Active Improvement of Control Policies with Bayesian Gaussian Mixture Model
This work addresses the challenge of enhancing robot control policies for non-expert users, though it is incremental as it builds on existing active learning and BGMM methods.
The paper tackles the problem of improving generalization in learning from demonstration by introducing an active learning framework based on Bayesian Gaussian mixture models, which was demonstrated in a reaching task with a Panda robot.
Learning from demonstration (LfD) is an intuitive framework allowing non-expert users to easily (re-)program robots. However, the quality and quantity of demonstrations have a great influence on the generalization performances of LfD approaches. In this paper, we introduce a novel active learning framework in order to improve the generalization capabilities of control policies. The proposed approach is based on the epistemic uncertainties of Bayesian Gaussian mixture models (BGMMs). We determine the new query point location by optimizing a closed-form information-density cost based on the quadratic Rényi entropy. Furthermore, to better represent uncertain regions and to avoid local optima problem, we propose to approximate the active learning cost with a Gaussian mixture model (GMM). We demonstrate our active learning framework in the context of a reaching task in a cluttered environment with an illustrative toy example and a real experiment with a Panda robot.