Abstract
Learning from Demonstration permits non-expert users to easily and intuitively reprogram robots. Among approaches embracing this paradigm, probabilistic movement primitives (ProMPs) are a well-established and widely used method to learn trajectory distributions. However, providing or requesting useful demonstrations is not easy, as quantifying what constitutes a good demonstration in terms of generalization capabilities is not trivial. In this letter, we propose an active learning method for contextual ProMPs for addressing this problem. More specifically, we learn the trajectory distributions using a Bayesian Gaussian mixture model (BGMM) and then leverage the notion of epistemic uncertainties to iteratively choose new context query points for demonstrations. We show that this approach reduces the required number of human demonstrations. We demonstrate the effectiveness of the approach on a pouring task, both in simulation and on a real 7-DoF Franka Emika robot.
Highlights
L EARNING from demonstration (LfD) offers an intuitive framework for non-expert users toprogram robots
One of the main capabilities of probabilistic movement primitives (ProMPs) lies in the task generalization, which is usually achieved by conditioning the trajectory distribution to some desired keypoints
The contributions of this paper are threefold: (i) we propose a principled methodology for deriving epistemic uncertainties in ProMPs; (ii) we propose to use a closed-form lower bound of the differential entropy of the epistemic uncertainty as an information gain metric for an active learning of ProMPs; (iii) we show the applicability of the approach on a robotic pouring task
Summary
L EARNING from demonstration (LfD) offers an intuitive framework for non-expert users to (re)program robots. One of the main capabilities of ProMPs lies in the task generalization, which is usually achieved by conditioning the trajectory distribution to some desired keypoints. It is desirable and possible to generalize with respect to a context or external variable, which is known before executing the task (such as the mass of an object or the volume of a liquid to pour), by learning the joint distribution of the context variable and the trajectory [5, 6]. Task generalization is crucial for robotic applications. This requires a set of demonstrations to provide various executions of the task, whose acquisition is often costly.
Sign-in/Register to view highlights & summary Sign-in/Register to access full text options 
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have
Schedule a callDisclaimer: All third-party content on this website/platform is and will remain the property of their respective owners and is provided on "as is" basis without any warranties, express or implied. Use of third-party content does not indicate any affiliation, sponsorship with or endorsement by them. Any references to third-party content is to identify the corresponding services and shall be considered fair use under The CopyrightLaw.
