Real time Multi-channel emotion detection for effective

advertisement
Keynote: IEEE Int’l Conf. on Online Analysis and Computing Science (ICOACS’16), Chongqing, China, May 28-29, 2016
Easy with Eve: Real time multi-channel emotion detection for effective
learning
Abdolhossein Sarrafzadeh , Prof., PhD
Head of Department of Computing, Director of Centre of Computational Intelligence for Cyber Security (CICS)
Department of Computing, Faculty of Creative Industries and Business, Unitec Institute of Technology, New Zealand
Homepage: www.unitec.ac.nz/about-us/contact-us/staff-directory/prof-hossein-sarrafzadeh
Phone: +64 9 815 4321 ext. 6040 Email: [email protected]
ABSTRACT
Many software systems would significantly improve
performance if they could adapt to the emotional state of the
user, for example if Intelligent Tutoring Systems, ATM’s and
ticketing machines could recognize when users were confused,
suspicious, frustrated or angry they could provide remedial
help so improving the service. Research leading to the
development of Easy with Eve, an Affective System in the
domain of mathematics (an Affective Tutoring System- ATS)
will be presented. Eve adapts to users by detecting their
emotion through facial expression and gesture analysis, and
can display emotion herself via a lifelike animated agent
called Eve which is currently being tested in a school in New
Zealand.
Eve enjoys a facial expression recognition component
that was developed in-house. A gesture analysis component
has been developed and will soon be integrated with Eve.
Eve’s adaptations are guided by a case-based method for
adapting to user’s states; this method uses data that was
generated by an observational study involving 3000 video
clips of human participants. The observational study, the casebased method, the life-like agent and the ATS itself and its
facial expression and gesture analysis capabilities will be
briefly presented.
Keywords: Affective Tutoring Systems, Lifelike Agents,
Emotion Detection, Human Computer Interaction, Affective
Computing
ABOUT THE KEYNOTE SPEAKER
Dr. Hossein Sarrafzadeh is Professor
and Head of Department of Computing
at Unitec Institute of Technology. He is
well linked to the ICT industry both
nationally and internationally, and has
strong links to the international
research and academic community.
Leveraging these strong networks with
industry,
academia,
and
with
government, particularly in the field of IT security and computational
intelligence, he has attracted funding to his department and his
institution. He understands global IT trends, needs and opportunities
and has a firm grasp of the capability in NZ.
Dr Sarrafzadeh has developed one of the world's first real-time
facial expression and gesture recognition systems. He has also
developed a novel vision based targeted advertising system which
has been patented in New Zealand and internationally. Dr
Sarrafzadeh has led the establishment of New Zealand’s first cyber
security centre in partnership with the National Institute of
Information and Communications Technology of Japan and has also
been instrumental in the founding and development of the Centre
for Computational Intelligence for Environmental Engineering and
the Internet of Things (in partnership with Wuhan University and
LJDY (IT Company), in China and NIWA (NZ)).
Professor Sarrafzadeh is known for his work in the areas of
intelligent systems, cyber security, IoT, adaptive user interfaces and
software engineering. He has developed one of the world’s first realtime facial expression and gesture recognition systems, and also a
novel vision based targeted advertising system patented both in
New Zealand and internationally. He has published over 140
research papers and supervised over 40 PhD and Masters theses. He
is frequently invited as a keynote speaker at international
conferences.
Download