present suggested that it could also be used for
Graphic
mobility training (use of the cane).
Rendering
PC 1
UC 1
PC 1
UC 2
PC 2
UC 3
Interaction with the Blind Cane
Windows 2000 VxWorks
The cane is a familiar and intuitive metaphor.
However, there are different types and models of blind
Virtual Sun Speakers Screen cane associated with different sensations and manipulation
strategies. It is necessary to take into consideration the
variability of this component.
Figure 6. Software Architecture Furthermore, the force feedback applied via the cane was
dedicated to two different uses: an active exploration of
In order to match the different input formats of the surrounding space (collisions and textures feedback)
VORTEX and WTK, two different databases are used. as well as a passive guidance (navigation effects). In the
The visual database used for WTK is made of 30K latter use, the cane may be more assimilated to the lead of
textured polygons. It was modeled using 3DSMax the leader dog. This “double metaphor” has sometimes
software of Autodesk, based on the real model of the confused the users.
Cité des Sciences et de l’Industrie of La Villette. The Workspace
haptic database used by VORTEX is simplified when Within the HOMERE system the workspace of the virtual
compared with the visual one. It is limited to the objects cane is narrower than a real one. But when the physical
located along the pre-defined path. Due to the current workspace of the device was well explained to the subject,
specifications of VORTEX 1.5, the objects of the haptic there were only a few misunderstandings (i.e. confusion
database are simple primitives: boxes, spheres, cylinders between a collision and a thrust of the VIRTUOSE).
and cones.
4.2.2. The Sensory Modalities
4. Informal Evaluation
Force Feedback
Force feedback sensations are well appreciated by most
4.1. Conditions subjects.
On the one hand, subjects have a good perception of the
Nine people with visual impairments performed an presence of obstacles despite a bad perception of the
informal evaluation of the HOMERE system. One person geometry of some objects (maybe due to the lack of
was partially blind while the 8 other people were totally torque feedback or to a lowness of the maximum stiffness
blind. Among the 8 blind people, 5 people were born- simulated by the VIRTUOSE).
blind and 3 were late blind. Each person had The different textures of the ground are also well
approximately 15 minutes to test the different possibilities perceived via the cane even though this perception is more
of the prototype and the different types of sensory naturally achieved with the foot-plant.
feedback. However, the noise level and the temperature in On the other hand, the cane is perceived as very heavy.
the evaluation hall were slightly high for the experience. This is mainly due to the fact that the gravity was not
After the testing sequence, people had to complete simulated in the VE. This implies that one had to force on
orally a questionnaire. the cane to move it and to feel textures and obstacles.

Proceedings of the IEEE Virtual Reality 2003 (VR’03)

1087-8270/03 $17.00 © 2003 IEEE
 Auditory Feedback presence of torque feedback, a harder stiffness of the
The ambient atmosphere was very well perceived even haptic device, etc. Navigation in the VE must be
though the evaluation hall was quite noisy. The sound of particularly improved. While keeping the hands on the
footstep was also appreciated for its realism. It provided system, the user must be able to control the navigation
good information about the displacement of the subject speed and direction. He/she must also be able to regulate
(speed of motion and turns). auditory feedback and other parameters of the simulation.
Thermal Feedback Changes of direction can be accompanied by the system
It appears that only few subjects have well perceived the but must be initiated by the user.
thermal compass - i.e. the virtual sun. Other demands collected during the evaluation should
This might be due to the ambient temperature during the be tested: ask for additional information when necessary,
evaluation, which was already very high. In addition, explore pavements with feet rather than with the virtual
there were few changes of direction during each cane, standing and walking while using the system (with a
navigation (and consequently few changes of position of running treadmill for example), have a global map of the
the virtual sun), and people are probably more sensitive to environment at disposal.
changes of temperature. At last, two subjects tried the
system only inside the building, when the virtual sun is 5. Conclusion
not activated.
Additional work seems necessary to test the thermal This paper has presented the HOMERE system.
compass in appropriate conditions. HOMERE is a multimodal system dedicated to visually
impaired people to explore and navigate inside virtual
4.2.3. The Navigation environments.
Three main applications for blind people are addressed
Passive Navigation by this system: the preparation to the visit of an existing
The criticism that was made the most often concerns the site, the training for the use of the blind cane and the
“passivity” of the navigation provided by the system. In purely ludic exploration of virtual worlds.
real life, a blind person is active in his/her search for The user of the HOMERE system is immerged inside a
his/her route. He/she chooses his/her route according to virtual site. He/she is addressed on different sensory
his/her objectives and to the information at his/her modalities on both haptic and auditory modes.
disposal (obstacles detected with the cane, sound In addition to previous work, HOMERE proposes an
information, previous knowledge of the environment, etc). approach based on a multimodal immersion of the user on
Thus, being displaced by the system on a pre-defined path scale 1, an intuitive manipulation via a virtual blind cane
did not suit the people who tried the HOMERE system. and a use of thermal feedback to simulate a virtual sun.
Changes of Direction An informal evaluation made by visually impaired
Changes of direction that are haptically guided (i.e. turn people showed the potential of this system. All subjects
effects) are well perceived and understood. were enthusiastic about it. The detection of obstacles with
Meanwhile they are not always appreciated. Some the cane is very efficient. Auditory cues are extremely
subjects found it too rough or too quick. Furthermore, appreciated. Active navigation inside the VE seems very
most of the subjects would prefer to detect the changes of important. Thermal feedback seems apparently difficult to
direction without being imposed and directed by the use for the moment and its implementation must be
system. improved.

Speed of Navigation Future work. Future work deals with the improvement
Some subjects found the constant speed of navigation of the haptic feedback: simulation of torque feedback by
(1m/s) too slow while others found it too fast. using a 6 DOF in output device, simulation of contact
Thus, the speed navigation must be regulated directly by with complex shapes (trimesh objects), improvement of
the user and if possible this speed must vary according to the thermal stimulation, etc. The “active” navigation must
the user’s intention. also be implemented. At last, a more quantitative analysis
of the prototype is also needed to adapt our system to its
potential users.
4.3. Perspective
Acknowledgements.
This evaluation - though incomplete - is useful for
future developments. The authors would like to thank the Institut pour la ville
Detection of obstacles with the virtual cane proves en mouvement-PSA Peugeot Citroën for initiating the
very efficient even if some improvements must be done: a HOMERE project. They would like to thank Pr. P.R.
Persiaux for his valuable remarks. They would also like to

Proceedings of the IEEE Virtual Reality 2003 (VR’03)

1087-8270/03 $17.00 © 2003 IEEE
 thank M. Apel-Muller, F. Ascher, L. Chodorge, H. [20] M. Kurze, “TDraw: a computer-based tactile drawing tool for blind
people”, Proc. of ACM conference on Assistive Technologies, 1996, pp.
Corvest and the Cité des Sciences et de l’Industrie of La
131-138.
Villette, R. Fournier and G. Uzan for their precious help [21] O. Lahav and D. Mioduser, “Multisensory Virtual Environment for
on this project. Supporting Blind Persons’ Acquisition of Spatial Cognitive Mapping - a
Case Study”, to be published in Cyberpsychology and Behaviour,
http://muse.tau.ac.il/publications/69.pdf
References [22] J. McComas, J.R. Pivik, M. Laflamme, “Current uses of VR for
children with disabilities” , Virtual Environments in Clinical Psychology
[1] http://www.robotics.com/robomenu/guidecan.html and Neuroscience: Methods and Techniques in Advanced Patient-
[2] http://www.zform.com Therapist Interaction, G. Riva, B.K. Wiederhold, and E. Molinari Eds.,
[3] P. Arcara, L. Di Stefano, S. Mattoccia, C. Melchbiorri, G. Vassura Amsterdam, The Netherlands: IOS Press, 1998, pp. 161-169.
“Perception of Depth Information by Means of a Wire-Actuated Haptic [23] W.A. McNeely, K.D. Puterbaugh, and J.J. Troy, “Six-Degrees-of-
Interface ”, Freedom Haptic Rendering Using Voxel Sampling”, Proc. of ACM
Proc. of IEEE International Conference on Robotics and Automation, Siggraph, 1999.
2000. [24] M.D.R. Minsky, “Computational Haptics: The Sandpaper System
[4] C. Basdogan, C.H. Ho, and M.A. Srinivasan, “A ray-based haptic for Synthesizing Texture for with a Force-Feedback Haptic Display”,
rendering technique for displaying shape and texture of 3D objects in Ph.D. Thesis, Massachusetts Institute of Technology, 1995.
virtual environments”, Proc. of ASME Dynamic Systems and Control [25] M.S. O'Modhrain and R.B. Gillespie, “The Moose: A Haptic User
Division, DSC-Vol. 61, 1997, pp. 77-84. Interface for Blind Persons”, Proc. of the World Wide Web Conference,
[5] P. Berkelman, R. Hollis, and D. Baraff, “Interaction with a Realtime 1997.
Dynamic Environment Simulation using a Magnetic Levitation Haptic [26] S. Morley, H. Petrie, A.M. O'Neill, and P. McNally, “Auditory
Interface Device”, Proc. of IEEE International Conference on Robotics Navigation in Hyperspace: Design and Evaluation of a Non-Visual
and Automation, 1999, pp. 3261-3266. Hypermedia System for Blind Users”, Proc. of ACM conference on
[6] G. Burdea, “Force and Touch Feedback for Virtual Reality”, John Assistive Technologies, 1998.
Wiley & Sons, New York, 1996. [27] C. Ramstein, “Combining haptic and braille technologies: design
[7] C. Colwell, H. Petrie, and D. Kornbrot, “Haptic Virtual Reality for issues and pilot study”, Proc. of the ACM Conference on Assistive
Blind Computer Users”, Proc. of ACM Conference on Assistive Technologies, 1996, pp. 37-44.
Technologies, 1998. [28] G. Robles-De-La-Torre and V. Hayward, “Force Can Overcome
[8] C. Colwell, H. Petrie, D. Kornbrot, A. Hardwick, and S. Furner, “The Object Geometry In the perception of Shape Through Active Touch”,
use of a haptic device by blind and sighted people: perception of virtual Nature, Vol. 412, 2001, pp. 445-448.
textures and objects”, Improving the quality of life for the European [29] R. Rivlin and K. Gravelle, “Deciphering the Senses, The expanding
citizen: technology for inclusive design and equality, I. Placencia-Porrero World of Human Perception”, New York: Simon and Schuster, 1984, pp.
and E. Ballabio Eds., Amsterdam, IOS Press, 1998. 72-179.
[9] R.P. Darken and J.L. Sibert, “A Toolset for Navigation in Virtual [30] R.A. Ruddle, S.J. Payne, and D.M. Jones, “Navigating large scale
Environments”, Proc. of ACM Symposium on User Interface Software ‘desktop’ virtual buildings: Effects of orientation aids and familiarity”,
and Technology, 1993, pp. 157-165. Presence: Teleoperators and Virtual Environments, Vol. 7, 1998, pp.
[10] A.D.N. Edwards, “Adapting user interfaces for visually disabled 179-192.
users”, Ph.D. Thesis, Milton Keynes, UK: The Open University, 1987. [31] S.K. Semwal, “MoVE: Mobiltiy Training in Haptic Virtual
[11] S. Ertan, C. Lee, A. Willets, H. Tan, and A. Pentland, “A Wearable Environment”, piglet.uccs.edu/~semwal/NSF2001PS.pdf, 2001.
Haptic Navigation Guidance System”, Proc. of the International [32] C. Sjöström, “The Phantasticon: The PHANToM for Blind People”,
Symposium on Wearable Computers, 1998. Proc. of the PHANToM Users Group Workshop, 1997.
[12] F. Gosselin, A. Riwan, and F. Lauture, “Virtuose 3D: a new input [33] C. Sjöström, “Designing Haptic Computer Interfaces for Blind
device for virtual reality”, Proc. of the Virtual Reality International People” , Proc. of International Symposium on Signal Processing and its
Conference, 2001. Applications, 2001.
[13] A. Hardwick, S. Furner, and J. Rush, “Tactile access to virtual [34] P. Smythe, S. Furner and M. Mercinelli, “Virtual Reality
reality on the World Wide Web for visually impaired users”, Special Technologies for People with Special Needs”, Telecommunications for
issue of IEEE Displays: Developments in tactile displays. all, P.R.W. Roe Eds., Office for Official Publications of the European
[14] K. Hinckley, R. Pausch, J. Goble, and N. Kassell, “Passive Real- Communities, catalogue no: CD-90-95-712-EN-C, 1995.
World Interface Props for Neurosurgical Visualization”, Proc. of ACM [35] D. Tzovaras, G. Nikolakis, G. Fergadis, S. Malasiotis, and M.
CHI, 1994, pp. 452-458. Stavrakis, “Design and Implementation of a Haptic VR System for the
[15] D.P. Inman, “VR Training Program for Motorized Wheelchair Training of Visually Impaired”, Proc. of the Virtual Reality International
Operation”, Proc. of the Ninth Annual International Conference Conference, 2002.
Technology and Persons with Disabilities, 1994, pp. 199-202. [36] G.C. Vanderheiden, “Use of audio-haptic interface techniques to
[16] D.P. Inman and K. Loge, “Teaching orientation and mobility skills allow nonvisual access to touchscreen appliances”, Proc. of Human
to blind children using simulated acoustical environments”, Proc. of the Factors and Ergonomics Society Annual Conference, 1996.
International Conference on Human-Computer Interaction, 1999, pp. [37] E.M. Wenzel, “Localization in virtual acoustic displays”, Presence:
1090-1094. Teleoperators and Virtual Environments, Vol. 1, Num. 1, 1992, pp. 80-
[17] G. Jansson and K. Billberger, “The PHANToM used without visual 107.
guidance”, Proc. of the PHANToM Users Research Symposium, 1999. [38] W. Yu, R. Ramloll, and S.A. Brewster, “Haptic graphs for blind
[18] Y. Kawai and F. Tomita, “Evaluation of Interactive Tactile Display computer users”, Proc. of the Workshop on Haptic Human-Computer
System”, Proc. of International Conference on Computers Helping Interaction, 2000, pp. 102-107.
People with Special Needs, 1998, pp. 29-36.
[19] H. König, J. Schneider, and T. Strothotte, “Haptic Exploration of
Virtual Buildings Using Non-Realistic Haptic Rendering”, Proc. of the
International Conference on Computers Helping People With Special
Needs, 2000, pp. 377-384.

Proceedings of the IEEE Virtual Reality 2003 (VR’03)

1087-8270/03 $17.00 © 2003 IEEE