p26 - Dynamic Imitation in a Humanoid Robot through...

Info iconThis preview shows pages 1–2. Sign up to view the full content.

View Full Document Right Arrow Icon

Info iconThis preview has intentionally blurred sections. Sign up to view the full version.

View Full DocumentRight Arrow Icon
This is the end of the preview. Sign up to access the rest of the document.

Unformatted text preview: Dynamic Imitation in a Humanoid Robot through Nonparametric Probabilistic Inference David B. Grimes Rawichote Chalodhorn Rajesh P. N. Rao Department of Computer Science and Engineering University of Washington Seattle, WA 98195 http://neural.cs.washington.edu { grimes,choppy,rao } @cs.washington.edu Abstract We tackle the problem of learning imitative whole- body motions in a humanoid robot using probabilistic inference in Bayesian networks. Our inference-based approach affords a straightforward method to exploit rich yet uncertain prior information obtained from human motion capture data. Dynamic imitation implies that the robot must interact with its environ- ment and account for forces such as gravity and inertia during imitation. Rather than explicitly modeling these forces and the body of the humanoid as in traditional approaches, we show that stable imitative motion can be achieved by learning a sensor- based representation of dynamic balance. Bayesian networks provide a sound theoretical framework for combining prior kinematic information (from observing a human demonstrator) with prior dynamic information (based on previous experience) to model and subsequently infer motions which, with high probability, will be dynamically stable. By posing the problem as one of inference in a Bayesian network, we show that methods developed for approximate inference can be leveraged to efficiently perform inference of actions. Additionally, by using nonparametric inference and a nonparametric (Gaussian process) forward model, our approach does not make any strong assump- tions about the physical environment or the mass and inertial properties of the humanoid robot. We propose an iterative, probabilistically constrained algorithm for exploring the space of motor commands and show that the algorithm can quickly discover dynamically stable actions for whole-body imitation of human motion. Experimental results based on simulation and subsequent execution by a HOAP-2 humanoid robot demonstrate that our algorithm is able to imitate a human performing actions such as squatting and a one-legged balance. I. INTRODUCTION Imitation learning presents a promising approach to the problem of enabling complex behavior learning in humanoid robots. Learning through imitation provides the robot with strong prior information by observing a skilled instructor (of- ten assumed to be a human demonstrator). This paper presents a model for exploiting this prior information about whole-body motions gathered from observing a human performance of the motion. Although the observation of the teacher is informative, there is a high degree of uncertainty in how the robot can and should imitate. Our model accounts for some of these sources of uncertainty including: noisy and missing kinematic estimates of the teacher, mapping ambiguities between the human and robot kinematic spaces, and lastly, the large m a 1 m 1 m 2 k 1 d 1 o 1 b 1 s 1 a k 2 d 2 b 2 o 2 s 2 d 3 b 3 o 3 s 3 k 3 a 2...
View Full Document

Page1 / 8

p26 - Dynamic Imitation in a Humanoid Robot through...

This preview shows document pages 1 - 2. Sign up to view the full document.

View Full Document Right Arrow Icon
Ask a homework question - tutors are online