Using virtual reality in cognitive research - Università di Padova

juicebottleAI and Robotics

Nov 14, 2013 (3 years and 6 months ago)

89 views

Using virtual reality
in cognitive research
Stefano Bussolon - bussolon@psy.unipd.it
Diego Varotto - varotto@psico.unipd.it
Dipartimento di psicologia Generale
Università degli studi di Padova
laboratorio di realtà virtuale
Responsabile: prof A. Carassa
[
]
virtual reality as experimental tool
Virtual reality is a very promising
technology that can be used in many
different experimental researches.
Researches in cognitive representation
of space could exploit at best facilities
provided by virtual reality and, on the
other hand, can better evaluate pros
and cons, limits and capabilities of the
new medium.
[
]
practical problems
in spatial cognition research
To test hypothesis concerning the cognitive
representation of an environment it is necessary to
face many practical problems: experiments have to
be carried into buildings subjects never explored
before, free of undesired labels and landmarks;
buildings must be large and complex enough to
require a non trivial representation and, during
experimental session, have to be uninhabited.
Other problems arise during experiments:
movements and actions of the subjects can not be
precisely recorded, and data has to be rewritten in a
computer format to compute quantitative and
statistical analysis.
[
]
the advantages of virtual reality
In spite of it’s drawbacks, virtual reality could offer
the solution of many of the problems just exposed.
Experimenters can create ad hoc buildings with the
desired furnishings, where subjects can navigate,
explore, search; the software can directly log
movements and record actions with the desired
spatial and temporal resolution. Researchers can
create types of interaction that in real word can not
be realized.
[
]
the main goals of our work
The present work describes the realization of a
virtual reality environment used to test some
experimental hypothesis concerning the spatial
cognition.
The main concerns of our work were the following:

create a closed large scale virtual
environment
,
complex enough to require a non trivial spatial
representation;

manage the kind of
interaction
of the participants
inside the environment;

register
the movements of participants, and
provide a pictorial representation of their walk.
[
]
the architecture of the building
The environment
we created is a
relatively complex
one: a two floors
building with a
cross shaped plant.
The environment is
empty and
therefore no
furnishing can be
used as landmark.
[
]
many different path can be followed
during the building exploration
There are 4 stairs
connecting the 2 floors
of the building, one for
each side of the cross
plant. The sides of each
floor are connected by a
central room. Each side
of the building can be
reached by different
ways. This allows
participants to use
shortcuts of previous
learned ways.
[
]
different kinds of exploration:
active vs. passive
Our second concern was to manipulate the kind of
exploration of participants. We realized two kinds of
exploration: active vs. passive.
[
]
In the so called active condition
they can freely walk inside the
building. In the so called passive
exploration, on the other hand,
participants were required to
follow an avatar.
The avatar is a virtual puppet able
to walk inside the environment
driving the participants to the
different areas.
the avatar’s behaviour
In the passive condition
participants are nonetheless
free to move inside the
building, even if they where
required to follow the avatar.
The avatar waits for the
participants to avoid they loose
it, and stops it’s walk till they
don’t come close enough. The
avatar’s behavior guarantees a
good degree of freedom of the
participants’ movements even
in the passive exploration.
[
]
Changes made on the basic
environment are inherited
The exploration of a complex environment in a
passive mode requires the creation of a large
number of paths the avatar has to follow: in
our experiments we used 13 different paths.
The software we used provides the facility to
dynamically create the required paths, to save
them and to generate ex novo the desired
"avatar environment".
[
]
Changes made on the basic
environment are inherited
[
]
Bussolon, Varotto:
Using virtual reality in cognitive research
the basic environment
path 1
path 2
path n
the
avatar 1
environment
the
avatar 2
environment
the
avatar n
environment
The exploration of a complex environment in a
passive mode requires the creation of a large
number of paths the avatar has to follow: in our
experiments we used 13 different paths.
The software we used
provides the facility to
dynamically create the
required paths, to save
them and to generate
ex novo the desired
"avatar environment".
in the passive condition the
avatar follows 13 different paths
[
]
participants that explore the environment in the
passive mode have to follow the avatar trough
13 different paths.
Bussolon, Varotto:
Using virtual reality in cognitive research
To optimize the
work minimizing
errors we saved the
paths in distinct
text files. we
created 3 macros to
create, save and
load each path.
the avatar walks toward the next
step and waits for the participant
[
]
The required behavior of the avatar is complex:
it should walk trough winding paths, going up-
and downstairs, and should wait for the
participant.
Bussolon, Varotto:
Using virtual reality in cognitive research
The path is divided into steps. The task of the
avatar is to walk from step 0 to step 1,
standing there till the participant is close to it,
then walk from step 1 to step 2, and so on.
When walking the avatar calculates the
distance left from the target step. When over
the target (distance < 10cm) it stands, and it
calculate the distance of the participant. When
the participant is close (distance < 1m) it walks
toward the next step.
Use of augmented reality to
improve the interaction
[
]
The objects inside the building behave as we
expect: to open a door, participants have to fire
with the joystick to the andle.
Bussolon, Varotto:
Using virtual reality in cognitive research
Sometimes the
realistic behaviour
leads to problems:
the avatar is unable
to open the doors,
and the creation of
automatic doors
was requried.
collect the data
Virtual reality allows experimenters to register
movements of participants with a very high temporal
and spatial resolution: for each session of each
participant we saved a log file containing the
position of the subject with a temporal resolution of
500 ms. The high temporal and spatial resolution of
the data collected constitutes one of the most
definitive advantages of virtual reality: data are also
ready to be used, so avoiding expensive work of
data-entry.
[
]
Saving movement of
participants on a log file
[
]
One of the greatest advantages of the virtual
reality is the opportunity to register with a high
precision the position of the participant in the
environment. We used a function to log, 2 times
for each second, the position of the participant,
with x, y z and yrot coordinates. Yrot represents
the rotation of the head on the y axis. This leads
us to know which rooms were explored and in
which order.
Bussolon, Varotto:
Using virtual reality in cognitive research
A java applet is used to render a
pictorial view of the paths
[
]
We use a java applet to render a pictorial view
of the way followed by the participants. The map
represents the position of the participant at
each
Bussolon, Varotto:
Using virtual reality in cognitive research
500 ms. The color
represents the
time: yellow is used
at the beginning of
the session, red at
the end. The little
lines show the
direction of the
sight.
a pictorial view of the participant’s
movements
To provide a qualitative
analysis of the data we
created an external
application devoted to
read the log files and to
draw a pictorial view of
the movements of the
participant.
[
]
conclusions
Virtual reality has many limits and is still a very
expensive technology, but the many, meaningful
advantages it provides candidate it to an important
future role in some areas of cognitive research.
[
]