ACM SIGCHI General Interest Announcements (Mailing List)


Options: Use Classic View

Use Monospaced Font
Show Text Part by Default
Show All Mail Headers

Topic: [<< First] [< Prev] [Next >] [Last >>]

Print Reply
Hatice Gunes <[log in to unmask]>
Tue, 7 Jun 2011 20:20:14 +1000
text/plain (155 lines)
Dear all,

For those of you working in the field of Affective Computing, we would
like to announce the following call for papers:



International Journal of Synthetic Emotions

Special Issue on

Benefits and Limitations of Continuous Representations of Emotions in
Affective Computing

Affective computing has witnessed tremendous progress over the past
decade, in all four of its core areas: recognition of human emotions by
machines, affective user modeling, modeling of emotions in agent
architectures, and expression of emotions in virtual agents and robots.

Thus far, affective computing can best be characterized as an engineering
discipline. A number of methods, techniques, algorithms and architectures
have been developed in all four of its core areas. These have enabled
achievements in emotion recognition that are beginning approach human
performance and development of increasingly affectively realistic agents
and robots.

Yet in spite of these successes, the field still lacks firm theoretical
foundations and systematic guidelines in many areas, particularly so in
emotion modeling: the construction of computational models of emotion
generation and emotion effects in agent architectures. This is in large
part due to the difficulties associated with developing theoretical
explanatory frameworks for emotions and affective phenomena in general, as
well as the difficulties in obtaining the necessary empirical data about
transient, multi-modal affective states.

Recently, there has been an increasing effort to transition affective
computing into a more scientific discipline. This is evidenced by an
increased emphasis on validating emotion models, on developing systematic
guidelines for affective modeling, and in attempts to develop standards in
the form of affective markup languages and open source development
frameworks and tools.

An important component of this effort is the increased interest in
understanding the relative benefits and drawbacks of alternative
theoretical perspectives on emotions, and the associated representations.
Three dominant theoretical perspectives on emotions have emerged: discrete
/ categorical, dimensional, and componential. In emotion modeling,
emphasis has been on the use of cognitive appraisal as a means of
generating emotions in agents, with the majority of models using the
Ortony, Clore and Collins theory (OCC). In emotion analysis and
recognition, the emphasis has been primarily on categorical
representations, and, more recently, on dimensional representations.

The purpose of this special issue is to explore the benefits and drawbacks
of these alternative perspectives on emotion in the four core areas of
affective computing, with a particular focus on continuous representations
of emotions, within both the dimensional and the componential theoretical
perspectives. The former typically using the pleasure-arousal-dominance
(PAD) dimensions, and the latter using appraisal variables associated with
cognitive appraisal.

Specific topics of interest include:

-Empirical evidence for the existence of underlying dimensions and
appraisal variables characterizing affective states
-Applicability of continuous representations for characterizing multiple
modalities of affective states
-Continuous representations of emotions in emotion recognition from facial
expressions, gestures, movement, and speech
-Identification of continuous dimensions of emotions in text and music
-Continuous representation of emotions in models of emotion generation and
emotion effects
-Use of continuous representation of affective states as a basis for
emotion expression in virtual agents and robots
-Continuous representations of affective states in affective user models
-Annotation methods and tools suitable for continuous representation of
affective states

To enhance the cohesiveness of this special issue, authors are asked to
devote a section of their manuscript to addressing the following questions
regarding the use of a particular theoretical perspective and
representation of emotion:

-Why did you choose the particular representation? What are the benefits
and drawbacks of the selected emotion representation, as compared to the
other alternatives?
-How applicable is the selected theoretical perspective and representation
to other aspects of affective computing (e.g., if your focus is on
recognition, comment on how your representation might extend to emotion
modeling; if your focus is emotion generation, comment on how applicable
your representation would be for modeling emotion effects, across multiple
-How well does the theoretical perspective and representation accommodate
multiple modalities of emotions?
-How readily are the necessary data available?
-How would you approach a validation of your approach and what data would
you need?

Submission Guidelines:

Authors are asked to submit original manuscripts, between 5500  8000
words in length, following the guidelines available on the journal

If you intend to submit a manuscript, please email your interest to the
Special Issue editors who will provide you with the instructions for


Important Dates:

Paper Submission: August 15, 2011

Acceptance Notification: September 15, 2011

Final Manuscripts Due: October 15, 2011

Special Issue: Spring 2012

Special Issue Editors:

Eva Hudlicka, Psychometrix Associates, Blacksburg, VA, US, [log in to unmask]
Hatice Gunes, Imperial College London, London, UK, [log in to unmask]


Dr. Hatice  Gunes
Intelligent Behaviour Understanding Group
Department of Computing
Imperial College London
180 Queen's Gate
London SW7 2AZ, U.K.
Phone: +44 2075941117

                To unsubscribe, send an empty email to
     mailto:[log in to unmask]
    For further details of CHI lists see