IROS2010 Workshop on
Socio-synergistic
intelligence:
Learning
with Caregivers
Taipei,
Taiwan, October 22th, 2010
The challenge of making a robot
learn from human user/caregiver shares essential problems with other research
fields on infant development. Both of them involve issues of not only learning
motor/social skills but also simultaneous and long-lasting mutual adaptation
between the learner and the caregiver. A promising approach is a
synthetic one based on both the explanation theory and more importantly the
design theory that is expected to fill in the gap between the existing
disciplines instead of staying at one closed discipline, and to provide new
understanding of human cognitive development. However, many of existing
synthetic studies have still ignored the caregiver’s side, and therefore,
limited to the computer simulation.
This full day workshop
is intended to bridge the researchers who want to synthesize the bidirectional
processed of tutoring and learning in the real world and those who develops
hardware that enables such a bidirectional process in human-robot interaction,
and thereon explore new research direction.
List of invited speakers & titles of the talk (and
tentative time schedule)
8:45-9:00
Introduction
9:00-10:00
Chen Yu
Indiana University
“Sensorimotor Dynamics in Multimodal Child-Parent and
Human-Robot Interactions”
10:00-10:20
Coffee break
10:20-10:50
Minoru Asada
Dept. of Adaptive Machine Systems, Osaka University
Asada Synergistic Intelligence Project, ERATO, JST
“Mirror Neuron System Connects
Physical Embodiment and Social Entrainment”
10:50-11:30
Takashi Hashimoto
School
of Knowledge Science, JAIST
“Constructive
simulation for sociointelligenesis”
11:30-12:00
Yuichiro Yoshikawa
Dept. of Systems Innovation, Graduate School of
Engineering Science, Osaka University
Asada Synergistic Intelligence Project, JST
“Synthesizing
Social Development with Small Desktop Humanoids”
12:00-13:40
Lunch break
13:40-14:10
Tetsuya Ogata
Dept.
of Intelligence Science and Technology, Graduate School of Informatics, Kyoto
University,
RESTO, Japan Science and Technology Agency
“Mutual adaptive interaction between robots and human
based on the dynamical systems approach”
14:10-14:40
Yasser F. O. Mohammad
Faculty
of Engineering, Assiut University, Egypt
“Learning to Interact with the Caregiver: from
Interaction Design to Interaction Protocol Learning”
14:40-15:10
Hidenobu Sumioka
Department
of Informatics, University of Zurich, Switzerland
“Cognitive development induced by
sensorimotor contingency”
15:10-15:40
Coffee break
15:40-16:10
Takashi Minato
Asada
Project, ERATO, JST
“Development
of Humanoid Platforms for Studying Social Development”
Important Dates
Jul. 15, 2010 (extended): Submission deadline
for extended abstract
Jul. 20, 2010 (extended): Notification of
acceptance
Jul. 25, 2010: Camera ready
deadline
Oct. 22, 2010: Workshop
Submissions
Anyone who is interested in presenting contribution in this WS is invited
to submit an extended abstract using IEEE format. The page limit for the final
paper is 6 pages. Submission must be in PDF file format to yoshikawa@jeap.org
Motivation and objectives
Learning from human users or caregivers is one of the key requirements
for autonomous artificial system that can adapt to human society. It is also
one of the greatest features of human beings to develop motor/social
capabilities. Both of them involve issues of not only learning motor/social
skills/knowledge but also simultaneous and long-lasting mutual adaptation
between the learner and the caregiver. Therefore, related disciplines are not
simply robotics and AI but also brain science, cognitive science, developmental
psychology, and so on, and we share this challenge. An obvious fact is that we
have too poor and little knowledge and too superficial implementations based on
such knowledge to declare that we have only one unique solution to the mystery.
The first step to conquer this situation is to discuss how such
bidirectional and transitive processes can be synthesized. A promising approach
is a synthetic one based on both the explanation theory and more importantly
the design theory that is expected to fill in the gap between the existing
disciplines instead of staying at one closed discipline, and to provide new
understanding of human cognitive development. However, many of existing
synthetic studies have still ignored the behavior of caregivers, and therefore
stayed at the computer simulation.
In this
workshop, we revisit the synthetic approach (ex. Asada et al. 2009, IEEE
Transaction on Autonomous Mental Development) that provides a new approach to
understand how human infants could learn from their caregivers, and expands
discussion toward the principle of learning with caregivers. However, the
approach to understand the aspects of mutual adaptation in human interactions
by faithfully synthesizing seems to involve methodological difficulties.
Because, one’s behavior is unconsciously influenced by subtle differences such
as how one feels the other, which would prevent us to simplify the interaction
to be modeled. Furthermore, target phenomena to be synthesized have not yet been
clearly figured out since the results of mutual adaptation might appear in long
history of interactions. Here, we argue that introducing robots or onscreen
agent into human communication would be another synthetic approach for
understanding human mutual adaptation because creating social applications give
us targets to be modeled and has social contributions as it is. For such a new
direction, to bridge existing synthetic research to the real world, the
researchers who develop platform that can be potentially applied to model/solve
the mutual adaptation in human interaction will be also invited. Finally,
general discussion towards the new research direction of learning from
caregiver will be held.
List of topics
Infant/child robot,
Synthetic approach, Infant development, Human-robot interaction, Physical
interaction, Social learning, Human-in-loop learning, Teaching by
demonstration, Learning by imitation, Teaching by touching, Social behavior,
Caregiver, Motherise/motionese, Parental behavior, Motor development, Social development,
Long term adaptation, Reading intention, and Theory of mind
Intended audience
Any researchers who are
interested in this topic, especially who wants to extend their synthetic work
in the computer simulation to the real world, and/or who wants to explore new
research direction in the intersection between robotics and human science
Organizers
Yuichiro Yoshikawa, Osaka University / Asada Synergistic Intelligence Project, JST,
yoshikawa@jeap.org
Takashi Minato, Asada Synergistic Intelligence Project, JST
Hiroshi Ishiguro, Osaka University / Asada Synergistic Intelligence
Project, JST / ATR
Contact Info
Please contact to
Yuichiro Yoshikawa (yoshikawa@sys.es.osaka-u.ac.jp)
if you have any questions.