|
Invited Lecturers
Prof. Klaus Scherer
Universität Genf
Additional Lecturer:
Prof. Elisabeth André
Universität Augsburg
Agenda
PDF Download
Contact:
For information and/or further details please email ifa@unibw.de
Organizers:
Prof. Berthold Färber
Universität der Bundeswehr München
Prof. Gerhard Rigoll
Technische Universität München
Dr. Frank Wallhoff
Technische Universität München
|
Motivation
What makes a robot genuinely intelligent and user-friendly? Even though we taught
robots to walk or even dance, to play chess, or to recognize objects, they are still
perceived as artificial and alien in interaction situations. This has only little
relation to their technical appearance and a great deal to do with their inability
to behave really human-like. In the case of human-human interaction, we would feel
rather irritated if important nonverbal messages like a helpless facial expression,
an angry undertone or a demonstrating silence would remain ignored. However, contemporary
HRI is a largely one-sided issue characterized more or less by robots following some
programmed routines while their human peers try to adapt to this unnatural kind of
interaction the best they can.
But how can we make robots understand and react to social and emotional humane signals?
What type of modalities contains important information? Which emotional states need to
be taken into account? Which methods of signal processing or pattern classification are
best suited to this kind of data? Unfortunately, these questions are mostly being
investigated in separate research fields, with psychology focusing on emotional states
and their underlying dimensions, while electrical engineers and computer scientists
concentrate on robot control or elaborate data analysing methods.
Even though specialized in-depth knowledge is certainly important, efficient collaboration
needs at least basic insights into related research areas. Thus, we are convinced that an improved
HRI can best be achieved by combining knowledge from human as well as technical science. Accordingly,
this proposed workshop aims at bringing together experts and researchers from both fields to
provide a holistic view on HRI and to learn from each other.
Target audience
This workshop is aimed at researchers from all HRI-related disciplines who are interested
in both psychological as well as technical aspects of nonverbal communication.
Workshop Structure
I. Invited Lecturers
In line with the goals set for this workshop, two experts will give talks on psychological as well as technical aspects of HRI:
1) Prof. Scherer:
“Authenticity in emotional expression: Production and detection”
In the evolution of expression and impression of emotions, the strategic manipulation of
expressive signals plays a major role. In consequence, decoders are sensitized signs of
faking to allow a judgment of the authenticity of the other's signals expressing reactions
and intentions. Encoders do their best to pass their expression of as authentic,
especially when it is not. While a major debate in ethology and psychology over the years,
this important issue has not yet been centrally addressed in the domain of robots and
virtual agents. This contribution will raise a number of fundamental issues, especially
with regard to production rules and multimodal synchronisation.
2) Prof. André:
"Challenges for Automated Multimodal Emotion Recognition:
From Offline Analysis of Emotion Corpora to Agents with Realtime
Response to Emotional Signals"
Human conversational partners usually try to interpret the speaker's or
listener's emotional cues and respond to them accordingly. Recently, the
modelling and simulation of such behaviours has been recognized as an
essential factor for successful man-machine communication. So far, most
of this research has been concerned with the offline analysis of
available or specifically created speech corpora. Most affective-aware
applications, such as the interaction with social robots, require,
however, that the user's emotional state is recognized in real-time
while he or she is interacting with an application. In my talk, I will
present various strategies we developed in order to meet the challenges
of real-time emotion recognition. I will present a framework for smart
sensor integration that is especially suited for real-time multi-channel
emotion recognition. Core of the framework is a dynamic buffer
management that offers simultaneous and synchronized access to a pool of
data streams measuring input of connected sensors or manipulated
versions of this input. The work will be illustrated by means of
affective interfaces we developed within the European Network of
Excellence Humaine (Human-Machine Interaction Network on Emotion) and
the European projects e-Circus (Education Through Characters With
Emotional-Intelligence And Roleplaying Capabilities That Understand
Social Interaction) and Callas (Conveying Affectiveness in Leading-Edge
Living Adaptive Systems).
II. Presentations
Thereafter, researchers from the cluster of excellence “CoTeSys” (Cognition for Technical Systems)
will give short presentations on their current interdisciplinary projects dealing with different
facets of nonverbal communication comprising gaze-based interaction, implicit negotiation,
performance parameters, intention recognition or physiological patterns of non-extreme emotions.
This enables not only insights into different measuring and data analysing techniques, but also
gives an example of how interdisciplinary collaboration can be fruitfully used for the benefit of improved HRI.
III. Discussion
In the third part of the workshop the concepts of the presented studies, as well as questions
or remarks from the audience, shall be discussed with the attending experts and colleagues.
Thus, this workshop provides not only the furthering of theoretical knowledge but also the
opportunity to receive practical advice for one’s own scientific work. Even if we might not
be able to find a concluding answer, workshop attendants will surely profit from the final
discussion “HRI and nonverbal communication – where should we go and how do we get there?”.
|
