Imitation by observation is an approach for learning from expert demonstrations that lack action information, such as videos. Recent approaches to this problem can be placed into two broad categories: training dynamics models that aim to predict the actions taken between states, and learning rewards or features for computing them for Reinforcement Learning (RL). In this paper, we introduce a novel approach that learns values, rather than rewards, directly from observations. We show that by using values, we can significantly speed up RL by removing the need to bootstrap action-values, as compared to sparse-reward specifications.
@article{arxiv.1905.07861,
title = {Perceptual Values from Observation},
author = {Ashley D. Edwards and Charles L. Isbell},
journal= {arXiv preprint arXiv:1905.07861},
year = {2019}
}
Comments
Accepted into the Workshop on Self-Supervised Learning at ICML 2019