[iva] Talking Robotics #1 - Michael Jae-Yoon Chung, University of Washington

[Talking Robotics]

\\\\\\\\\\\\\\\\\\\\\\\\
CONCEPT
How is it called? “Talking Robotics”. See more on our website:
talking-robotics.github.io

What is it? Talking Robotics is a series of talks to reflect, dialogue, and
network. We focused on the field of Robotics and its interaction with other
relevant fields, such as Artificial Intelligence, Machine Learning, Design
Research, Human-Robot Interaction, among others.

Where is it? It is held virtually. Since COVID-19, researchers have been
unable to normally attend conferences and network as we used to. This
situation is expected to last some time. We created a virtual space where
robotics researchers can present their work and engage with the audience to
promote collaborations, feedback, and a sense of community.

When does it happen? We host virtual sessions bi-weekly, i.e., every other
week. We allocate 30min for presentation time and 30min for discussion and
networking. Our first session will be on Sept 4, Friday at 9am PDT. Details
are below.
\\\\\\\\\\\\\\\\\\\\\\\\

\\\\\\\\\\\\\\\\\\\\\\\\
TALKING ROBOTICS #1

SAVE THE DATE
September 4, Friday, 9am Pacific Time
Zoom link: https://washington.zoom.us/j/95943842399
Meeting ID: 959 4384 2399
*The talk and discussion will be recorded and shared publicly.

\\\\\\\\\\\\\\\\\\\\\\\\

SPEAKER
Michael Jae-Yoon Chung is a graduate student at the University of
Washington whose research focus is on end-user programming for authoring
interactive robot behaviors.

TITLE AND ABSTRACT
Iterative Repair of Social Robot Programs from Implicit User Feedback via
Bayesian Inference

Creating natural and autonomous interactions with social robots requires
rich, multi-modal sensory input from the user. Writing interactive robot
programs that make use of this input can demand tedious and error-prone
tuning of program parameters, such as tuning thresholds on noisy sensory
streams for detecting whether the robot’s user is engaged or not. This
tuning process dealing with low-level streams and parameters makes
programming of social robots time-consuming and inaccessible for people who
could benefit the most from unique use cases of social robots. To address
this challenge, we propose the use of iterative program repair, where
programmers create an initial program sketch in our new Social Robot
Program Transition Sketch Language (SoRTSketch), a domain-specific language
that supports expressing uncertainties related to thresholds in transition
functions. The program is then iteratively repaired using Bayesian
inference based on corrections of interaction traces that are either
provided by the programmer or derived from implicit feedback given by the
user during the interaction. Based on experiments with a human simulator
and with 10 human users, we demonstrate the ease and effectiveness of this
approach in improving social robot programming and program outputs that
represent three common human-robot interaction patterns. We also show how
our approach helps programs adapt to environment changes over time.

TEASER VIDEO
https://youtu.be/LbcMxC_3x1U

\\\\\\\\\\\\\\\\\\\\\\\\

TALKING ROBOTICS ORGANIZERS

Patrícia Alves-Oliveira, University of Washington (USA)
Silvia Tulli, Instituto Superior Técnico (Portugal) and Soborne University
(France)
Miguel Vasco, Instituto Superior Técnico (Portugal)
Joana Campos, Instituto Superior Técnico (Portugal)

TWITTER: @group_robotics


Talking Robotics
 [image: Twitter] <https://twitter.com/group_robotics> [image: YouTube]
<https://www.youtube.com/channel/UC-ildOqpbG04qaNXbhNYqMQ> [image: General
(Enter full link)] <https://gaips.inesc-id.pt/hri-reading-group/>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.uni-bielefeld.de/mailman2/unibi/public/iva-list/attachments/20200831/96c614f1/attachment.html>