Abstract
Most current methods for learning from demonstrations assume that those demonstrations alone are sufficient
to learn the underlying task. This is often untrue, especially
if extra safety specifications exist which were not present in
the original demonstrations. In this paper, we allow an expert
to elaborate on their original demonstration with additional
specification information using linear temporal logic (LTL). Our
system converts LTL specifications into a differentiable loss. This
loss is then used to learn a dynamic movement primitive that
satisfies the underlying specification, while remaining close to
the original demonstration. Further, by leveraging adversarial
training, the system learns to robustly satisfy the given LTL
specification on unseen inputs, not just those seen in training.
We show that our method is expressive enough to work across
a variety of common movement specification patterns such as
obstacle avoidance, patrolling, keeping steady, and speed limitation. In addition, we show how to modify a base demonstration
with complex specifications by incrementally composing multiple
simpler specifications. We also implement our system on a PR2 robot to show how a demonstrator can start with an initial
(sub-optimal) demonstration, then interactively improve task
success by including additional specifications enforced with a
differentiable LTL loss.
Original language | English |
---|---|
Title of host publication | Proceedings of Robotics: Science and System XVI |
Editors | Marc Toussaint, Antonio Bicchi, Tucker Hermans |
Number of pages | 10 |
ISBN (Electronic) | 978-0-9923747-6-1 |
DOIs | |
Publication status | E-pub ahead of print - 1 Jul 2020 |
Event | Robotics: Science and Systems 2020 - Duration: 12 Jul 2020 → 16 Jul 2020 https://roboticsconference.org/ |
Conference
Conference | Robotics: Science and Systems 2020 |
---|---|
Abbreviated title | RSS 2020 |
Period | 12/07/20 → 16/07/20 |
Internet address |