Reparameterized Variational Divergence Minimization for Stable Imitation
Abstract
While recent state-of-the-art results for adversarial imitation-learning algorithms are encouraging, recent works exploring the imitation learning from observation (ILO) setting, where trajectories \textit{only} contain expert observations, have not been met with the same success. Inspired by recent investigations of -divergence manipulation for the standard imitation learning setting(Ke et al., 2019; Ghasemipour et al., 2019), we here examine the extent to which variations in the choice of probabilistic divergence may yield more performant ILO algorithms. We unfortunately find that -divergence minimization through reinforcement learning is susceptible to numerical instabilities. We contribute a reparameterization trick for adversarial imitation learning to alleviate the optimization challenges of the promising -divergence minimization framework. Empirically, we demonstrate that our design choices allow for ILO algorithms that outperform baseline approaches and more closely match expert performance in low-dimensional continuous-control tasks.
Cite
@article{arxiv.2006.10810,
title = {Reparameterized Variational Divergence Minimization for Stable Imitation},
author = {Dilip Arumugam and Debadeepta Dey and Alekh Agarwal and Asli Celikyilmaz and Elnaz Nouri and Bill Dolan},
journal= {arXiv preprint arXiv:2006.10810},
year = {2020}
}