Generative adversarial training of product of policies for robust and adaptive movement primitives

TL;DR

This paper introduces a generative adversarial training method for product of policies to improve robustness and adaptability in learning movement primitives from demonstrations, addressing dependencies often ignored in simpler models.

Contribution

It proposes using approximate trajectory distributions as discriminators within a GAN framework to enhance learning stability and speed, while incorporating product of Gaussian policies and ensemble methods for robustness.

Findings

Validated on a 7-DoF manipulator

Improved adaptability to varying contexts

Enhanced robustness to perturbations

Abstract

In learning from demonstrations, many generative models of trajectories make simplifying assumptions of independence. Correctness is sacrificed in the name of tractability and speed of the learning phase. The ignored dependencies, which often are the kinematic and dynamic constraints of the system, are then only restored when synthesizing the motion, which introduces possibly heavy distortions. In this work, we propose to use those approximate trajectory distributions as close-to-optimal discriminators in the popular generative adversarial framework to stabilize and accelerate the learning procedure. The two problems of adaptability and robustness are addressed with our method. In order to adapt the motions to varying contexts, we propose to use a product of Gaussian policies defined in several parametrized task spaces. Robustness to perturbations and varying dynamics is ensured with the use of stochastic gradient descent and ensemble methods to learn the stochastic dynamics. Two experiments are performed on a 7-DoF manipulator to validate the approach.