Robotics Graduate Student Seminar: Garrett Katz, "Robotic Imitation Learning"
Tuesday, March 28, 2017
1146 AV Williams
For More Information:
301 405 4358
Robotics Graduate Student Seminars
Robotic imitation learning using a model of cause-effect reasoning
PhD candidate, Computer Science
Advisor: Prof. James Reggia
Manually programming robots is currently very difficult, impeding more widespread use of robotic systems. In response to this problem, increasing efforts are being made to develop cognitive robots that use imitation learning. With such systems a robot learns a procedure by watching a human perform the task. However, most imitation learning systems focus on accurately replicating a demonstrator’s actions rather than on obtaining a deeper understanding of why those actions are occurring. To address this issue, we have developed a framework for cognitive-level imitation learning based on cause-effect reasoning that infers the intentions of a demonstrator. As with imitation learning in people, our approach constructs a plausible explanation for a demonstrator’s actions, and generates a plan based on this explanation to carry out the same goals rather than trying to faithfully reproduce the demonstrator’s precise motor actions. This facilitates generalization of a single demonstration to new situations. Our approach has been implemented and validated empirically using a physical robot, which successfully learns and generalizes skills involving bi-manual manipulation of objects in three dimensions.
About the Robotics Graduate Student Seminars
The Robotics Graduate Student Seminars at the University of Maryland College Park are a student-run series of talks given by current graduate students.
The purpose of these talks is to:
- Encourage interaction between Robotics students from different subfields;
- Provide an opportunity for Robotics students to be aware of and possibly get involved in the research their peers are conducting;
- Provide an opportunity for Robotics students to receive feedback on their current research;
- Provide speaking opportunities for Robotics students.
This Event is For: Graduate • Undergraduate