The Understanding of Meaningful Events in Gesture-Based Interaction - Intelligent Video Event Analysis and Understanding - page 14

in response. Austermann et al. [1] investigated how people interact with both the hu-
manoid ASIMO and the dog-like robot AIBO and found that, although there was no
difference in the way commands were sent to the robots, the positive or negative
feedback were appropriate to the robot type. For example, the pet AIBO was re-
ceiving rewards similarly to a real dog by touch and encouragements while ASIMO
received polite personal expressions for gratitude after had accomplished a given
task. With great similarity with respect to human-human interactions, a study of
Mutlu et. al [35] showed that people perceived ASIMO as being more sociable and
more intellectual in a collaboration task than in a competitive one.
Even if the interface is not specifically designed to make use of such human-like
elements, the tendency of humans to anthropomorphize is inevitable. This is caused
by the emotional implication that humans tend to put into everyday activities which
also applies to computers: interacting with a computing system tends to develop a
social side as if the machine were another person (see for example the Milo and
Kate demo of Lionhead working with the technology of Microsoft Natal project 5 ).
Another possible explanation would be that anthropomorphism is used as a means
to help assimilate new technology. It is also important to note that the feedback of
a system will influence humans' responses at the affective, behavioral or cognitive
levels.
Other studies have also observed and argued that persons see and interpret feed-
back by associating and augmenting it with human-like characteristics. For example,
Ju and Takayama [20] noticed how people interpret automatic door movements as
gestures and found evidence that supports this idea. This goes back to an instinctive
reflex of human beings that interpret emotionally the actions they perceive hence the
attribution of human characteristics to non-human or even unanimated objects. In an
early study, Heider and Simmel [17] showed that people interpret objects that are
moving in their visual field in terms of acts of persons. With this respect, emotions,
motivations and other human characteristics are being attributed to unanimated
objects.
It is therefore important to consider feedback mechanisms that are not only ap-
propriate and adequate in the form they take (visual, audio, haptic, etc.) but also
with respect to the emotions they are able to induce. This will transform feedback
into a true interaction event perceived as such by the human participant and will
lead to an increase in the satisfaction of the overall interaction experience. This is
strongly connected to human-human communication where particular gestures are
expecting corresponding notification and responses and the same should be expected
from human-computer interaction. Appropriate feedback of a correctly detected and
recognized gesture can thus become an important communication event.

4Con lu ion

Gesture commands can be detected in video sequences using events that define them
in conjunction to location, posture, touch or other application or scenario specific

5

http://www.xbox.com/en-US/live/projectnatal/