Computational Model of Humans:
Adaptation and Influencing

Proposed RSS 2019 Workshop
June 22-26, 2019
Freiburg, Germany

Freiburg, Germany

Key Facts

Length
  Full Day
Contact Information
  dlosey@stanford.edu

Abstract

Front Figure

Robots are increasingly becoming members of our everyday community. Self-driving cars, surgical and rehabilitation devices, and social and assistive robots operate alongside human end-users to carry out interactive tasks. In order for robots to transition from isolated systems to intelligent partners, however, these robots need to understand the humans they are interacting with: how to learn from human actions, how humans adapt to robots, and how robot actions can influence nearby humans.

This workshop promotes a discussion on the role of computational models of humans in robotics, and highlights the different ways in which robots can model human partners to achieve fluent, efficient, and successful interactions. We're bringing together researchers to explore challenges in computational human-robot interaction, where we plan to discuss technical advances in:

We seek contributions and insight from researchers whose interests span human modeling, human-robot interaction, and learning from humans.

List of Confirmed Speakers

Tentative Schedule

Logistics: 25 minute talks followed by 5 minute discussions.

08:50 AM - 09:00 AM Workshop Introduction
09:00 AM - 09:30 AM speaker 1
09:30 AM - 10:00 AM speaker 2
10:00 AM - 10:30 AM Discussion and Coffee Break
10:30 AM - 11:00 AM speaker 3
11:00 AM - 11:30 AM speaker 4
11:30 AM - 12:00 PM Panel w/ First Four Speakers
12:00 PM - 02:00 PM Lunch Break
02:00 PM - 02:30 PM speaker 5
02:30 PM - 03:00 PM speaker 6
03:00 PM - 03:30 PM speaker 7
03:30 PM - 04:00 PM Poster Session and Coffee Break
04:00 PM - 03:30 PM speaker 8
04:30 PM - 05:00 PM speaker 9
05:00 PM - 05:30 PM Debate w/ Last Five Speakers

Audience

This workshop will focus on two emerging areas in computational HRI: adaptation and influencing.

Adaptation considers how robots and humans modify their behavior to adjust to one another. One common example of adaptation is learning: we want robots that learn the right behavior from human partners. But there is another side to adaptation---just as the robot learns from the human, humans can also modify their behavior when interacting with robot partners. Both sides of adaptation will be discussed in this workshop.

Influencing refers to how a robot can leverage its actions to proactively change the behavior of nearby humans. Influential behavior can arise when a single human and robot are interacting---for instance, a self-driving car slowing down to cause a human-driven car to follow suit---as well as when teams of humans and robots are working to complete a task. This workshop will discuss the relationship between computational human models and influential robot behavior.

This workshop will be of interest to researchers who study computational HRI, and want to discuss, learn about, and present recent advances in adaptation and influencial behavior.

Organizers